it's been nearly two years since I closed the book on Project Diva Future Tone. my old roommate's now ex-girlfriend ran off with a 19 year-old mechanic at a go-karting place. whoops :( we haven't played project diva together in a while

I wrote that review under the impression that beatmania iidx was gonna take over my rhythm game grind, and for all intents and purposes it has. both future tone and megamix have unfortunately been uninstalled in order to make way for other titles, while I'll still sit down for iidx sessions at least once a week. however, around the time that review dropped, our local arcade began branching into imported rhythm games. I had been frequenting the place for a couple years by that point (also that same old roommate used to work with the arcade's owners at a barcade down the street back when we were in undergrad; small world!), and the owners had granted my request of kicking off their rhythm game selection with a real project diva cabinet. unfortunately my declining interest in the game had already taken hold, and I never really worked on transferring my skills over to the arcade version as I had initially hoped.

it wasn't until a chance meeting later that year that my interest was reignited. while visiting the arcade with friends from out-of-town and trying out a newly acquired jubeat cabinet, I cycled onto project diva to run a quick set after a girl who had been working on some hard charts. she stood beside me, enraptured in my skill, and quickly introduced herself afterwards. she was a local arts student and miku aficionado who had stumbled upon the switch game and began sinking countless hours into it, propelled even further by her later discovery that we had an arcade cabinet for the game locally. we talked for nearly an hour that day, leaning against pinball tables in this humid shop backroom, discussing our favorite songs while she asked endless questions on hand placement, chart details, and various other strategies.

arcade rhythm games pose significant barriers to those looking to excel at them. each cabinet prides itself on its foreign, sometimes unintuitive interface, all completely divorced from the dual analog and keyboard+mouse conventions we've grown accustomed to. practicing regularly requires either dumping money into a local cabinet, which may be hours away at best, or purchasing an expensive custom controller to play at home, which may require procuring a crack of the game and circumventing layers of copy-protection and environmental inconsistencies in order to run. so to get this social experience, where I could not only cheaply access a local cabinet but also train someone else to enjoy it just as much as I had over the years, was a small miracle. there's something beautiful in that that's difficult to approximate through another type of game; passing down my hard-earned knowledge about this confusing, niche machine to someone else, watching her progress through the ranks of extreme and exex charts, and crystallizing my own insights into something I could actually express to someone else.

as of the last few weeks, she finally surpassed me by clearing Disappearance of Hatsune Miku exex. its notoriously precise ending walled me back when I was playing the game daily, leaving it one of the charts I still have never cleared (along with both ex and exex intense voice and denparadigm exex, plus a couple other ones I've never played much bcs I never owned them at home). a magfest acquaintance of mine who streams project diva regularly and owns his own private cabinet still struggles to clear it every time I've watched him play it; clearing this is such a massive accomplishment. a lot of my study of the game for the last year has been spurned on by her talent and passion, sending me back to the lab to learn parts I had previously skimmed over and spool out new strategies. at the same time, I've transitioned into an adjunct teaching position that's monopolized my time in a way I haven't experienced since high school. I think it's time to close the book again on this game, even though I'm certain I'll play it again in the future.

much of my writing this last year has been on action and puzzle games; genres that are often analyzed through the lens of their depth and capability of decision-making. rhythm games are fully execution tests by comparison, which makes them less straightforward to view through that lens. consider this my first attempt to square that circle and analyze what makes an arcade rhythm game compelling at multiple levels of skill.


Background

The unexpected grassroots popularity of the Hatsune Miku Vocaloid software by Crypton Future Media quickly translated to licensing deals, and Sega managed to squeeze the first Project Diva game out for Playstation Portable in 2009, just months after Sony Music Japan reissued Supercell's landmark self-titled album and Miku's popularity went fully mainstream. This rudimentary early iteration of the series was helmed by Sega's CS3 studio (previously Overworks, known for Skies of Arcadia and the Shinobi reboot) in collaboration with shovelware producers Dingo Inc. With the release of a second Project Diva title subtitled "2nd" and its subsequent expansion "Extend", sales for the series began to dominate Japanese charts. Sega parted ways with Dingo Inc. after the PSP market declined and let CS3 handle two further entries for the Playstation 3 and Vita solo, which were known as Project Diva F and F 2nd. CS3's era ended with the release of Project Diva X on Playstation 4, which fell prey to a cool fan reception, a recall in Korea due to explicit lyrics in the song Holy Lance Explosion Boy, and the overall decline of the Miku brand. In their prime, these games were not only advertising vehicles for the primary Vocaloid products and characters (as well as shallow "pet" simulators outside of the main rhythm mode) but also an important venue for up-and-coming artists to spread their work. This series served an important tastemaker function; just look at the number of comments that mention Project Diva on the Niconico page for baker's excellent track Sound (they don't seem to be on this version of the site anymore but I found the old comments by using Google Translate across the whole page with a Firefox extension).

Meanwhile, the continuing relevance of arcade rhythm games led Sega to task a post-Yu Suzuki AM2 with developing an arcade translation of the PSP game. A prototype was undertaken using a variant of the older Virtua Fighter 5 engine before eventually debuting as a standalone title roughly a year after the release of the first PSP title. While some elements were retained from its sister series, including using Sony's trademark triangle/square/cross/circle symbols for its buttons and many of the same promotional videos (PVs), the underlying mechanics, scoring, and note chart design differed. Along with many tracks pulled from the PSP titles, Sega solicited fans to submit tracks in a contest to be added as arcade-only songs, with custom PVs also being produced via the PSP game's unique video editing mode.

AM2 continued to update the game through a series of minor revisions each containing a selection of new songs, while two major additional revisions added in a bulk of songs from the mainline series, with Version A covering Project Diva 2nd and Version B covering Project Diva Extend. A few final revisions trickled in songs from AM2's subsequent Nendoroid-inspired 3DS series Project Mirai before a full overhaul of the game was released in late 2013 as Project Diva Arcade Future Tone. This title reflected both a move from Sega's older Ringedge arcade board to the Nu platform as well as the addition of a touch-controlled slide bar. A slew of songs from the first Project Diva F were incorporated, with further small revisions bolstering the roster. However, the influx of arcade-original songs was minor, and virtually all songs from the Future Tone era were ported in from the mainline games and Project Mirai 2. The relative failure of Project Diva X signaled the end of active development on the series as a whole, and Future Tone was sundowned in 2016 with only one song from Project Diva X included. After six years of arcade exclusivity, Future Tone received a Playstation 4 release compiling nearly all of the songs from the history of the arcade game, with various DLC that further expanded the game. Fittingly, the arcade version of Future Tone received a late update in 2017 after a year of inactivity which added in two new songs, one of which was the incredibly popular Miku 10th anniversary song DUNE/Sand Planet. A slew of additional note charts for old songs (placed under the "Extra Extreme" label) were added over the course of 2018, with no new songs in their wake. Project Diva Arcade has been defunct since.


The Cabinet

Project Diva Arcade uses a traditional single-person cabinet size and structure, with a controller width of 930mm (around three feet), a depth of 865 mm, and a height of 2190mm (seven feet and change). Official images of the cabinet may be seen on Sega's official site. The main display sits roughly a little over four feet up and is gently tilted back for viewing from above. Its bottom edge is level with both the smooth, fiberglass-esque slide bar and four large hemispheric buttons. Above the screen sit two small speakers on either side of the assembly, although for those who find them quiet there is thankfully a headphone port as well. Virtually everywhere I've ever played this cabinet has necessitated use of the headphones, even when placed next to an identical cabinet.

The buttons are very similar to those used in Pop'n Music: they are around as large as the palm of one's hand and are gently curved. These use an Sanwa optical switch with a 200g spring. Third-party controllers often use a 100g spring instead for a quicker activation. I find the latter to be more comfortable; I was fortunate enough to experience a modded Diva cabinet at MAGfest one year that featured 100g springs, and I noticed virtually every high-level player immediately gravitate towards this cabinet. For a game that necessitates double-handed rolls on single buttons, the 100g springs feel much clickier and less prone to "squishing" into the switch such that they don't rebound when switching from hand to the other at a fast rate.

Regardless, each option affords an accurate experience for single-handed "jackhammer" sections where the same note is repeated many times. I have comfortably been able to execute consistent eighth notes at ~200 bpm (a little under 7 hits per second, such as seen here in Denparadigm Ex) and short rolls of sixteenth notes at ~140 bpm (a little over 9 hits per second) on both stock and modded cabs. However, these buttons are known to frequently experience failure; many of the cabs I've played on in my life have at least one button that will intermittently stick during rolls. This is less than optimal in a game so laser-focused on sequences of rolls, and it begs the question of whether a different type of button would have been more suitable or if, like the console series starting with Project Diva 2nd, they should have switched to a system that used a separate set of four buttons for each hand. Official arcade-style controllers for the console games indicate this could have been a reality. Alas, as Project Diva Arcade released shortly before 2nd, it is likely that their controller design paths diverged too early in the maturation of the series for this to have ever been under serious consideration.


Basic Mechanics

Project Diva Arcade retains many of the surface-level characteristics of the original Project Diva on PSP. It uses a four button layout featuring the traditional triangle/square/cross/circle symbols of the Playstation controller, although the buttons are laid out in a row in the aforementioned order instead of in a "plus" formation as on a Dualshock. The game does not use a traditional manner of scrolling notes; rather, the notes pop into existence at chart-specific positions as the song progresses. The notes are initially hollow (or rather, opaque black with a white outline) as they appear, but are filled by a corresponding colored opaque note that flies in from the edge of the screen. The moment at which these two notes intersect is the moment at which the correct button or set of buttons should be pressed.

It should be noted that this method of indicating timing has some unique problems associated with it. Most rhythm games use fixed, scrolling playfields that give each note an identical length of time from entry onto screen to the so-called "judgement line," or rather, the line at which intersecting scrolling notes should be met with an associated button press. However, Project Diva Arcade's judgement marker equivalents - the hollow notes - appear at wildly varying points on the screen, and thus the consistency for where notes appear from and where they are moving to is lost. Had there been no forethought, this would also result in different lengths of time for each filled note before it hits the corresponding hollow note. Project Diva Arcade compensates for this by ensuring that notes always enter from either the top or bottom sides of the screen (depending on which the hollow note is further from) and by giving each note a slight curvature to its path that varies depending on the distance between the judgement marker and the edge of the screen.

Of course, this creates a further issue: the note pathing is now non-linear and variant across different notes. It begs the question of why this method of marking the notes on screen was chosen in the first place. One answer may be to add visual flair to the note charts and incorporate them with the PVs, such as in this cheeky rendition of the Eiffel Tower seen in Paris Cinema Girl, and another may be found in the influence of the Osu! Tatakae! Ouendan series, which popularized a similar non-scrolling charting style reliant on notes placed on different parts of the screen. The original Project Diva developers on PSP may have taken inspiration from the Ouendan games for their casual appeal on a handheld console; however, Ouendan explicitly factors its moving notes in as an execution check on the player's ability to move their stylus around in tandem on the Nintendo DS's touch screen. The Project Diva series never ties its similar charting style to the mechanics in any meaningful way, as the note positions on screen have no bearing on play at all. This makes the choice more of an obstruction to clean chart design than anything else.

Nevertheless, both console and arcade Project Divas band-aid over the non-linear note pathing in the same way: by adding a marker with the appearance of a hand from a clock. When a hollow note appears on screen, it starts with a clock hand in the 12 o' clock position. This hand immediately starts rotating around the note clockwise until it reaches the 12 o' clock position again; this is the point at which the player must press the appropriate button(s). This lessens the overall clarity issue by localizing the judgement marking to the note itself, which keeps the player focused on the static hollow notes appearing instead of the filled notes flying in from off-screen. Still, this is an acquired taste, and one that is much more complex than notes scrolling down a playfield. It also does not provide a practical or consistent manner of indicating the temporal distance between two notes due to each clock hand being tied to a different note with no ability to directly compare two hands on the same note (except in rare cases with overlapping notes, such as this post-chorus breakdown in Soiyassa on Extra Extreme). As a result, the point at which Project Diva Arcade approaches unreadable note density is much lower than that of games with traditionally scrolling notes, creating practical limits on what is reasonable in high-level charts. High-level Project Diva Arcade players should expect to spend a significant amount of time studying endgame charts at slow speeds to grok unusual or poorly conveyed rhythms.

Each note press is graded on the accuracy of its timing on the enumerated scale of COOL, GOOD, SAFE, BAD, and WORST. These not only influence the player's score but also affect the player's remaining life. Project Diva Arcade features a life bar at the upper left of the screen that decays on missed or inaccurate notes (BAD and WORST), and when it drains it immediately ends the song (unless the game is played on the "premium" mode, which is one single song all the way through with no limitations). Getting COOLs and GOODs back to back will also create a combo multiplier; getting a SAFE or below will end it. One additional quirk is the WRONG grade, which occurs when the correct timing for a note is achieved with the wrong button pressed. The developers chose to make the life loss for this grade negligible, making it possible to skillfully mash through some sections that are otherwise too tricky. However, because this kills combos, it's not really exploitable, and the dire consequences of ruining the actual timing make it so that the player still needs to have some comprehension of the song's rhythm.

