First released in 1984, Star Force is one of the early shoot'em ups that have not yet followed the trend initiated by Xevious (1982), thus maintaining a very minimalist approach in its gameplay. In this regard, the game exemplifies the early days of Tecmo – then still known as Tehkan. The player is in command of a starship, the Final Star, whose goal is to destroy the planet Gordess and its defenders, in a succession of twenty-four levels, each represented by a letter of the Greek alphabet. With few frills, the title only requires the player to shoot down waves of opponents with their lasers. The only upgrade item allows the rate of fire to be increased, but the title does not offer any new weapons or protective features. This somewhat archaic design increases the difficulty of the title, as each projectile can be fatal, but it allows the player to get back into the game reasonably quickly, without suffering a complete setback.

Several issues nevertheless plague the game: firstly, the ad nauseam repetition of enemy waves during twenty-four levels quickly tires, since nothing breaks the monotony of the gameplay loop: the end-level bosses are excessively bland and do not provide any additional challenge. On the other hand, the American port is surprisingly difficult compared to the Japanese version. Indeed, the speed of the enemies has been greatly increased, so that the first stages already have the difficulty of the Japanese mid-game. The screen quickly becomes saturated with projectiles, which are particularly tough to avoid, as all the opponents shoot towards the player. Unlike a bullet hell where the projectiles follow comprehensible patterns, here the player has to avoid shots that come from both the front and the back, with very limited room to manoeuvre. Despite the Final Star's high rate of fire, it is easy for enemies to get through the barrage of shots, quickly making the situation uncontrollable, especially because the starship moves so slowly. These structural problems become particularly critical in the final levels.

Star Force thus struggles to impress, although it is not a particularly bad game: it suffers from a simplistic nature, perhaps too pronounced. This assessment has surely been made by Tecmo, since Super Star Force (1986) complexified the formula, with an emphasis on exploration, in rather unexpected isometric views.

While Super Pitfall was intended as an extension of the formula developed by David Crane with Pitfall II: Lost Caverns (1984), it is perhaps surprising that their critical reception was the polar opposite. Whereas Pitfall II is seen as a flagship of the Atari 2600 for its ambitious platform and exploration gameplay, Super Pitfall is widely regarded by the general public as one of the worst titles on the NES. This difference in opinion can be explained by the difference in technical limitations, which make certain features of the gameplay far less excusable for an NES title, but also by comparison with other games, which elevated players' expectations.

The player controls Harry, who has gone on an expedition to the Andes Mountains, both to save his lion Quickclaw and his niece Rhonda, but also to find the Raj diamond – leaving aside the cultural contradiction, which only seems to illustrate the orientalism of the Western creators. Unlike Pitfall II, where the final objective is visible from the very first screens, Super Pitfall places the player at the top of a gargantuan level, with no indication. Only by starting to explore can the player begin to get their bearings. This beginning is not unlike Metroid (1986), but a critical difference is that the first chamber of the latter is closed. In practice, there is only one way forward: find the Morph Ball on the left, before progressing eastwards and exploring the rest of Brinstar. Super Pitfall does not respect this progressive widening of the scope of exploration. Instead, it is possible to go to the caves on the right or to dive to the depths, without any hints as to what the objectives are.

The main map of Super Pitfall is divided into four sections, which are particularly artificial. Unfortunately, exploration is not unrestricted, as some roads are blocked by doors that the player can only open if they have the associated keys. Such keys are scattered throughout the caverns, but they are invisible and may only be revealed by jumping erratically near them. This highly random and cryptic approach is not unique to the 1980s exploration video game, but it makes it feel cruel and impossible to progress without a guide. When one is perused, it appears that exploration is simply a matter of meandering through the guts in a very specific order, while revealing all the important items. In reality, Super Pitfall is particularly linear and exploration would be almost boring, if the location of objects were openly communicated by the game.