Project Diva Arcade Future Tone added a new input method alongside the normal buttons: the slide bar. This bar is touch-activated with a light array underneath that tracks points of contact. Sliding is represented in the game via orange arrows, which point either left and right and may have a "tail" of any length. This tail signals the length for which the slide must be conducted, with no tail requiring no particular slide distance and a long tail potentially requiring as much or more than the length of the entire bar. However, slides have no timing restrictions beyond initial contact. The tail technically consists of many smaller hollow slanted bars arranged in a line; tiny slide notes fly from the outer edges of the screen into the hollow bars along with the timing of the song much like regular notes. However, depending on the distance the player traverses sliding their fingers across the physical bar at the moment of the initial slide, a corresponding segment of the tail's hollow bars will fill blue. This will ensure that the player does not lose the slide when the filled hollow bars make contact with their associated notes, and thus it avoids the player needing to make physical contact with the slide bar the whole time the slide is in effect. Effectively, the player may buffer in slide inputs at the slide's initiation and spend the remaining time on hand positioning, although no song to my knowledge ever requires the player to babysit a long slide while simultaneously pressing buttons unfortunately. A fascinating side effect of this is that it also allows a long slide to be filled by multiple short slides, as long as slides past the first are input before the song's progression reaches a point where there are no more filled bars in the tail left. A player may choose to not only break up a long slide but to also switch hands in the middle of a slide without the threat of needing to maintain contact with the slide bar while doing so, which may be useful in scenarios where a slide effectively "crosses up" a player (i.e. a player has to perform a long slide from left to right and immediately jump back to the left buttons).

Slides in Project Diva Arcade may also be presented to the player as "chords," where multiple slides must be input simultaneously. While chords exist for the standard buttons as well, slide chords have the added wrinkle of their control surface being contactable at a continuous range of points; the only restriction for chords is that the two points of contact where the slides are initiated must be a nominal distance apart, perhaps a couple inches. There are four types of slide chords: both left, both right, pointing inwards towards each other (or -> <-), and pointing outwards from each other (or <- ->). A common way of dealing with these chords is through using the index finger of each hand simultaneously; the song Requiem of the Phantasm makes this clear in its Extra Extreme chart by tying this physical movement to the conductor-like gesticulating of Miku in its PV at multiple points, including twice in this clip here. However, astute players will note that the two slides don't need to originate from two separate hands; two fingers on one hand work as well if spaced correctly. Soiyassa on Extra Extreme has a good example of where this is useful in its end section, where the player must alternate double-button chords and slides in the first two measures and single buttons and slide chords in the second two measures. The second half gives the player the opportunity to create parity with the first half by continuing to use one hand on the slide bar and one on buttons if they choose to use two fingers on the same hand to slide; I often accomplish this with my index and ring fingers. Likewise, inward and outward slide chords can be accomplished with the thumb and middle fingers moving towards and away from each other, much like adjusting the zoom on a smartphone's screen. The integration of the slide bar into the player's finger and arm movements prevents them from keeping their hands glued to the buttons, which begins to form a picture of where Project Diva Arcade succeeds in creating opportunities for the players to plan physical routes for their hands through its unique controller.


Advanced Mechanics

The original Project Diva on PSP had a major mechanic that did not get directly ported to the arcade edition: held notes. For these notes, a full snake-like note would fly in from the edge of the screen with the shape of the corresponding button on both ends, and the player was required to press down on the button at the start of the snake and then release precisely at the end; both the press and release here were graded as separate notes. Again, visually this summons images of the similar "phrase marker" mechanic from Osu! Tatakae! Ouendan, where the player drags their stylus over a ball that rolls down a snake-like track from one point to another.

Project Diva Arcade replaces this with a twist. The arcade hold mechanic is effectively an overlay on top of the regular note chart that does not directly affect the player's performance in terms of grading or combo breaks. Rather, when a note or set of notes is marked "HOLD," they can be optionally held down past the time where they need to be activated for additional points. The longer the note is held, the more points are earned, and the more notes held at once, the greater the multiplier on said score. A hold will automatically end at the five second mark, at which point an additional "max hold bonus" will be awarded to the player; otherwise the hold will end when one of the notes that is being held must be pressed again for a new note, even on versions of the game where one could theoretically press a note without breaking the hold. Holds can also be added in progressively much like slides, where consecutive held notes may be chained together to raise the hold bonus multiplier without breaking the previous hold. The key here is that while a note or set of notes is held, the player must continue playing the rest of the song. Notes other than the ones being held as well as slides will continue appearing on screen, necessitating that the player proceed with the track as usual while accounting for the held buttons up to and until the max hold.

Project Diva Arcade's take on hold notes is certainly unusual; most games hew closer to the model in the original Project Diva. However, Project Diva Arcade's version brings much-needed complexity to the otherwise drab four button layout. Let's consider the primary "home" position for the hands across the four buttons: virtually all players will play the triangle and square buttons with their left hand and the cross and circle buttons with their right hand. I personally play in a "penguin"-style, where I use my thumbs for the two inner buttons and the rest of my fingers on each hand for the outer buttons. This arrangement allows both symmetric coverage of the buttons and a fair range of movement where virtually everything is in reach with a bit of forward arm movement. All buttons lie directly beneath fingers, with the primary movement challenge being switching between the home position on the buttons and the slide bar.

The hold mechanic effectively removes forward arm movement, forcing the player to rethink the way they approach the control space. Consider a case where one must hold down the cross button (which is second from the right). If they are using their thumb for the hold, then all right arm movements must pivot around the thumb for the duration of the hold. Some aspects of play that this alters are relatively nuanced. If one is performing a trill between the cross and circle buttons (rapidly alternating between the two), they are likely going to quickly rock their wrist back and forth to shift weight between the thumb and the fingers in succession. Here the primary force is exerted by the forearm, with the thumb and fingers simply transmitting the force. If we then consider a situation where one is holding down on cross and then must rapidly play jacks on the circle button (the jackhammer rapid notes mentioned before), then this force exertion will not work, as twisting the wrist will release the cross button inadvertently. Instead, the muscles of the fingers must be directly used to "flap" downward onto the circle button, with care taken not to lose the thumb's grip on the square button during the movement. The same goes in the opposite direction, where one must hold down the circle button with the fingers and proceed to play jacks with solely the thumb. This may sound trivial, so hold your right hand up and attempt to flap the thumb back and forth independently at speed without jittering the rest of your hand. You'll find it to be difficult and potentially uncomfortable to do for more than a few seconds. If you press your fingers against a table you'll find it easier, but note that the buttons have a built-in spring force pressing against your hand that a static table doesn't have. Even the buttons for something like beatmania IIDX or Sound Voltex use significantly lighter springs designed for high accuracy and little input force from the player, making a button's response against the player while being held not much more than that of the table. The specificity of Project Diva Arcade's cabinet design enables its hold mechanic to diversify the way the player physically routes their body parts while playing the game.

The slide bar further complicates matters, as it sits out of range of the hands without movement of the arm. If we continue with the hypothetical held cross button of the previous example, we could theoretically slide our thumb forward over the cross button to reach the slide bar when needed without losing our hold. However, not only does the button's heavy spring render any major movement across the button dangerous due to the potential for an accidental momentary release, but the button has a curved surface which makes the requisite force magnitude and direction for holding the button down slightly different depending on where one's thumb is positioned. If the thumb is positioned near the center, as is intuitive, then one will likely not notice that this is an issue, but sliding towards the back of the button will quickly require the player consciously exert force fully downward even when pressing on a surface that would otherwise be sloped, diverting some of the player's force in a horizontal direction. Therefore, one must fully pivot their elbow around the thumb as a fulcrum in order to reach the slide bar, preferably with a 90 degree angle at the elbow with the fingers pointed towards the left side of the cabinet. The initial swivel can be used as momentum to begin a left-direction slide if the song calls for it, and the fingers can proceed to slide back and forth without much further arm movement. However, this blocks off access to the circle button completely, as it is now under the forearm. One may be able to press the circle button down with the arm, but doing this without the fine motor skills of the fingers is clumsy, and thus even using this technique in the first place may be only applicable to circumstances where the player feels comfortable jettisoning the circle button. Consider further that the circle button is being held down by the fingers and we are now required to attempt to slide by pivoting around said fingers. Trying to even figure out a suitable way to both hold and slide at the same time in that configuration may be more effort than it's worth.

While in many cases you can use a hand that is not holding any notes in order to perform the slide, there are certainly situations where the above becomes highly relevant. Consider this particular ending section from Transparent Watercolors ExEx. At the start of the linked section, the player must slide left, double tap triangle, and hold a triangle/cross chord. Note that this held chord not only requires both hands (unless one has a particularly wide hand span and could hold both with one hand) but also involves one outer button held with fingers (triangle) and one inner button held with the thumb (cross). Immediately after, the player must repeat the inverse of the same pattern, where they double-tap square, slide to the right, and hold down both square and circle, effectively causing a full four-button hold if done properly. The key interest here is the restriction that the held triangle and cross add when the phrase is replicated. The square must now be pressed without the thumb disturbing the held fingers on the left hand, and the fingers of the right hand must slide over the bar and immediately onto the circle button without disturbing the thumb holding down the cross button. I find this creates an incredibly pleasing sensation of holding down the left side of each hand and then flourishing over to hold down the right side of each hand, with a couple button taps in the middle.

Immediately afterwards, the rhythm is repeated yet with the holds changing; the first hold is now performed with a triangle/square chord and the second hold with a cross/circle chord. As the second part of the phrase now starts with a tap on the cross button, the partial sequence must now be performed fully by one hand, with a double tap of the thumb, a rightward slide across the bar with the fingers, and a return to home with both thumb and fingers holding down the chord. These two sequences illustrate situations where the hold may complicate otherwise straightforward patterns by restricting the movement of the player. Others with more comfort with different techniques may find alternate ways to route these physical movements, such as using palms to allow holding a button while freeing up the same fingers to perform slides or choosing specific fingers for each task; I personally use my middle finger to slide in the second variation of the above phrase and then chord the notes using my thumb and pinky, transforming the tap, slide, and hold steps all into a single wrist motion. This is made possible by the richness of routing in the physical domain, where the player can adapt their playstyle and approach to each song's chart according to their own bodily abilities.

The same could be said could be said for many other games with hold mechanics: Pop'n Music, having nine bubble-shaped buttons compared to Project Diva Arcade's measly four, not only has a more traditional hold mechanic but also frankly features much more prominent physical routing at its top level of play, as demonstrated in this video of top player TATSU. Note how elegantly his hands move across the board, swapping responsibilities for each button between hands as the song demands and avoiding getting too comfortable in any home position. Project Diva Arcade never quite reached this level of sophistication when it came to chart design. However, there are two major wrinkles that Project Diva Arcade brings with its approach to holds. The first, more minor one is the way the timing of the hold is completely decoupled from the rhythm of the song itself. Hold notes in most rhythm games directly correspond to phrases and rhythms within the song; the timing of the hold here is purely an overlay onto the regular chart. The player must account for the length of the hold note as a separate internal timer, even as the hold continues into separate phrases beyond the one where it was initiated.

The second, more important reason is the scoring. Project Diva Arcade uses two separate kinds of scoring: the regular score and a Completion Rate (this links to the Project Diva Wiki, shortened URL used due to embedded parentheses breaking on Backloggd). For the purposes of this critique, I will focus on Completion Rate, as it directly ties to song ranks. As with most rhythm games, Project Diva Arcade ranks the player on their performance with a set of separate, fixed grades: Standard, Great, Excellent, and Perfect (although the latter is not score-based and solely records whether the player hit every note in a song). These ranks are given as the player exceeds certain thresholds of earned Completion Rate given as a percentage, with the Extreme and Extra Extreme difficulties putting the thresholds at 70%, 85%, and 95% for Standard, Great, and Excellent. Completion Rate is calculated primarily as a ratio of the player's combo and timing score (without considering holds) to a hypothetical "maximum" score given by playing every note perfectly throughout the song with COOL timing, hence the use of a percentage to represent the score. This means that the primary component of a player's completion score (and therefore the rank they earn on a given song) rests on their basic ability to accurately hit notes.

On paper this sounds reasonable: if a scoring system is a game designer's way of conveying the importance of particular aspects of play to the player, then the scoring system here encourages the player to hit as many notes as possible while also focusing on timing. However, this ignores a key facet of score gained from hitting notes: the combo bonus. Project Diva Arcade applies a flat point bonus of 50 additional points per note multiplied by the tens place of the combo counter. For instance, hitting a note with a combo of between 10 and 19 notes will give a bonus of 50 points, hitting a note with a combo of between 20 and 29 notes will give a bonus of 100 points, and so on up until the combo reaches 50, at which point the combo bonus saturates to a bonus of 250 points per note hit from that point onward until the combo ends. This is not an insignificant amount; notes hit with a COOL timing give 500 points as a base, meaning that a bonus of 250 points will increase the score by a factor of 1.5. Notes hit with FINE timing give 300 points as a base, and the bonus increases them up to a whopping 550 points, which is worth more than a COOL timing with no combo. This scoring system favors consistency through maintaining long combos of correct over timing, the latter of which can be partially mitigated through the combo bonuses.

Looks are deceiving here, however. Consider that I have a song of some relatively long length, such as 500 notes long, which gives plenty of opportunities to build up a 50+ length combo even if it is dropped multiple times. Let's say I drop the combo twice, once around a third of the way through, and once around two thirds of the way through. I have now lost 7250 potential points from the full bonus each time I have dropped my combo (ignoring the points lost for actually missing a single note, which will vary depending on timing), which ends up being a total loss of 14500 points overall. Each combo drop results in the subsequent 50 notes each getting less points than they would have if the combo was not dropped, with the first nine notes being 250 points lower, the next 10 notes being 200 points lower, and so on. On a different attempt, let's say I again drop the combo twice but this time ten notes after the start of the song and 10 notes before the end of the song. I now only lose 2450 points on each combo drop, for a total of 4900 points lost total; a mere third of that compared to the first hypothetical. This is because on the first combo drop I will be losing only 50 points per note for the next 50 notes due to only lagging the original bonus by 10 notes instead of 50, and on the second combo drop I'll lose 250 points per note for only 10 notes as compared to 50. If the maximum score is 367750 (using the formula given in the previous wiki link), then the first hypothetical will result in a 2.6% higher Completion Rate than the second hypothetical even though only two spaced-apart notes were missed in each case. Although not earth-shattering, this is certainly enough of a difference to tip the scale over getting a particular rank, and I on multiple occasions have won or lost ranks thanks to missing single notes at the "right" or "wrong" times. It's difficult to enforce the idea of "consistency" through a scoring system such as this, as it ends up favoring longer combos to the extent that it prioritizes certain notes over others without any thought given to the context of the song. The lifebar does serve as a backup method of ensuring consistency by deducting life on every miss and regaining life on each correct note, but the weights allocated to the lifebar in this particular game make it so that one will rarely die outside of rapid flurries of notes that will instantly kill the player if missed; it has no way to punish long-term inconsistent play because the life regain from correct notes is too powerful. Therefore, the scoring system struggles to emphasize a method of play outside of maintaining large combos, which is applied inconsistently.

However, there is a secondary component to Completion Rate: the hold bonus, which is this second wrinkle for how the game handles its hold mechanic. The ratio of points received to maximum points only considers points gained from hitting notes, both in terms of points gained from accurate timing and the points awarded from longer combos. The hold bonus provides additional Completion Rate percentage points directly on top of the ratio up to a maximum of an additional 5% per song. For example, getting a four-button hold to the full five seconds (which is uncommon but certainly not rare) provides 18000 points, which will give a full additional 1% to one's Completion Rate in the aforementioned hypothetical song according to the formula given in the wiki link above. Under ideal circumstances, one can actually exceed 100% with exceptional performance on a given song (the maximum is 106%, with 100% coming from all COOL timing and a full combo, 5% coming from maxing out on the hold bonus, and an additional 1% coming from a full lifebar bonus per-hit that I've glossed over up to now). As the hold mechanic provides a much-needed complexity factor to a player's routing for each song, centering it as a vital mechanic with the scoring system in this way ensures that players are incentivized to pursue it as often as possible. It also papers over some of the issues with the combo scoring in terms of allowing the player to "make up" lost points from combo breaks with skilled play instead of inconsistently punishing them for their mistakes. For those primarily focused on achieving Excellent ranks on each song, the addition of this percentage cushion predicated on the advanced mechanic provides relief as long as one is willing to learn what makes the game unique compared to its competitors.


Chart Design - Pre-Future Tone

Each song in Project Diva is given between three and five different note charts on different difficulty levels: these are denoted as Easy, Normal, Hard, Extreme, and Extra Extreme, with the former and latter not making an appearance for certain songs. The Extra Extreme difficulty in particular originated in the Future Tone era as new charts with sliding mechanics for old songs that did not have them. Each chart is independently rated on a scale from 1 to 10 stars, including half-star denominations between ratings (i.e. 7.5*, 8.5*, etc.). While ratings for any chart on any difficulty can theoretically fall anywhere in the scale, charts for a given difficulty level tend to fall into a specific range; for example, Hard charts tend to be found in the 5* to 7* range, while Extreme charts are virtually always 8* and above. As a note, I'll be referring to Extreme and Extra Extreme charts as "Ex" and "ExEx" respectively throughout the rest of this analysis.

Many of the basic elements of Project Diva chart construction can be found through analysis of Denparadigm Hard, which, at an 8* rating, sits well above most of the other Hard charts in the game. From the opening verse, one may notice the general structure of the game's charts: each measure (four beats in this case, if you were to count along to the tempo of the song) has its own horizontal line of notes on the screen, which zig-zag back and forth across the screen moving downwards on each measure until reaching the bottom of the screen. Each string of notes on Hard tends to consist of only a single button, although at the blistering speed of Denparadigm it may as well be a jackhammer pattern as discussed before. If we review the layout of the buttons from left to right (triangle, square, cross, circle) we may also note that most of the adjacent phrases tend to use physically adjacent buttons as well, and often times the buttons appear in ascending or descending orders, such as four phrases starting at this timestamp moving from left to right until jumping back to square on the final phrase.

Indeed, I would expect most people watching this video after having read the Advanced Mechanics section to be puzzled at the lack of many of the previously discussed concepts appearing here. Chords are completely absent until nearly two minutes into the song and disappear after the bridge and final pre-chorus, holds only appear at the end of phrases before pauses, and slides seem tossed in haphazardly. Many of the aforementioned mechanics exclusively appear in high-level play, even as in many other arcade rhythm games you would find similar mechanics in much earlier difficulties. Perhaps this was done to assuage issues with the awkward visuals for determining the judgement of a note, or it may have arisen from trepidation about making a rhythm game with a strong appeal for casual fans of the subject material quite as brutal as its contemporaries. Such bland chart design permeated even the highest levels of the console series, with the infamous boss song Intense Voice of Hatsune Miku from Project Diva 2nd challenging its players to 20 straight seconds of sixteenth notes at 220 bpm in the exact same ascending pattern as seen in Denparadigm Hard (although for Project Diva 2nd, this would've been counter-clockwise around the face buttons). Both the console series and Project Diva Arcade illustrate teams with no prior rhythm game experience struggling to articulate novel and challenging chart designs on top of the inelegant base of the original Project Diva, and while it's inarguable that Denparadigm Hard would easily fluster a newcomer with its quick tempo and long strings of notes, it's not a terribly interesting chart either.

With that in mind, it would be more fair for us to analyze a true high-level chart in the arcade version. The Project Diva Arcade version of Intense Voice released two years after its original appearance in Project Diva 2nd, and it thankfully attempted to provide a more considered challenge on Extreme than the lazy button-mashing of its original iteration. As the arcade version's first 10*, we can immediately begin to see some more sophisticated design principles. Relatively close to the beginning of the song we can see rapid triplets: first with cross-square-cross, and second with circle-triangle-circle. This presents a twist on the design of button locality that we first glimpsed in the way that Denparadigm Hard laid out its button phrases.

If we recall the home position for the player's hands specified in the Advanced Mechanics section, we can specify relationships between each of the buttons based on how the player will interact with them in generic cases. There are two major pairings that give insight into how challenges can be inscribed into charts: a pairing based on separating the left and right hands, and a pairing based on the part of the hand used to hit the buttons. The first pairing is straightforward: we divide the cabinet down the middle, with a triangle/square pair tied to the left hand and a cross/circle pair tied to the right hand. The second pairing we see reflected in the above triplets: we have an "outer" pair controlled by the fingers consisting of triangle and circle and an "inner" pair controlled by the thumbs consisting of square and cross. The second pairing is significant because it's symmetrical; in the above triplets, the thumbs are used for the first triplet (cross-square-cross) and the fingers are used for the second triplet (circle-triangle-circle). This ensures that when the player alternates the buttons, they're doing so symmetrically on each hand, which allows the player to activate the same group of muscles on each hand in succession. The alternative, such as switching between circle and square rapidly, would require the player to move separate groups of muscles on each hand at a rapid, consistent rhythm, which would be more difficult. Another subtle aspect of these triplets is that they start with the right hand in both cases, ensuring that the majority of players will lead with their dominant hand. As the first significant execution barrier in the song, these triplets provide a tricky yet digestible challenge for the player.

A few more seconds into the song, we also see an appearance of three-button chords (which I'll refer to as tri-chords). With two-button chords (which I'll refer to as di-chords), there are multiple common combinations that can be done with a single hand, specifically triangle/square, square/cross, and cross/circle. No tri-chord could ever be pressed with a single hand however, making them generally more challenging. As seen in the former clip, these chords are often laid out in a triangle formation, and most often they are chosen based on button locality, with the two most common tri-chords being triangle/square/cross (a left-side tri-chord) and square/cross/circle (a right-side tri-chord). Shortly after the appearance of these we often see the sole quad-chord, where all four of the buttons must be pressed simultaneously. Unlike the tri-chords, where the player must discern which button is missing from a set of three, the quad-chord provides instant clarity into what buttons must be pressed due to there being only one possible combination. It also follows the aforementioned symmetry principle between the hands in an easily identifiable way: both hands must simultaneously press both of their assigned buttons at once. While Project Diva Arcade is nowhere near as chord-centric as many of its contemporaries, the game still deploys these chords frequently in higher-level charts.

In the second main section of the song, we can begin to see ways that chords are utilized in series to create meaningful challenges. Their first application here is in a variant on the staircase configuration, which generally refers to when multiple buttons are pressed in sequence in order from one side of the controller to the other. In this version, one button is repeatedly pressed as a base (first circle, and then triangle) three times in a row, with each press accompanied by each of the other buttons pressed in succession, ascending from one side to the other. This particular idiom lends itself well to moving one hand to the base button while using the other hand to ascend the staircase independently, shifting out of the home position in the process. The final aspect of this is key: by forcing the player to shift out of the home position, the designer immediately has the power to disorient the player by throwing notes at them that would be difficult to do without full coverage of the buttons, which requires the player to budget in time to return to the home position. We examined this concept previously in the Advanced Mechanics section regarding the slide bar, the sole purpose of which is to mandate that the player moves their hands away from their default position; a pre-Future Tone chart such as Intense Voice Ex does not have this luxury unfortunately. However, by moving away from home position, the player may have the ability to reduce physical complexity, such as in this example where a player may avoid fine motor control and asymmetric inputs in the home position (using individual parts of the hands to hit each button while hitting the last note of each staircase as a single-handed chord) by transitioning it into symmetric inputs with gross motor control (playing each side of each chord with a separate hand consistently). In this case, the long pause after the rapid staircase makes it trivial to return to the home position if one leaves it in order to execute the staircase.

I've covered the thought process without considering how the holds affect it, but we can clearly see that a number of these buttons have holds attached to them. In this particular case, the hold mechanic allows the player to optionally pursue the more complex home position route by requiring that they not only press each button in the ascension part of the staircase but hold them as well. This creates further asymmetry between the base button fingers and the remainder of the actionable bodily appendages by giving them different roles (one repeatedly presses, the others press and hold). Although a subtle example, we can see here how the hold mechanic is able to provide the incentive for more difficult physical routes through a sequence, giving players of different physical ability levels separate options for dealing with the section. We can see a Future Tone-era spin on this concept in Transparent Watercolor ExEx, where this particular staircase pattern is followed up immediately by a slide, with a return to home complicated by the immediate follow-up of the opposite staircase pattern with much smaller breaks.

Returning to this section in Intense Voice Ex, we can now observe a different kind of staircase, where the three separate adjacent di-chords are used in succession. If we look back at the common di-chord list, we can see that the first (cross/circle) and last (triangle/square) in the series are di-chords meant for a single hand in the home position, while the middle one is the "thumbs" di-chord (square/cross). Again, the player has the option to either let one of their hands leave the home position to hit the middle di-chord or go for the more complex hold route, where they quickly perform a single-hand chord, a chord with both thumbs, and the opposing single-hand chord in rapid succession. After two of these staircases (first starting at the right side and then the left) and a similar pattern where the single-hand chords occur first before the "thumbs" chord, we see the escalation of this challenge, where the two single-hand chords occur before the "fingers" chord (triangle/circle). In this instance, the less complex route of navigating away from the home position is removed due to the distance between triangle and circle; while square and cross could be pressed simultaneously with a single hand, the same does not exist for triangle and circle as they each sit at the outer and opposite edges of the cabinet. Therefore, while one may progress up to this point without the necessary fine motor skills to transition from full-hand movements to finger-specific movements, they must demonstrate them in this final phrase of the section in order to progress without a life penalty.

I've chosen these examples because they're relatively comprehendable when watching and provide a good entry-level understanding of the subtleties that come from these kinds of constructs when encountered in high-level play. This begs the question of why this kind of design exists in a top-level chart, which should theoretically have much more difficult design. The section after the above is mainly rapid-fire single-button presses, although the designers still are placing jackhammer sections on one button before trills between two buttons, indicating that they still have not realized that the former is harder than the latter due to the stiff button springs and precise timing needed to let the button rebound between successive presses. The real challenge comes within the last six measures, where extremely fast triplets give way to rapid single-button staircases (looping each button from triangle to circle and then later circle to triangle) and shifting trills between the two hands. This is an unfortunate example of a song infamous for ending with an sudden volley of notes, which makes the song aggravating to practice; every failed attempt at the very end of the song results in needing to replay over two minutes of comparatively easy gameplay to get back to the hardest section. While the original PSP version of this song had a similarly brutal outro, it was at least shrouded in a Chance Time mechanic that game had that removed the lifebar while giving the player additional points during the segment. It's also worth noting that this kind of frustration has been addressed in later songs, such as Leia ExEx putting its most technical section at the beginning of the song, making one's poor performance evident from the beginning of the song rather than the end (it's also worth comparing this song, a 9.5* in the Future Tone era, to the 10* Intense Voice Ex in terms of its more complex chord patterns in the chorus).Given that the pattern is not particularly complex at slower tempos and solely subsists on the speed to trip up the player, the song overall doesn't necessarily do an excellent job illustrating the higher levels of play.

Unfortunately, this may have been one of the more interesting endgame charts of its era. The next 10*, Two-Faced Lovers Ex (my apologies for the scantily clad Miku outfit that top-level player hisokee is using in this video), demonstrates a common kind of chart from this era: charts heavily based on the syncopation and rapid syllables of Miku's vocals. For those less familiar with music lingo, syncopation essentially refers to rhythms that prominently feature beats that fall "off-beat"; if you imagine tapping your foot or nodding your head along to a beat, syncopated rhythms will try to exclusively use beats in-between the "on-beats" made by your tapping. As such, most of this song's challenge is centered around long lines of single notes tightly woven together with few chords to speak of. This style of charting and the jackhammer patterns within present their own conflict for the player with regards to moving away from the home position: what segments should they attempt to mash with a single hand while keeping all buttons covered, and what segments should they attempt with two hands in order to ease the strain on a single hand over longer sets of jacks? However, in a chart with few chords to encourage button coverage or holds to encourage single-handed play, the easiest strategy is simply going for two hands as often as possible. Because the decision between remaining at home base or roaming the controller is given so many times, it dilutes its potential meaning and ends up making the chart feel long and samey. By the end of the song, the notes are so densely packed together that, with no adjustable scrolling speed or appearance rate for the judgement markers, they are difficult to even make out.