Indeed, beyond its cryptic nature, Super Pitfall has a rather execrable control system, even for NES standards. Harry feels either too heavy or too floaty; collisions are generally poor and the game handles ladders and crouching very badly. While it is possible to shoot enemies, it does not allow to attack ground-crawling foes, making progression frustratingly slow at times. The redirection of jumps in the air is erratic, so that sequences with vines are always awkward. Other more minor issues punctuate the title and worsen the archaic nature of the exploration.

This is a shame, because the structure of Super Pitfall could have resulted, in theory, in a decent game. Secret passages are never properly communicated - one must either jump on birds that look like the standard enemies or jump to a specific location without any indication - yet it leads to visually different sections. The Dark World where Rhonda is imprisoned has the true trappings of a final level, with much more powerful enemies: the atmosphere is quite different with huts perched on huge trees growing inside a cave. In the middle of this section, a sort of massive temple adorned with improbable skulls serves as a prison for the hero's niece. It's a memorable passage, within a generally bland game, and it's unfortunate that the overwhelming majority of players never reached this mystical place, due to the cryptic nature of the game design.

It's hard to find qualities in Super Pitfall, as the execution is so poor. Even with the best will and arming oneself with paper to draw a map – in the pure tradition of early Western games – the title remains stodgy. It's worth noting that in 2016, for the game's thirtieth anniversary, Nesrocks released an extensive romhack that addressed the game's plethoric flaws. The result is far from an exceptional game, but it is certainly much more accessible to the general public. This is probably what Super Pitfall should have been.

Ocarina of Time is one of those gems adored by the gaming community; time has had little effect on its reputation and it was interesting to revisit it with an fresh eye in 2021. For a transition into 3D, the Zelda franchise gives itself an exceptional performance, taking the core ideas of A Link to the Past, while bringing in the allegorical freshness introduced by Link's Awakening. Indeed, it is impressive to see how the title builds a coherent universe, which makes for a very metaphysical reading. Link undergoes a real tragedy, which is only readable thanks to the overhanging aspect of our status as spectators. The fable told by the game unfolds in several stages, making very skilful use of time travel to establish a discourse on its endless flow and the irretrievable flight from the present. The adult sequences do an excellent job of repetition, to accustom our gaze to the ruins and desolation - mimicking Link's loss of innocence – even as the wounds of the world are gradually healed. Mechanically, the game is very interesting and already offers dungeons of a rare quality – the one in the Forest, in particular –, which reuses the different items well. One can certainly have reserves about the concept of mini-dungeons, which always end a little prematurely, but the change of atmosphere in the different zones is very appreciable. Some elements of quality of life have suffered a little over time: it is difficult to return to Z-targetting after the dynamic camera of Breath of the Wild, and it is regrettable that teleportation does not arrive until so late; the extent of the Hyrule plain is certainly impressive at the time of the release of the game, but it is somewhat painful to cross with Link as a child. Besides, a nice philosophical fresco is laid out before our eyes: it's less about Link than about what he represents to the average teenager, through the cluster of characters we meet. If the game suggests that Link is the individual outside of society, it is clear that each secondary character is particularly bizarre or has an atypical personality. Link is thus faced with a number of different guests, who paint a strange world, where his own benevolence appears to be out of the ordinary. Link accepts, to the point of sacrificing his identity. Ocarina of Time thus manages to draw up a subversive discourse on the world and its tragedy, even before Majora's Mask. In fact, it remains an excellent title that is particularly relevant today.