Two-Faced Lovers is a bit of an outlier example in terms of the severity of its devotion to jacks. However, most of the top charts from that era played with similar design paradigms. How'd It Get To Be Like This? Ex features less prominent syncopation and no jackhammers in favor of playing more with odd note localities and staircases, such as in this inspired sequence that both places staircases of different lengths next to one another in the same phrase as well as creating sequential staircases that actually "loop" from triangle on the left side to circle on the right side seamlessly. Po Pi Po Ex hews closer to Two-Faced Lovers regarding syncopation, with its primary chorus rhythm being iterated upon in adjusted staircases with mini-trills intertwined. Of this era, Paradichlorobenzene Ex may be the best expression of this style, with a rolling samba-esque rhythm underpinning some very creative extrapolations of the staircase featuring note triads that advance left and right across the controller. This chart in particular does an excellent job weaving chords into this playstyle, and in the second half of this clip it highlights pairings of notes that we glossed over earlier: button pairs with a single other button between them. This pairing, consisting of triangle/cross and square/circle, both exist across the hands in the home position and feature asymmetric pairings of hand muscles, with each one requiring use of the thumb of one hand and the fingers of the other. This pairing is essential for ratcheting up difficulty after particular rhythms are initially introduced with the two types of pairings mentioned prior, as they demand an additional layer of asymmetric coordination that the other pairing types do not require.

One design aspect each of these charts struggles with is their integration of chords, exemplified best by Saihate Ex. The first half of this chart is innocuous enough, replicating many of the same patterns we discussed previously regarding Intense Voice ExEx (another indicator that the latter chart punches well below its weight in terms of difficulty in its first few sections). The sheer volume of chords in some of these sections strains the player's ability to read them accurately on a first pass, exacerbated by their close locality, minute variations between chords, and lack of distinctions in color between the notes. If we slow this section down, it doesn't look complex on paper, with alternating inner and outer di-chords switching to alternating right and left tri-chords before doing a sort of muddled staircase from right to left using a mix of di-chords and tri-chords. One may feel compelled to leave the home position for the tri-chords, where they might use their right hand to hit the right-most buttons and the left hand to cover the single remaining button, which can then be simply moved one button to the side to hit the alternate tri-chord. Leaving the home position has next-to-no penalty for the player, as the alternating tri-chords end on the same chord they started on and can be easily transitioned out of into the final two di-chords, leaving the player back in the home position regardless. Given the poor visibility, it's more of a knowledge check than anything else; something a player would check in a video before playing the song in order to avoid being overwhelmed by the visual flurry of notes. One way to solve this would be to transpose the notes on screen to match the physical layout in front of the player, mimicking a vertically-scrolling rhythm game in the process. The chart does use this method at multiple points, begging the question of why they didn't choose to deploy this more frequently.

This is not to impugn all of Saihate Ex necessarily, as the song does introduce us to some other chord patterns that will be deployed again later, such as the additive staircase used in the third phrase of this clip that requires the player to actively add each button in successively larger chords before eventually making a quad-chord and subtracting all of the buttons in succession. However, what one may notice from watching the song in full is how segregated the single-button sections and the chords are in general. At this stage in the game's development, the designers seem to conceive of each type of note as a completely separate challenge, with the highest point of escalation being divided between fast jackhammer phrases such as in Two-Faced Lovers Ex and dense chord phrases such as in Saihate Ex. Intense Voice Ex may be one of the better examples of a song attempting to integrate the two, as it repeatedly uses these plinked notes between single buttons on one hand and chords on the other, and the era's final endgame chart World's End Dancehall Ex shows the designers finally giving blended sections a chance in its outro. The rest of the charts we've brought up this far (with the arguable exception of Paradichlorobenzene) fall into this trap pretty noticeably though: examine these examples from How'd It Get To Be Like This? Ex and Po Pi Po Ex.

This kind of chart design artificially limits the kinds of phrases that the player must learn, and it ultimately reduces the physical routing tradeoffs required in the learning process. Too many jackhammer sections on their own don't incentivize much other than two-handing each section, and chord spam can easily trap the player in the home position without confounding factors forcing one to move away. Holds, which should theoretically solve both of these issues, are underutilized as well; other than the smart example found in Intense Voice Ex, I found virtually no instances of holds that weren't at the end of phrases before a pause or spread throughout a section of single-button presses. There are glimpses of excellence here, but much of the older Project Diva Arcade material feels undercooked. Thankfully, the changes undertaken in the transition to Future Tone would remedy many of these issues.


Chart Design - Post-Future Tone

Instantly with the addition of the slide bar we can breathe a sight of relief; the game has finally added a mechanic that forces one to leave the home position as a rule, allowing the kind of physical routing I spoke of in the Advanced Mechanics section to finally blossom. There is a reason many other rhythm games that uses buttons has some sort of non-button control surface to add this kind of complication, from the turntable in IIDX to the knobs in Sound Voltex. Pop'n Music, which is closer to Project Diva Arcade in its button dimensions, solved this issue prior simply by having over twice as many buttons, far more than any person could ever cover simultaneously.

To succinctly demonstrate the drastic rise in the charting chops of the designers at this point in the game's development, let's examine this clip from the song Knife Ex, our first 8.5* we've examined. I've chosen this clip in particular because it does not use the slide bar; it simply shows how much more nuanced the original version could have been with better charting. The clip immediately begins with a hold on triangle that segues into a relatively slow trill on square and cross (our inside "thumb" buttons), where each trill phrase ends with di-chord on the right side of the controller (cross/circle). It's followed by an interlude that ends with a hold on circle, and the pattern repeats in an inverse fashion, with each trill phrase ending on the left side of the controller (triangle/square). Without the hold, this pattern is simple: alternate the thumbs for the trill and end each phrase with the full hand chord of whichever thumb you would have otherwise finished on. With the hold, suddenly our issues with accidentally releasing a held button on fingers while using the thumb to consistently play notes arise, as discussed in the Advanced Mechanics section.

Without the chord, one could move their non-held hand to the inside buttons and trill on a single hand. With all of these together, the player must now choose: do they consistently move between home position and the inside buttons in order to avoid using the thumb on the held hand, or do they attempt to do each section in the home position while carefully avoiding releasing the held button? This still finishes in the home position regardless, a complaint I levied in the previous Chart Design section, but note that this is an 8.5*; theoretically "easier" than the other charts we've covered up to now. It also has the bonus of being completely unique in its execution, compared to many of the prior examples, which used patterns repeated from song to song. Note as well how the held circle at the end hits max time not at the end of the phrase but a few notes into the next one right before a square/cross hold, perfectly segueing holds together! The sophisticated charting here in an 18-second clip perfectly demonstrates how one might have leveraged the mechanics of Project Diva Arcade before the slide bar was added.

A chart with more parity with the heavily syncopated examples from earlier is Gaikotsu Gakudan to Riria Ex, which originally debuted in the Project Mirai series. From the jump it becomes apparent that the old note layout style - long horizontal rows of notes moving from one side of the screen to the other - has been made less rigid, with more aesthetic flourishes and verticality featured by comparison. This layout style not only gives each song more individual flavor (such as in the Paris Cinema Girl example given earlier) but also can assist readability when deployed correctly. For example, the ending section uses it almost immediately to make a trill between circle and square more clear by staggering the two with a vertical offset. This eliminates one of the issues brought up in the previous section: the inability to parse tightly clustered groups of notes. This is made additionally important in the next phrase, which bucks the tradition established by many of the clips up to now by not fully mirroring the phrase before it. Instead of trilling on the same notes used in the chord as in the first phrase, cross and circle (the latter of which had not been used in the phrase before) are trilled, and by offsetting the cross notes in this phrase to be level with the cross of the chord before it, this twist immediately pops out of the screen differently than it would have it every note in the sequence was level.

The next phrase then starts at the bottom of the screen and works its way up to two inner di-chords, which are placed such that they line up with the controller buttons. The slides here additionally sandwich the di-chords instead of appearing on the same plane, indicating to the player that they should use their fingers to swipe the slide bar while holding with their thumbs. We previously saw this used to good effect in Saihate Ex, and by adding a non-horizontal lead-in, the effect is improved. Another small example of where this kind of charting can be an improvement is seen in these two clips from Step Forward Ex, which actually displays a staircase pattern as a literal staircase. Much like a section of How'd It Get To Be Like This? Ex that repeated a staircase multiple times with various notes cut off and ordered differently, Step Forward Ex repeats the same trick, but the visual layout of the staircase now immediately indicates where notes are truncated or where the staircase is reversed. While the original layouts undoubtedly stuck to simple patterns in order to keep the unorthodox judgement mechanics from becoming overwhelming, the designers eventually figured out how to make the quirks of their charting system work for them to enhance readability.

However, this kind of charting can also work against the player; in some cases the Future Tone-era charts actually employ obfuscation of the chart as a challenge. One subtle example seen in Gaikotsu Gakudan to Riria Ex actually relates to holds, where a chance for a seemingly juicy four-way hold immediately gets preempted by other holds. In this case, a left-side di-chord hold (triangle/square) has a circle hold immediately following it, signalling that one should hold all three simultaneously. However, the circle is not followed up with by a cross hold but rather another left-side di-chord hold before finally yielding to a cross hold; the player will optimally hold the di-chord, drop it right before holding the circle, and then re-hold the di-chord and then the cross in order to make a full four-way hold. I'll certainly grant that the conceit is clever in the way it upends the player's expectations on how the game sets up sequential holds, but at the same time this is a pure knowledge check, with as little depth as the long chord phrases that I criticized in the previous section.

A type of obfuscation more pertinent to visual layouts can be found in Jugemu Sequencer ExEx, effectively an entire song built around disorienting the player when attempting to sight read. This includes multiple different box-style layouts, a section that purposefully displays the notes out of order, notes surrounded by a chord that must be hit on each downbeat repeatedly, and trills where the alternating notes diverge away from each other. Clearly the chart designers had fun with this song as they tried to push the note layout system to its limit. Again, this song falls prey to being primarily a long series of knowledge checks, especially since, as a 10*, the actual note-to-note difficulty is probably notably lower than most of its cohort even when compared to the pre-Future Tone charts. Since it's only a single chart built entirely like this, I can let it slide, and since it's not a particularly brutal challenge comparatively, I actually enjoy it a fair bit having sat down and learned the whole thing.

Still, there are instances where these practices accidentally slip into more serious songs. One instance can be seen Denparadigm ExEx, where two right-hand tri-chords overlap with notes in-between the two chords placed between them. Evidently this is meant to both follow the standard of having tri-chords be an equilateral triangle while also having it appear as though they are actually separated from one another and squished to the top and bottom; an admirable attempt, but the overlap between the lines in each chord gives the impression that they're adjacent. Likewise, in Negaposi*Continues Ex, this particular chord sequence attempts to adhere to the faux-vertical scrolling template but jumps up to the top of the screen right as it throws a curveball triangle/cross di-chord instead of the adjacent di-chords preceding it. In this situation, it may have been better to squash the whole phrase vertically and let the curveball chord sit at the bottom so it's clearly visible as a twist by comparison.

Which is a shame, because otherwise this latter sequence is an extremely interesting pattern that far exceeds the examples we covered in the previous section. It starts off with a right-hand di-chord followed by two circle hits in what begins a triplet pattern in an odd time signature. However, as soon as the identical second di-chord is hit, the pattern shifts from double-tapping the right-most note (circle) and instead requires a double tap on the left-most note (cross). The game then turns this into an augmented staircase where the di-chord and its subsequent left-button double tap advance leftwards twice until the triangle is double-tapped, at which point it then throws the curveball triangle/cross chord to disrupt the staircase after only three iterations of the pattern. That curveball is meant to signal that the player must return to the original position (cross/circle) and pattern (right-button double tap as opposed to left), after which they then get thrown into a motif from earlier in the chart utilizing a rapid di-chord staircase that returns to center. This phrase not only completely defies the phrase symmetry that older charts relied on, but it also still somehow returns to its starting point as if it was symmetrical.

Diving into the nitty-gritty of the rhythm in this clip, the song takes the unusual tact of switching into a 3/4 rhythm (three beats in a measure) that is repeated five times. What's more, the actual beginning beat of the repeated pattern the player must hit falls just before the actual start of the 3/4 rhythm, and the accent continually hits preemptively on triplets over top of an already-3/4 rhythm, creating a desync between the actual beat (which is already an odd time signature) and the rhythm the player inputs. The two slides in between the end of the phrase and the start of the motif are essential because they effectively re-synchronize the player to the beat of the song before the motif, letting both the player and the song meet on the downbeat of the fourth measure of the overall clip. The marriage here between the song's whirlwind, atypical rhythm and the complex reinterpretation of a staircase pattern in its image is so crystalline and beautiful that it may be the most perfect five seconds of gameplay this cabinet has to offer; a point at which the game not only completely shears itself from traditional rhythm, but invites the player to partake in the spectacle. Attempting to explain the many ways this may be routed based on player preference could easily eat up an entire other paragraph. Truly phenomenal, even without holds or many slides to add complexity.

Indeed, some of Project Diva Arcade's best moments after the release of Future Tone come from these moments where the game encourages the player's two hands to cooperate past the point of straight lines of chords and into something polyrhythmic. In my interpretation of the above phrase, I completely exit the home position and treat each hand as a separate half of each chord, passing the baton of jackhammer notes from my right to my left and back to my right in the process; the two hands constantly swap different rhythms between one another. Disappearance of Hatsune Miku ExEx features a similar type of rhythm in its notorious ending, where the player plays jacks on one end of the button layout while tossing in other buttons every three notes. The escalation here is interesting, as the first phrase only requires the player to make chords using square and triangle against circle, thus allowing them to easily stay in the home position, while the second phrase with jacks on triangle weaves in a left-hand dichord (triangle/square) halfway through, suggesting that the player may need to move their right hand out of the home position. This rhythm doesn't quite tickle the itch of Negaposi*Continues due to it cinching up each measure in a trill that resyncs the player to standard rhythm on a consistent basis rather than at an odd time as in Negaposi*Continues, but it does innovate with the interwoven nature of chords and single presses more than what was seen in endgame charts in the early days of Project Diva Arcade.

Some of the game's versions of this concept instead fully suggest the home position, mandating fine motor control in the process, such as in this sequence from Denparadigm Ex. The phrase again uses triplets in a similar di-chord double-tap order, but here all the di-chords are outer di-chords (triangle/circle), the special combination that we saw restricts the ability to move from the home position in Intense Voice Ex. The double taps are a staircase from left to right where the player must continually tap the outer di-chord at each outset of each set of three while moving between the different individual notes in the process, while the second half of the phrase switches to alternating squares and crosses in place of the double taps. This hews closer to being a strict execution check; while I see the logic in making a section like this more free-form, I also personally enjoy the game foisting an unusual rhythm and pattern like this upon the player. If the rhythm wasn't triplets, I might be less impressed.

Perhaps the chart most exemplary of the strengths of the post-Future Tone-era design is Piano × Forte × Scandal Ex. This chart doesn't have a particularly awe-inspiring section a la our previous three examples (perhaps due to it being a 9.5* and not a 10*), but instead we see a consistently varied charting style that effortlessly weaves in many of the best elements of the game. The song itself serves as particularly fertile soil: it's not only syncopated but also much looser and airier than the older compositions, owing to composer OSTER project's talent for freewheeling big-band arrangements, and it features an unusual structure with multiple instrument solos interspersed between the traditional verses and choruses.

Even in the first verse we can already see some of the changes highlighted throughout this section, such as the specific placement of holds such that the logical end of each one transitions smoothly into a separate hold instead of large gaps between them. Of particular note is the way this chart deploys obfuscated hold patterns as shown here, here, and here. In each of these instances, continuing one hold with another will cause it to be cut short by an upcoming occurrence of the button used in the first hold, suggesting that the player should cut the first hold short right before beginning the second. On their own, I previously complained about these being knowledge checks; however, executed several times through a song, and when considering the limit of 5% Completion Rate that one can accrue through holds, I see this iteration of the concept as giving players multiple different hold routes throughout the song to consider based on their own personal competence, where even missing the most "optimal" hold in one instance could theoretically still result in the best score as long as enough of the other holds are hit. In the second of these instances, correctly switching holds to avoid ending the bonus also results needing to slide outward with the fingers while the thumbs hold in the center as previously seen in Gaikotsu Gakudan to Riria Ex, and this example switches back to the buttons from the slide bar without breaking the hold as well.

Beyond the acumen in its hold implementation, Piano × Forte × Scandal Ex has excellent fundamentals in terms of its use of the core buttons. One of the tricks the chart pulls repeatedly are slides before a tri-chord, putting the player away to the slide bar fractions of a second before they need to return both hands to the home position. However, shortly before this clip, the designers actually sneak in the same trick using jacks to pull the player's hands from the home position. The jacks could theoretically be done one-handed, but their length and transition between multiple buttons make them much more reasonable to do with two hands, which will require a quick move back to home at the end of the jacks on square in order to hit the tri-chord. Trills are also woven in with more care than we've previously seen, such in this short phrase right before the ending. Note that the first trill actually starts on cross, which is the inner button on the right hand, and it transitions seamlessly into a second trill that also starts with the inner yet is on the left hand. For those composed enough to do so, they could theoretically stay in the home position the whole time. However, this also presents the opportunity to move both hands to the right side of the controller and start the trill with the left hand, transitioning over to the left side in between trills. I personally do something a bit stranger: I cross my hands over one another to begin the trill on the right hand, with it on the cross button and my left hand on circle. I then execute the second trill the same as in the prior explanation, having performed a double tap from cross to square with my right hand while my left hand transitions over to the left side of the controller. While this sacrifices the smooth alternating pattern by adding a double tap in the middle, I find that it gives my hands an easier time moving over, as the right hand executes two hits on adjacent buttons while the left hand has those two presses to move over as compared to only a single one. To provide this abundance of physical routing from a single button-only section is rather genius.

There's another interesting section from earlier in the song that I find equally fascinating. This is a similar trill between cross and square where the cross hits suddenly drop out, leaving the square to wander along the off-beat until the triangle gets subbed in. Again, because the square must constantly be played here, I find myself starting in the home position with right hand on cross and left hand on square until crossing my right hand over my left in order to weave in those triangle notes. The way the chart designers extrapolated the lazy air of that piano line nearly trailing off as it edges higher until finally returning to the trill really sings the praises of what this system can do in the right hands. When Future Tone's charts are at their best, they paper over many of the issues brought up in previous sections, fulfilling the promise of a true arcade-quality rhythm game for the Vocaloid technology.


Analysis of High-Level Play

In the clips I've used up to now, you may have noticed that many of them feature a Dualshock controller in place of the arcade controller I have been describing. Since the release of Project Diva Future Tone on Playstation 4, the majority of the community has transitioned to widely available home versions of the game, which include Future Tone on Playstation and MegaMix/Mega39's across both Switch and PC. This also includes most of the top-level players using regular controllers, which would require a completely different line of analysis than the one used above given the considerable differences in techniques between using an arcade controller and a traditional gamepad. This version of the game is completely fine, and the way the control gets reinterpreted actually has many fascinating quirks of its own, if also sacrificing some of the mechanics in the process. However, it would be preferable if we could take a look at how an actual high-level player takes on the arcade version that we've been discussing. To do this, I'll be using this lovely full combo video of Sadistic.Music∞Factory Ex from top-level player F.SinA.

Sadistic.Music∞Factory Ex is overall a jackhammer- and trill-heavy chart, though it's occasionally sparse due to the designers gunning for 666 total notes in the song, which falls below most of its 10* brethren. Still, from the get-go watching this video we can immediately see why it makes an interesting watch: the amount of jacks necessitate quite a bit of two-handed playing on single buttons, which in turn means we'll see him leaving the home position quite a bit. There actually is a particular physical technique that can get around this: the "tickle," which you may have noticed F.SinA doing while screwing around in one of the breaks. We won't see him deploy it much, likely for its tendency for losing the even spacing needed to maintain good timing on jacks, but we get to see a good shot of it nonetheless. If you come from a game like Guitar Hero, you may be surprised that he doesn't drop a combo or lose health by doing this; generally the rules on when you can and can't hit notes are much more relaxed in arcade rhythm games, allowing him to freestyle in this otherwise empty section.

Getting into the first verse, we begin the fundamentals discussed above expressed. Holds are common but generally feature only one button at a time, yet we still get to see how it affects the way he navigates across the controller. He seems to not feel particularly strong about one-handed jacks with just his thumb, so you'll see him do short sets of jacks with his fingers even when on an inner button, and when he's performing holds on one of the outer buttons, he'll bring his alternate hand over to assist his thumb on inner jacks. Also pay attention to just how far he brings his hands up between each hit of the jacks, as doing otherwise would likely end up squishing the button in place. The most creative physical routing he incorporates here can be seen in this clip: during a section where he needs to hold triangle through three successive sets of inner jacks (square, cross, square), he initially uses his fingers for the hold before smoothly sliding his arm over the button to use his forearm for the hold such that he can use both hands on the cross button on the other side of the board. Such a clever incorporation of a non-hand body part, and one that requires some serious finesse in order to not lose the hold when sliding over the bump where the palm comes up to meet the wrist.

After one of the song's more iconic sections and another verse of jacks, we finally get to see F.SinA play some chords during this clip. A few important key points stand out to me about his play here. First, you'll notice in the opening di-chord staircase he favors the right hand to play both the right-side di-chord and the center one; as a fellow right-hander, I tend to do this as well. This is followed by a bit of a break section with some slides during which he holds cross and circle simultaneously throughout. Be careful to note the incredibly tight timing with which he releases cross and then represses it before another slide, ensuring that he gets the max hold bonus in the process. Surprisingly, he then chooses to play the next section by one-handing each di-chord and then two-handing each trill afterwards, again preferring his right hand. We then see some quick alternations between triangle/cross and square/circle di-chords followed by the usual tri-chords, where F.SinA chooses to leave the home position for the first set and then execute the same strategy I mentioned earlier for Saihate Ex for the tri-chords. Indeed, I learned that strategy from watching this video; I used to alternate my forearms across three buttons at once to hit these alternating tri-chords! Not a good strategy if you're hoping to avoid bruising unfortunately.

The song enters a lull for a bit before launching into its next-most iconic section: a 12-second sequence of constant triplets. The fluidity of F.SinA's movement is impressive here, as he successfully gets all of the available holds in this section without dropping a single note. His initial route through the jacks section is interesting: he starts with his left hand on triangle as one would expect, before switching to right hand for square and then cross before coming back to his left hand for a repeat on the square. My theories for this are either that it was a snap reaction, or that he does it on purpose to favor his right hand, as we've seen that he prefers it in ambiguous circumstances. During the next section with triplet trills and a hold on circle, he alternates between his right thumb and left fingers for trills between cross/triangle and later triangle/cross, but surprisingly when the trill switches to triangle/square he rapidly moves his left hand between the two. It begs the question of whether he finds his left hand (or either hand) to struggle with trills between thumb and fingers on the same hand, as we've seen repeatedly before that he generally elects not to go that route when possible. The rest of the sequence plays out roughly as we would expect.

The following sequence is somewhat obfuscated but ultimately not much different than we've seen earlier in the song, with more presence of di-chords but otherwise a lot of trills and jacks. When we reach the ending sequence, two particular choices jump out at me. At the beginning of the above clip, F.SinA initiates a hold on circle before playing jacks on cross, a square/cross di-chord, and then jacks on square. Those jacks on square in particular we seem him for the first time attempt one-handed, and with some careful attention we can see he actually begins trailing off after the first three jacks in the sequence, causing him to only get FINE ratings on the last two. It's an extraordinary amount of swag for what I imagine could be little more a couple hundred points; what he gains from holding circle a little less than a second would be just under 600 points, and the two FINEs lose him 400 points from COOLs. I respect his dedication to squeezing as many points out as he possibly could. The second fascinating section is the literal ending, which quickly alternates circle with the cross/square di-chord and then triangle with the same di-chord. Rather than stick to the home position, F.SinA goes to cross his arms over each other, using the right hand on both circle and triangle in order to preserve the alternation of the hands.

From this commentary I've provided on his play, hopefully it is evident how much of each person's playstyle and physical routing depends on their own personal preferences and abilities. F.SinA far outstrips me or most other players in skill, and part of his talent comes not only from emphasizing his strengths but also identifying his weaknesses. He comes up with routes that focus on his strong ability for two-handing jacks and trills and avoids instances of one-handed versions of the same, to the extent that he even incorporates a hold using his forearm in order to keep both hands in play at once. While I wouldn't put Sadistic.Music∞Factory Ex in my list of favorite songs, the way F.SinA approaches provides many examples of the kind of play that Project Diva can excel at when charted correctly: highly demanding charts that allow for physical experimentation and routing on behalf of the player.


Final Thoughts

When I first encountered Project Diva Arcade two years ago at MAGfest, it reinvigorated for me what had become a stale game on gamepad and has kept me coming back off and on since. I would be the first to admit that it has serious flaws in its basic design, as I have extensively covered above in the mechanics sections. These aspects have left me reluctant to invest serious time past what I already have into the game, especially when it comes to the often confusing judgement method that prevents me from playing on reaction as much as I would with a vertically scrolling game. This is also confounded by the fact that there aren't many charts left for me to learn, and the few that still are present tend to lean heaviest into the need for extensive prior study before tackling them appropriately. In a much more mature game like IIDX, I persistently have new charts to tackle, giving me experience without resorting to repeatedly attempting charts. Meanwhile, Project Diva Arcade's endgame didn't extensively develop until the last few years of the game, leaving the top-level chart selection a bit barren.

At the same time, when I get to play one of these many charts listed above, ones that I've spent a considerable time working my way through the intricacies of, I still find myself enraptured by simply how physically engaging the game is. Even though I found myself light on interesting examples of charts using the slide bar (and I left out goofier slide charts such as Po Pi Po ExEx), the sensation of running my fingers across it and seamlessly returning to the buttons thrills me. The hold mechanic is even more genius, adding a significant amount of complexity in planning to a game that may have otherwise felt overly straightforward. Its these elements that I find are totally unmatched when not on the original arcade controller, and if you find one of these cabinets in the wild, it would be worth your time to try it just to get a glimpse of what these mechanics feel like in practice.

I would still wholeheartedly recommend the console ports of this game to anyone, especially those who want an entry point into the genre. However, if you're a rhythm game aficionado and have otherwise passed this one up, or you find you have a cabinet for this local to you, I still believe it's well worth your time to learn this particular version of the game. Although the skill ceiling falls comparatively lower than most games in the genre, it'll still provide many dozens of hours worth of playtime to explore, especially if you find yourself enjoying the music.

This is the only phone game I’ve touched that made me go back to do every achievement, feeling ever more things clicking while shooting for the 4-jump-orb only clear, and enjoying myself enough to live through a 9 hour flight with zero wifi. It really does strike the wonderful tonality of an arcade game where you keep finding a thing to master with each stage and learning fruit arc priorities and what combination of the like 3 perks you can have for the 4-clear. Elated when finally got that last medal fucker in the last half hour of the flight home.
It’s just a really good game!! Need more phone games like this :(

Played on default settings with Survivor.

Rain World's.. well, world, is utterly beautiful and captivating, there's something about it that strikes me as endlessly interesting with just how diverse each region is despite what could broadly be described as "post-apocalyptic industrial"; it effortlessly outclasses many in this aesthetic genre, with a sprinkling of Buddhist theming that elevates it to masterclass status in my book. It feels like a mashup of Girls Last Tour and Made in Abyss visually. I feel like this is where the intrigue for many comes in, especially in the last couple of years or so with the cutesy Slugcat fanart and gifs people love sharing around (I did so for ages before even playing), but I feel something is miscommunicated about the game from this angle by the fandom.

When people talk about a game having disregard for a player in a good way, the vast majority of the time they're talking about Demon's Souls or sometimes Dark Souls 1, or cheap shots in mostly old PC games; but these games always give the player the tools to overcome, especially the lauded Souls franchise and adjacent. Rain World does not, or at least not really; you might be given a spear, but the amount of hostile enemies you can actually kill with it? Like, two, at least somewhat reasonably. The world does not care if you wish to walk across that bridge, there is an eldrich horror larger than you could have imagined casually feasting on one of the beasts that previously demolished you. Wait... There's your chance to get past, RUN! Jump over that rock, hurry! Scurry along you silly little Slugcat and pray the monster is still engorging itself. Finally you reach the pipe, and on the other side, another beast, instantly snaps you up.

Press [start] to continue.

So you've learned something at least, about how this ecosystem interacts; but you died, despite how hard it may have been to get where you were. You can't go back though, you know it's a dead end, the worm nudges in that same direction still. You have to push on, Survivor. So you push forward again, with that depleting Karma meter you haven't payed much attention to only a couple hours in or so. You claw and climb your way through a (to you) arduous section with some beasts and perilous jumps, until you finally reach the door and shimmy through, to be met with a Karma Gate at tier 3, but alas you are only tier 1, and hungry. Time's almost up, the screen begins shaking and the sound becomes thunder, soon you're washed away into oblivion. How utterly frustrating, I thought, how could they make the player go through all of that knowing they'd likely die and be unable to get through? There's a lot of moments like this, and an area in particular that drove me insane, I constantly asked "how would I change this?", and each time I'd conclude, simply, that I wouldn't. To fix that or place that on a cycle or something rigid would defeat the purpose or at least severely crack the game's vision. To approach these challenges from the perspective of "how would I change this for the player experience" is to directly detract from the experience of the Slugcats.

This isn't about you, this is about a Slugcat which in this game no matter how good the player may be might as well be regarded as only a couple rungs above insects. The act of killing something that could also kill you is rare, and reliant on often single-use tools you happen across. You WILL be fumbling with Slugcat's movement to the very end and it's all the better for it. Why should things be easy? The decision as well to make the game fixed-screen was also massively beneficial to selling the uneasiness in a 2D space, it places extra emphasis on immersing yourself in the soundscape, listening to every little audio cue you can to clue yourself in on what's just around the corner. It also makes the game feel even larger than it already is, while still retaining measurable distance over "samey" screens such as certain waterways or long tunnels. The sound design is also masterful, and it synergizes perfectly with the visual presentation and how it impacts progression through the game.

With the Downpour DLC came a free update that optionally adds visual cues to things such as indicators for off-screen enemies and accessibility options such as an engine slowdown similar to Celeste's; I always appreciate these sorts of updates and tweaks on principle, but I would still strongly urge trying to Survive and to only tweak downwards if you're genuinely going to drop the game over these gripes. They are in my mind what makes Rain World, Rain World, but I get it.

Ultimately Rain World represents what I believe to be the most realized vision in a video game, period. Unfaltering in its indifference to the player and their usual power-trip-seeking behavior and drip feed dopamine addictions, it has a story to tell about the cutest thing in the world simply trying to reunite in an apocalyptic world, and it refuses every step of the way to fall victim to dissonance between its gameplay and its narrative.


In one word: Primordial.

The more time passes, the harder it is to wrestle with how weird this game is to recommend. I want to tell all my sweaty action gamers to start on Critical and skip all the cutscenes, and I sincerely believe it’s the way to play, but I also realize that’s not a realistic experience whatsoever lmfao.

That intrusive thought verbalized, it doesn’t change how I feel about the game itself. Not sure I have more to add on top of what I said in this video: https://youtu.be/RYdFpv9Ay7o?si=XKNPVk13CDzhNzn8

the simultaneous success of dmc1 and failure of dmc2 put a variety of studios in an interesting position of "who could develop the next essential character action title," and team ninja really capitalized on the opportunity to try to take the crown. this became apparent to me during the game's beefy runtime considering just how different it is than today's CAGs in terms of structure and focus. in many ways ninja gaiden '04 (and by extension black) is a bold attempt at the 3D action-adventure game that sits as a bit of an evolutionary dead-end as CAGs have become increasingly combat-focused.

much of this can be seen in parallels between ninja gaiden and devil may cry 3, which released just a year after the former. dmc3 still retains some of the resident evil-derived exploration and interconnected world of its predecessors, but really pushes towards a focus on flashy and dynamic combat at every turn. ninja gaiden instead truly leaned into the scenario design that dmc1 played around with. instead of focusing on a single castle awkwardly chopped up into different missions, ninja gaiden generally focuses each chapter on a completely different area of the map, serving as mini-dungeons with their own puzzles, unique hazards, enemy types. as the game comes to a close these areas become more and more entwined, but the level segmentation is tremendously well-planned, and it clears up much ambiguity over what key items work where and how to progress the story. thanks to helpful notes from ayane, objectives are explained to the player, and if you pay any attention to the various locks around the game world it's easy to keep track of where to go. this kind of simple exploration really does wonders to vary the pace and avoid the trap of "I just did combat arenas for an hour straight and it's all memory-sludge in my head."

on the contrary, each scenario sticks out so precisely in my head that it's a wonder they came up with so many excellent ideas. taking down the airship by destroying the main generators and slaughtering a cyborg on the roof while flying through a storm, descending into the crypt and rising back up to place the chalice at the altar only to fight a giant skeleton monster and watch it crash to the bottom, where the floor has collapsed into a egyptian-tinged den of traps, the half-ice half-magma section where you switch between sides in order to eventually relight a furnace; hell, they nailed a 40-minute underwater section that serves as a great breather between non-stop action otherwise. unlike devil may cry, ninja gaiden manages to entirely switch its mood and playstyle on a dime when it wants to and surprise the player with some new idea at every turn. the variety kept me going "one more save point" over and over again where most CAGs generally lose me at 60/90 minute sessions, which is a credit to how itagaki and the team understood how to perfectly scale the intensity of an adventure over a 16 hour runtime.

of course, with so many great ideas thrown into the mix, there's bound to be some bad ones, and unfortunately there are some major clunkers to be found here. the entire military base raid for one is suffocating both in how the enemy forces exclusively use firearms which stagger the battle pacing and how there are not one but three different vehicular bosses in this section, all of which necessitate use of the clunky OOT-style first-person bow aiming controls. another particular sore point for me near the end of the game was the dreadful phantom pirahna labyrinth, where the game explicitly tells you to spam your flail light attack to avoid needing to mash out of their bite grab that bleeds your health as more and more leech onto you. the fact that these fish will blip into existence as others die really ratchets up the frustration during these sections, which frankly I died to the most during the last two chapters of the game (and ghost doku lmao). smaller puzzles and platforming sections will likely jolt you as the game wears on much as they did me, and it occasionally requires taking a break and coming back to it another time when the brutality of failure in one of these less-than-fair sections becomes too heavy.

thankfully overall I found the rest of the game surprisingly fair. given that ryu's powerful ultimate technique attack harnesses "essence" drops from enemies (colored orb drops shamelessly cribbed from dmc), currency and health drops are exceedingly frequent, and you will have an abundance of currency and health at any point in the game so long as you balance your UT usage. health consumables are also in fair supply, and the smaller variety are rather cheap at the in-game store; I virtually always bought a full set without significantly denting my essence reserves. save points are smartly placed in high foot-traffic locations and replaying sections will rarely take more than a few minutes as long as you diligently save. this latter point is frequently harped upon as a major problem with the game, and I have to assume people bothered by it are simply not used to having to replay sections of games at all; the devs thankfully didn't force us to restart the mission entirely as in dmc1 or pre-SE dmc3, and the backtracking necessary in most levels makes frequently passing by save points common. on that note: I never found the lack of lock-on frustrating given that ryu's soft lock-on is intelligent and don't really see a need for it in a game without firearms anyway. rarely did I whiff without it being completely my fault, and if you're blocking and countering frequently you should already be lined up with your foe.

I've neglected to discuss the combat given that I wanted to highlight less-discussed elements of the game, but yes, the combat is phenomenal. another aspect of this game's odd branch off of dmc1 is its more traditional and less combo-focused combat that instead favors rapid decision-making and opponent reads. it's not surprising to me that this was developed by a fighting game studio, because the combat first-and-foremost reminded me of something like tekken. in those games I am frequently weaving around opponents, waiting for openings, and then seizing on split-second encounters to score a juicy launcher and deal the bulk of my damage. here it's the same, removed from the one-to-one paradigm to instead wider groups of enemies all vying to attack you at once. izuna drop was my bread-and-butter and securing that light attack (or simply whiffing and catching them with the launcher) was my prize. much like a fighting game enemies will punish you for repetition of simple maneuvers or grab you out of a block frequently, and thus staying instinctual and reactive during combat is a must. it's free-flowing and easy to come up with new links or surprise enders given ryu's strong moveset and stable of weaponry, and while I primarily stuck to dragon sword during my playthrough, I still felt like I was discovering new things and improving the more I played.

a few minor quibbles about the combat (mainly controls) that I would like to bring up however: for one, the density of different button combinations and the amount of contextual ones meant that it occasionally felt like a crapshoot on what would actually come out at any given time. flying swallow was a particularly noticeable one for me, as I didn't really nail down its use until I started incorporating the homing jump into my play, and even afterwards it was a bit dodgy even when I was clearly pointing towards an opponent. incorporating the platforming moves in also never came naturally to me, especially since ryu's wall-run can be spotty on whether it will go vertical or horizontal, and the battle camera further aggravates that. in general I would say ryu's platforming is a little slippery: for 2004/5 it's excellent and an improvement on dmc1's stiff movement and a contemporary like sands of time's rigid traversal system, but it definitely shows its age and results in some unneeded frustration.

bosses are also mostly good, but not particularly ambitious in terms of movesets of fight setup. larger bosses mostly boil down to typical [look to the windup -> dodge the move -> spam an attack for a bit] flowcharts that leave a bit to be desired, especially for bosses like the dragon where I felt like I was mainly just chipping away with the windmill shuriken. humanoid bosses tend to suffer from ryu being too fast and dynamic and the boss needing to stay somewhat calm in order to be fair and predictable, and thus to balance it away from ryu's favor there's a lot of indeterminate blocking that I found annoying and perplexing. using two flying swallows back to back with one clinking off and one hitting makes little sense, and it ruins some of the immersion of trying to figure out the fight when it feels like random chance and not taking advantage of openings. none of them are particularly bad though (except the tanks/copters), and all of them are absolutely manageable or even a little easier than you would expect given the notorious difficulty of the game. the endgame fight where you float on a chunk of rock was also way better than the space harrier section of dmc1, so major props to team ninja there.

there's so much to discuss with this game, and I'm not sure I can even accurately give an opinion given that I've only done a normal playthrough and will likely not go for a hard run anytime in the near future (that's just not how I play games unfortunately). there's so much more content waiting for me when I get back, and hopefully I will given how much this game has imprinted on my brain. equally invigorating, anxiety-inducing, thrilling, and frustrating in all the best of ways.

Potentially the best Tengo Project/Natsume title and easily my favorite of the bunch. Know I’m intensely biased towards Wild Guns, as it represented one of the first arcade (or, maybe more accurately, arcade-inspired) games I ever put any amount of time into, but it’s one I’ve only grown more fond of as I’ve played through it over the years.

If there’s anything close to a lynchpin mechanic here, it’s the Vulcan Meter- shoot down enemy bullets and you’ll build up towards a mode that allows for a brief period of invincibility and extra damage, crucial for some of the tougher bosses. The catch is that it triggers automatically, so the process of managing it upends some of your dominant defensive strategies: it’d be easy to get complacent and shoot down every bullet, but because this will leave you exposed on some of the tougher fights, tools like jumping, rolling, and bombing all get their place in the defensive hierarchy- anything to avoid unnecessarily activating it. It additionally helps in keeping some of the repeat fights engaging thanks to inherent variance brought on by these shifting priorities: fight the dualist miniboss on the first stage, and there’s little reason not to try and stun him to build a bit of extra meter, but later encounters will have you re-examining those same strategies, frantically jumping around shots you’d be otherwise tempted to counter.

It’s a system that also speaks to the game’s uncommon generosity among arcade titles, the combination of the game-changing nature of the Vulcan Meter and the way an extra life is always in reach thanks to the low scoring threshold making for runs where pushing through attempts, instead of endlessly resetting them, is encouraged. Improvisational too! Aside from the deviations of the brought on the player’s fluctuating resources, there are also randomized item drops, sometimes granting you extra-powerful weapons, sometimes saddling you with the near-useless “P-Shooter.” The result is a game where multiple runs can vary wildly from one another- and given that a successful clear only takes about 30 minutes to complete, makes the prospect of going for another one endlessly inviting.

I could go on: the way the scoring system encourages you to stay one place and keep firing to build and maintain your multiplier, eating through your special weapon’s ammo in the process, or the constant temptation to lasso and stun targets- a move that'll leave you momentarily vulnerable, but can be so, so worthwhile on some of the deadlier fights.

Remarkable that everything here gels so well together, especially given that many of the other Natsume titles at the time had the benefit of being based on older titles and the tight window of time for the game’s development. Would never have guessed its origin would be so unromantic, but regardless, the end result is one of my favorite titles of all time- if you’ve been thinking of jumping into this style of game, I’m hard-pressed to think of a better entry point.

///

While we were waiting for our next subcontracted work to begin, our boss told us to develop a new game with two conditions: quickly and cheaply. I believe from the initial planning to the finished app it took us about five months. I’m sorry the answer is so boring!

Wild Guns creator Shunichi Taniguchi https://shmuplations.com/wildguns/

When I started planning out my next run through the game while I was stuck on the final boss, I knew I was in. When I started looking up the utility of individual moves and philosophy behind Wude, I knew this had jumped to becoming something very special for me. It took a long time to get here- the joke being that it took the game being sold as a beat ‘em up for me to pick it back up again (thanks Raph!), but I’m glad I did.

One of the most telling things for me was when I started considering it alongside some of my favorite games: Thought a lot about ZeroRanger and the way it secretly primes you to see the appeal of going for a one-credit clear- and the basic logic of the death counter serves a similar function, getting you to really take note of your successes and failures as you plot out the best way through each of the levels: Take a shortcut, even if it means missing a few helpful upgrade statues, or try your hand at some of the tougher encounters, where you might take some extra deaths but end up improving chances of success long-term.

It’s a great tension, something that keeps the appeal of a 1CC alive over multiple levels, though I think some levels capitalize on it better than others- the third stage is the obvious standout, with immediate access to the boss and complete control over which encounters you want to pursue. Out of all the levels, it’s the one where those questions change the most on each new attempt, debating the merits of trying to hit every Shrine in order to best prepare for one of the toughest fights in the game. It’s a spike in complexity and difficulty that I don’t think the game is ever able to quite match again, and between it and the incredible production of Stage 2, as you flicker in and out of reality, I’m sort of inclined to say that the game peaks by the halfway point- though the latter half is still excellent, and further speaks to the way the systems here bolster so much of the action.

Even a fight that might seem inconsequential in the moment has ramifications that can really spiral outwards: what weapons you used, how much damage you took, whether or not you used your Focus meter and now have to build it back up again- and with the ability for some enemies to activate a more powerful second form if you perform a takedown on them, even the simplest interaction in the game carries some considerations with it. So often with action games, I find you need to set some intrinsic goals for yourself- the reality being that playing the game as-is never really pushes the mechanics far enough. I feel like I’ve spent more time on the structure of Sifu than on the granularities of combat, but that’s partly because it’s the framing of these mechanics that really pushes the game over the edge.

The game’s unusual cohesiveness reminds me a lot of Inaba discussing Hideki Kamyia’s directing- that the quality of his games wasn’t just the result of raw talent, but time and effort and a wide variety of inspirations. Reading on Slo-Clap’s variety of influences, from arcade games to martial arts cinema to drawing from the experience Pak Mei experts, it’s hard not to see Inaba’s sentiment borne out here as well- just a remarkable amount of consideration for everything here.

Some stray thoughts:

- I have no basis for this, but I feel like the destructible physics objects were something that someone fought like hell to keep, and if so, I’m totally glad they did. Running around and breaking chairs is legitimately one of the high points of the game- maybe made all the better by how measured the game is otherwise.

- The Arenas mode reminds me a lot of what originally liked about the Streets of Rage 4 DLC, a complimentary mode that’ll keep the game evergreen. Especially like the way cheats and gameplay modifiers are added to some of the scenarios, really distinguishes the fights from one another and shows a side of the game I’d never opt into normally. I think the biggest compliment I can pay the mode is that, when I needed a break from Sifu, I just played Sifu: Arenas.

- Camera is far better than most action games, with certain objects in the environment going transparent for the sake of clarity, but the few moments where it lapses feel far more devastating (especially given the penchant for being cornered in Sean’s arena) given the penalty for failure is so high- Oni once again reigns supreme among 3D action games with its camera design.

- And this is intensely pedantic, but I spent a lot of time wondering if the aging mechanic should have been framed a bit differently- I love the idea of becoming a wise old master, but in practice, getting that old is normally a sign of a lifetime of chronic screw-ups. Something that tapped into the sense that you were unnaturally re-animated, like the corruption of your character in Shadow of the Colossus, becoming more pallid and deathly over time (and as an added bonus, it could work as another cinematic reference in a game full of them- add The Crow to the list!), but at that point, when I’m workshopping some alternate way of visualizing these systems, it’s probably more a sign of how deep in I am than some major fault of the game.


All that said, there’s not really a clean way for me to close this out- I told myself I could only talk about the game if I could clear it at under 50 years old, but in doing so, I realized how little of the game I could really speak to- terrifying for the purposes writing on it, wonderful for the realization that there’s so much left to uncover. In a genre full of games with unspoken caveats and that hint at what could be, Sifu ends up being something remarkably fully-formed, unusually complete.

///

References:

But instead of being some genius with ideas coming out of thin air, it’s more due to his nature and he gets ideas from a lot of different places and inspiration from diverse things and he thinks about things for a long time… his process needs space.

- Producer Atsushi Inaba, discussing Hideki Kamyia’s directing style, From 1UP

Both Oldboy and The Raid were inspirations for the design and art teams working on Sifu. In general, the game is filled with references and inspirations from many action & martial arts movies, such as John Wick, Sha Po Lang/Killzone, and Tom-Yum-Goong/The Protector… There were many different inspirations and references for the team, but for gameplay two good examples are Sekiro: Shadows Die Twice as well as God Hand on the PS2.

- From a Q&A with GameRant

The style we are focused on for Sifu is the Pak Mei style. Jordan (our creative director) and myself, as well as others in the studio have various backgrounds in martial arts but Jordan specifically trained in the Pak Mei style. We’ve worked closely with a sifu here in France to make sure the details of the Pak Mei study are represented accurately in the game. There’s so much history and nuance across all of the styles but we’ve been super inspired by diving deep into this one for Sifu.

- SloClap Executive Producer Pierre Tarno, on the martial arts depicted in the game, From mp1st

///

Some additional resources on the game:

How to enjoy Sifu (thanks Ziad!)

Sifu - Why you should use... Charged Backfist

text by tim rogers

★★★★

I hope you guys had fun reading this website, because I’m probably not going to bother writing on it anymore!



The reason would be Treasure’s Bangai-O Spirits, a game so perfect and so deep that when it comes right down to the wire, I’d probably rather be sealed in a sensory deprivation tank playing it against my best friend for millenia than working, vacationing in the Bahamas, or even having pseudo-violent sex. That’s the sign of a good game, right there — if you’d willingly give up an opportunity to let a girl live out her fantasy to rape a man just so that you could deflect projectiles for a couple more hours. I get this feeling, playing this game — if I keep deflecting projectiles, eventually I’ll be so good at deflecting projectiles that I won’t even need to close my eyes or tense my knees to ejaculate. I will simply be. I will flow.

Maybe that’s not enough for the videogame company PR dudes who troll this website (hi dad!), so let’s put it into more back-of-boxy language:

Bangai-oh Spirits is “Brain Training” for God. He spake unto the earth: “I am the Alpha and the Omega, the beginning and the end, the First and the Last, the Street Fighter II and the Street Fighter III, the Sin and Punishment and the Ikaruga, the carbon dioxide and the oxygen, the Bangai-oh Spirits for the Nintendo DS.”

This isn’t just the best game Treasure has ever made — hell, it isn’t even just the best game they can make, it’s a 210-gun salute to the head of the pathology they’ve waved proudly for the last decade and a half of their illustrious careers designing quirky little games that only morbidly obese people like myself seem capable of enjoying. It’s them dropping all their cards — five aces, including the elusive Golden Ace of Knives — on the table and putting a pistol in their mouth and saying heck you, I will pull this trigger, I swear I will.

This is hardly hyperbole.

Bangai-O Spirits is a puzzle game, a shooting game, a fighting game, an MMORPG. It is Pac-Man and Super Mario Bros. and Street Fighter III at the same time. In this age of Everyday Shooter and dozens of other one-man games that let you use two analog sticks to move and shoot at the same time, Treasure steps in with an update to their own Robotron riff, and they manage to outclass literally everyone, even without the multidirectional shooting gimmick.

What’s so good about the game? It’s kind of hard to put it in a few words, so we’ll have to break with tradition and actually discuss the way the game plays — not yet, though, don’t worry. I know, you wish I’d just keep talking about my penis for the duration of the review, though hey, sometimes we humans have to endure boring stuff.

The tutorial in Bangai-O Spirits is pretty boring. However, unlike an actual school, the long-winded tutorial in Bangai-O Spirits teaches you how to be objectively awesome. Each tutorial stage opens with a dialogue between three characters — a commander with an eyepatch, a young girl with large breasts, and a too-cool youth in a hip military uniform. In addition to teaching you how to use weapons, the commander also manages to enrage the cool youth: as the tutorials get more complicated and wordy, toward the end, the kid literally says that he didn’t realize he signed up for “an actual robot anime” — he just thought the robot and the anime characters on the box were there to draw in anime fans. The professor then gets defensive, insinuating that the end of Evangelion was “the best part”. Not making this up! When you beat the tutorial, the commander says, “We’ve defeated the Tutorial Army!” And the kid says, “That was fast. If we sell this game back now, do you think we can get at least 3,500 yen for it?”

Bangai-O Spirits: It burns!

The professor — and the girl — are quick to point out that the game has over 160 stages in the “Free play” mode, and that you can edit your own stages and play multiplayer, if you want. The kid shrugs this off, says that “stage-editing is for losers”.

And just like that — you’ll never see those three anime stereotypes again.

This is refreshing, when you consider that Treasure games have always had problems expressing story and character. Dynamite Headdy, their attempt to break into the mainstream with a “mascot” character, was a huge jumbled cluster of near-impossible challenges; it requires some degree of arcane voodoo magic to even get past the tutorial in that game, and it requires nerves of steel to like any of the bizarre characters. The music is Chinese water-torture. (The game itself is pretty alright.)

Treasure has gone through the weirdest phases of wanting to be popular and not giving a heck. Gunstar Heroes was an attempt to capitalize on a perceived opening in the action game market for people who didn’t like the look of Contra. It was apparently conceived as a kind of Sonic the Hedgehog with guns and fighting-game inspired close-quarters combat. Gunstar Heroes thrives whenever it’s not trying to tell a story, or be otherwise interesting. Each stage plays differently from the last, with a short shoot-em-up runup to the gimmick segments or boss parade: the Dice Palace stage, cited as the “best stage ever!” by game “journalists” with maybe a quarter-piece of their cranium missing, is all at once Treasure’s fragmentary genius and infuriating shortcomings rolled into one. It’s a line of pleasant little challenges — one-offs involving their pristine game engine — interrupted by an absolutely lazy hub where you roll a dice to move through a “board game” that’s just a straight line.

Gunstar Heroes begins to shine in stage five, which is titled “DESTROY THEM ALL” — contrary to what the lobotomites writing top-hundred lists will tell you, “DESTROY THEM ALL” is the “best stage ever”: all semblance of “level design” falls away, in the best way possible, and it’s just you, an entire enemy army, and occasional platforms. Flying enemies drop bombs, you jump up, grab them out of the air, and chuck them at towering robot walkers. Grunts run full-steam at you, and if you see an opening, you grab one and use him as a projectile, too. You do this over and over again for ten minutes, and the word “ENJOY” never vanishes from the stock-market ticker running across the bottom of your brain.

Why can’t Gunstar Heroes just be six straight hours of this? Why can’t it just never end? Let the computer just keep crunching the numbers and stuffting out bad guys — the more I kill, the more the computer stuffs. Zen Masters will tell you that there has to be balance, that without low points, high points wouldn’t feel as special, shadow lends context to light, et cetera, though really, Pac-Man Championship Edition is non-stop fun, and anyone who complains because there aren’t any NPCs to talk to is probably also a chimpanzee.

At any rate, Gunstar Heroes was something of a failure in Japan because

1. Most people on earth literally can’t read
2. It was on the MegaDrive, which was essentially the Xbox of its day.

Back then, people who owned MegaDrives in Japan were called “MegaDrivers”, just as the first wave of Xbox players called themselves “Xboxers”. Dynamite Headdy was Treasure’s idea of appealing to a broader audience — at least, in theory — and when the three-year-olds of the world found out that the instruction manual contained more words than “Happy” and “Fun”, it all went south.

Thus spake Alien Soldier, a schizophrenic, shattered velvet bag of demonic battle fragments. The full title of Alien Solider is, and I quote, “VISUAL SHOCK! SOUND SHOCK! NOW IS THE TIME TO THE 68000 HEART ON FIRE ALIEN SOLDER FOR MEGADRIVERS CUSTOM“. In other words, they’re weren’t lying to themselves — or anyone, really. Alien Soldier‘s title screen contains perhaps the longest story crawl of the 16-bit era, and no amount of resetting the console will make it any less ridiculous. Watching the superplays embedded on the Sega Ages compilation disc are like attending a waterski wedding on the surface of the sun. The game’s arcane mechanics are a Treasure staple: encouraging the player to play in the most ridiculous, unattractive way possible. To wit: the strongest attack at your disposal isn’t the flaming, screen-spanning dash that you can only perform with full health (and even then, only once) — it’s the pixels-wide flames that erupt from your character’s back as he’s doing said dash. Playing the game to completion is equal to breaking it — finding the pixels that work, freezing there, and TKO’ing every opponent in the first instant of the fight.

Eventually, Treasure tried to make money, and mostly failed, because they thought a game needed a “story” to be a phenomenal hit, and the story in the relatively high-budget Sin and Punishment was so loopy and bizarre that you can’t overlook it; to say it was written on acid or after smoking much reefer is a criminal overstatment. It’s more like something written in one day by a man with really bad indigestion cascading into diarrhea. For its mechanics, it’s easily one of the best games of all-time, and though the cut-scenes are fiercely skippable, the very fact that it has a story (and that its music sounds like pornography for librarians) keeps it out of the specialest place in my heart. It also — and this is pure conjecture — kept the game from being released outside Japan, even though it was written entirely in English and intended primarily for US release.

All throughout the years, as Treasure floundered with the weirdest cartoon licenses (Tiny Toon Adventures, for god’s sake), I wished they’d just make a full-on fighting game with Gunstar Heroes mechanics. They kind of almost did that with Yuyu Hakusho for MegaDrive, and later Guardian Heroes for Saturn — and most recently two Bleach-based fighting games on the Nintendo DS — though the games always felt too focused on “personality” and “selling points”. And they never set the world on fire.

Treasure’s Ikaruga, which resulted seemingly from a sudden desire to make the cleanest, simplest shooting game of all time as a kind of counterpoint to their Radiant Silvergun, the clusterheckiest shooter ever conceived, ironically, was perhaps their most internationally acclaimed game; maybe this has had something to do with Treasure’s recent game design roll-backs.

Bangai-Oh! for Dreamcast and the Nintendo 64, has always been my favorite Treasure game, not so much for its “hilarious quirky dialogue” as for its crunchy shooting action. There’s very little fat in the game, and it’s quite an accomplishment that Bangai-O Spirits is even leaner.

ACTUALLY DISCUSSING THE GAME NOW

I’m not even going to bother discussing how this game differs from Bangai-Oh, because it doesn’t matter: Bangai-Oh hardly exists anymore. Instead, I’m going to try to explain to you why you should buy this game, and not something like, I don’t know, Super Smash Bros..

In Bangai-O Spirits, you control a giant robot — which happens to be only the size of one “block” of the playing field. I believe this was the core concept of the original game — control a cursor-sized player character that looks like a giant robot.

Your goal in this game is to destroy various targets in stages. A rough map of the stage, with the required targets highlighted, takes up the top screen. You move around on the lower screen.

You have five general kinds of attacks. Before each mission, you’re free to select which weapons you want to use.

Press the Y button to use one basic weapon, and the B button to use another one.

Press the A button to dash in the direction you’re facing. Press any direction while dashing to change direction.

Press the L button or the R button to use one of two “EX Attacks”. Hold the button to charge the EX attack, and release to unleash it. The more enemy bullets in your general area when you release the trigger, the fiercer the attack you’ll unleash. If you let go of the button just as a bullet is hitting you, the missiles emanating from your body will even increase in size.

There are seven types of basic weapons. None of them are “better” than any other. Each has a definite strength and a definite weakness. All bullets are capable of canceling out other bullets, meaning that “the best defense is a good offense” and “the best offense is a good defense” are more than just something karate teachers tell kids.

I will explain the weapon classes briefly:

1. Bound laser: Laser-like missiles that reflect off walls at 45-degree angles. Pros: Can be used to access unsuspecting enemies in tight spots. Cons: Very average strength.

2. Homing missiles: Missiles that chase enemies. Pros: will hunt enemies down for several seconds; no need to even aim, most of the time. Cons: Even weaker than the bound shot; sub-par mid-air turning radius.

3. Break shot: Special shots capable of canceling two enemy bullets instead of one. Pros: can be used to cut through an enemy’s bullet curtain. Cons: even weaker than homing shot.

4. Napalm shot: Super-powerful shot. Pros: relatively twice the power of a bound shot. Cons: very easily canceled — normal bound / homing shots can eat through two napalm shots before being canceled themselves; break shots can eat four napalm shots.

5. Sword: It’s a sword. Pros: ultra-powerful, ultra fast. Cons: ultra-short-range. Hard to move forward and close in while attacking.

6. Shield: It’s a shield. Position it by tapping a direction and pressing the attack button. Pros: eats any and all bullets endlessly. Stays in the direction where you put it. “Movable cover”, essentially. Doesn’t fade even when you fire your other weapon. Cons: kind of hard to get used to! (Though in a Treasure game, “Kind of hard to get used to” is often used to describe The One Thing The Game Designer Wants You To Do All The Time.)

7. Bat: A giant baseball bat. Pros: can deflect any projectiles back at the enemy. Can deflect dozens of projectiles at a time, even. Can even turn enemies into projectiles. Can also be used to hit physics objects (conveniently shaped like soccer balls, baseballs, basketballs). Cons: Kind of slow. Not much attack power on its own.

Some of these weapons can even be combined in Radiant Silvergun style — homing and bound, for example — by pressing both buttons at once.

Your default attack mode is “free” — you can move while shooting — and your robot can fly all around the stage, with delicious crispy coasting physics. However, should you tap an attack button twice and then hold it the second time, your character will enter “fixed” mode, and you’ll stop anywhere (even in mid-air) and be free to aim in any direction. Useful in so many ways.

Now I’ll explain the EX attacks. Man, I’m like IGN over here:

1. Homing: Fires a multitude of homing shots. Good for: wide-open spaces with lots of enemies.

2. Bound: Fires a multitude of bound shots. Good for: twisty tunnels and lots of projectiles.

3. Break: Fires a multitude of break shots. Good for: lots of bullets, feeling tricky.

4. Napalm: Fires a multitude of napalm shots. Good for: feeling lucky.

5. Freeze: Freezes enemy bullets in the air. The “Treasure Factor”. Good for: super-playing and “puzzle”-solving. The longer you hold the button, the longer the bullets freeze.

6. Reflect: Works like a bat, except in all directions all at once. Good for: people who know what the hell they’re doing.

All EX attacks (except Freeze) can be “focused” in a direction of your choosing. In the previous Bangai-Oh game, you could only blast the bullets in a circle around your character’s body. In Spirits, you can press a directional button in one of eight ways while charging your attack, and then release the button to unleash the attack in only that direction.

And that’s it. That’s the game’s engine. The instruction manual doesn’t even describe the weapons — it just tells you to play the tutorial and find out for yourself. It knows that if you’re holding that instruction manual, you’re a Treasure-gaming freak who doesn’t read manuals anyway.

The tutorial contains the game’s only instances of “Story”. Its little lessons cascade into something huge, and the last stage is a frustrating, amazing, idiotic feat of level design with an infuriating execution and an exhilirating climax. Once you get past it, you’re free — as you’ve always been — to play the “Free Play” mode, which has, yes, well over a hundred and four tens of stages. You can play the stages in any order. You don’t even have to clear one stage to move on to the next. The interface is DOS-like in its cleanliness. Press the right directional button to advance forward one stage. If you’ve cleared the stage, your best time and top score will be displayed at the bottom of the screen. If you want to play a stage, press the A button. Now you have the DOS-like weapon-select screen. Choose your arsenal and play the stage. If you die, sucks to be you. Choose “exit” to go back to the stage-select menu. Try different weapons.

Many stages are beatable with any weapon selection; few absolutely require use of the bat to obtain the “easy” solutions. Many stages feature hang-ups that force you to open the menu and restart. Some stages feature extended side-scrolls climaxing in epic one-on-five battles with steel-trap-minded opponents of equal abilities. Some stages are massive labyrinths with multiple targets and pursuing grunts. Some stages are vertical scrolls that force you to avoid cascading unbreakable blocks while breaking the correct breakable blocks in order to reach the target. Other stages begin with the player surrounded at point-blank range by four giant honkin’ beam cannons; the 3-2-1 countdown to the start of the stage is basically a countdown to the trigger being pulled. You have literally half of one second to figure out how to win.

All of the praises heaped on Wario Ware instead belong to this game; the yokels of the world jump on Wario for featuring Nintendo Characters, who are virtually like game journalists’ “childhood friends”, only better, because they actually exist. Wario Ware has “amazing” stages that begin with a screen of The Legend of Zelda — there’s Link, and there’s a cave in a wall. When the game starts, you have five seconds to get into the cave. Get in, and the game pats you on the head and moves to another challenge. Sure, there are some nice things to say for the vision of Wario Ware, though at the end of the day, it doesn’t make the player feel spiritually complete, much less any smarter: there’s something to be said for the game’s glee in forcing the player to sharpen his mind to a point where he can acknowledge in a microsecond what he’s supposed to do, though the game ultimately fails in the eyes of god (and me) for never telling the player why. Bangai-O Spirits is about a son-of-a-bitching giant robot with fantastic weapons; your character is a soldier fighting the bad guys. You have to survive so that you can fight again. The only way to survive is to DESTROY THEM ALL.

Sometimes it’s a puzzle game, sometimes it’s a platformer. Sometimes it’s a hardcore bullet-hell shooter. Sometimes it’s a fighting game meant to be played by robots with laser-eyes. It is Every Genre. Most of the time, it doesn’t matter what it is, because it’s Bangai-O. It’s mind-expanding. It’s a diamond-sharp core engine joyfully level-designed gang-bang-style by a team of dudes who have been putting the “crunch” in “scrunch” for over a decade now. On the one hand, they don’t give a stuff what order you play the levels in. On the other hand, they need you to play this game every day for the rest of your life. For this purpose, they have created a level-edit mode.

The level-edit mode is more than just level-design — it’s a shockingly deep elementary course in game design. I described the weapons above, right? If you have any imagination at all, your mind should be teeming with possibilities. Make a puzzle stage, make a rock-hard traditional-style platform stage. If you’ve got a crush on the “Reflect” EX attack, make a stage that puts a new spin on it. The game is saintly in its generosity. Like I said above, Treasure aren’t just shooting themselves in the foot with this game — they’re shooting themselves in the throat. If you be a man who loves hardcore action games, you don’t need anything else.

The platinum icing on this golden cake, however, is the multiplayer mode. I spent six minutes throwing together a long rectangular stage with soccer balls and two evil enemy Bangai-O clones in it, and then played a deathmatch with a friend. We used bats and EX reflect, and just had a screaming orgy, squealing like distorted schoolchildren for literally 25 rounds in a row. Remember when you were ten, and you played River City Ransom with your buddy, and you stood on a garbage can and your buddy kicked the garbage can, which made you slide toward the edge of the screen, so you jumped and kicked a guy in the head? Then, years later, remember when you first played Halo co-op, and your friend got in the driver’s seat of the Warthog and you got on the turret, and you guys drove around hecking stuff up? Well, even in this day and age where four friends can get on Xbox Live and play Halo 3 transcontinentally, with one guy driving the warthog, another guy manning the turret, a third in the passenger’s seat firing a sniper rifle, and a fourth on a Ghost flying around shooting lasers — Bangai-O Spirits manages to provide gaggingly huge thrills with tiny sprites and solid-color backgrounds. It’ll make you remember Halo, and then remember River City Ransom, and then remember Bangai-O Spirits — the game you are playing right now. At the low points, it’s like playing the best game ever; at the high points, what is on the screen in your hands is pretty much the same thing that huge-headed aliens see when they mind-heck each other for sleepless, foodless dozens of hours on end.





What might go down as the game’s biggest “mainstream” innovation is its ability to transmit data via audio files. It says on the back of the box: “For the first time in history, custom map and reload data can be transmitted as sound files.” That’s kind of neat. Here I’d thought, for the longest time, that Treasure despised the aural sense (what with the brapping scrapping scraping nonsense-bullstuff music and limp sound effects that tend to populate their games, this one kind-of included). It’s nice to see them acknowledging that human beings have ears, I suppose. To load map data, you put it on an iPod and then just hold a headphone over the DS microphone. Sounds wacky, I know. The first time it actually worked, it kind of scared me a little bit. I wouldn’t stuff my pants with amazement if this cute little feature, above all else, is what makes Treasure world-famous after all. Failing that — well, listen to a stage for yourself. Game or no game, I know some people who would pay good money to stand in a dark room and listen to grown men in Darth Vader suits make sounds like this for two hours and call it “music”. Hell, one of those people is me.

At present, the Amazon.co.jp score for this game is three and a half out of five stars. The first review on the page describes the game as fun and nifty and all, though the tutorial is perhaps too confusing for beginners, so he has to give the game two stars. Boo-hoo. Another reviewer says, and I quote, “I wish they would have just ported the N64 version.” As a person who has played and loved the N64 and Dreamcast versions both, I say that this game is easily the better version. Not only that, you might as well not even compare it to the originals. It is its own game. And besides, everyone knows that the guys who write reviews of games on Amazon.co.jp are coke-bottle-glasses-wearing greasy-haired horse-teethed freakazoids who wear dirty pale-green electrician’s uniforms all day, every day, even though they don’t have a job.

I really don’t know what else to say to recommend this game to you. Don’t think about it in relation to any other games. Consider this game, for all intents and purposes, the only game in existence. If you read the description of weapons and it sounded purely awesome, if you’ve played every Kenta Cho game and often tell people “Yeah, this stuff is better than that Xbox stuff”, if you bought an Xbox 360 because Jeff Minter designed the music visualizer, if you know why Panzer Dragoon Zwei is exponentially better than Rez, if you possess enough cold hard evidence of the heart to prove in front of a jury that Cave Story murdered Super Metroid and Castlevania: Symphony of the Night, then even if you be a female, you are my brother, and we will join for great justice and we will love this game until and even after the end times come, until and even after we lose our corporeal bodies due to the explosion of the sun or nuclear war. We will live on, in spirit — in Bangai-O spirit.

This game is a joyful disease. Pass it on.

In other words:

Absolute highest recommendation.

–tim rogers

addendum: I just revisited the instruction manual and read the profiles of the three “characters” in the “story” mode. It says of Ruri, the girl, that it “seems as though she’s not in love with Masato”. It says of Masato, the boy, that it “seems as though he’s not Ruri’s brother”. Yeah, that’s not just game-of-the-year-worthy story exposition — that’s the Best thing ever, right there.

addendum the second: To clarify, the multiplayer mode doesn’t allow the two players to kill one another. Yes, it’s tragic, because this game would essentially destroy Senko no Ronde in that case, and, in all ways imaginable, become the Greatest Game Ever. The multiplayer mode is, instead, about shooting bad guys together, competing for points on a stage-for-stage basis. If you set up a custom stage so that there are a bunch of evil Bangai-O clones, working together is pretty fun! Like, really pretty fun. Amazingly fun, sometimes, especially if you set up a stage that, scientifically, just isn’t possible to win, and then proceed to lose fifty times in a row, until you suddenly win ten times in a row. Balancing a VS mode for something like this would have been code-murder. It would have been like making a whole separate game, which I’m sure Treasure didn’t have time for.