Making a sequel to the "best game ever" is a daunting task, especially when it has to be done in record time. Yet this is the crazy challenge that Aonuma is taking on, at the cost of his and his team's energy. If Ocarina of Time's approach was to create a golden standard for the series and the video game world, Majora's Mask has a more carnal and intimate feel, which is made possible by the main time loop mechanic. This allows for a smaller, but more chiseled cast of characters with their own timeline to focus on. It is the exploration of this tight world and the discovery of its inhabitants that occupies the central part of the title: admittedly, our objective is largely indicated from the start – to prevent the Moon from crashing into Termina and to retrieve Majora's Mask from the hands of Skull Kid –, but Termina acts as a gigantic social dungeon, where you have to get to know the NPCs in order to progress. The mechanics of time appear as a driving force of the playful and dramatic tension, instilling an urgency and sequencing the game phases, thanks to the owl statues. At the same time, the masks act as rewards and specific tools in the progression. The counterpart is the contraction of the number of dungeons, since there are now only four, albeit with a fairly substantial number of more or less optional mini-dungeons. This also reduces the number of non-mask items, compacting the exploration mechanics in favour of interaction with NPCs through masks. This time management and its consequences make full use of the video game medium and are difficult to transcribe elsewhere. A real social web is woven through the encounters with the inhabitants of Termina, which allows human and realistic sensibilities to come to the fore, as in the long quest of Anju and Kafei, which begins as soon as we discover the Romani ranch. Here, it should be noted that developmental angst has infused the game's atmosphere. The first hour is particularly confusing - in its gameplay and narrative. The game has a mysterious and eerie aesthetic, even if it is due to the frame-jumps of the NPCs and the inhuman expressions and movements of certain characters. Unlike Ocarina of Time, we are an adult in a child's body, as Link witnesses the absurdity of the world, which he does not necessarily understand. Majora's Mask paints a human picture of the end of the world, which echoes contemporary concerns. The spectre of death looms constantly, as does the emphasis on the remains, souls and emotions caused by death. Yet the promise of a new dawn is often apparent: the title discusses the importance of friendship, love and the cycle of life. It is in the search for a sense of existence that the game shines: with each return to time, the relationships Link has created with the inhabitants start anew, but the memories live on, as the mask seller points out in the concluding cinematic. As such, Fierce Deity Link is emblematic of Link's goodness and self-sacrifice in his quest to ease the anxieties of the inhabitants of Termina, opening the way for him to answer the simple and profound questions asked by the Moon Children. Majora's Mask is a rough, but grand and timeless game, with an unparalleled elegance in its setting. There was a before and after to Majora's Mask – if only because of Aonuma's prominence – and the Zelda franchise will be able to see new heights to climb.

Played with BertKnot.

As crime fiction is a genuine institution in Japan, it is not uncommon to see directors or scriptwriters trying their hand at it, on a whim. We could multiply the examples, so common is it: we have to admit that video games and dramas often see attempts, more or less successful, to tell detective stories. The Centennial Case seems to more or less follow this trend. Although the director, Koichiro Ito, had already delivered 428: Shibuya Scramble (2008) and TRICK×LOGIC (2010), the rest of the team seems to be quite new to the field. Junichi Ehara is most known for NieR: Automata (2017), while scenario writer Yasuhito Tachibana is primarily a director of TV series, which have never come close to the investigative genres. Of course, one should not too easily prejudge the quality of The Centennial Case on the basis of these elements: after all, The Portopia Serial Murder Case (1983) is one of the very first games by Yuuji Horii (Dragon Quest), while Famicom Tantei Club: Kieta Koukeisha (1988) was the brainchild of Yoshio Sakamoto (Metroid).

This background seems necessary to understand what the game is attempting to achieve. An FMV detective drama, which interweaves several cases into one another, it has a structure that is very reminiscent of the spirit of Japanese dramas. This is understandable given the presence of Tachibana as lead writer. The story features a crime novelist, Haruka Kagami, who is invited to investigate the Shijima estate. A skeleton has been found under the centuries-old cherry tree and the writer is asked to find the identity of the corpse. The investigation takes a mysterious turn when the family elder is poisoned just after a traditional ceremony. A murder follows the incident and Haruka begins to investigate, soon discovering that there seems to be a link between the mysteries of the present and others that have punctuated the Shijima family's history for the last century. To discover the truth about the present, she therefore solves thoses past cases, presented to her through several hanshi-bon. This approach allows the game to be chaptered and to alternate between different eras. The distinctive feature is that the game reuses the same actors across the periods, a choice that lies at the heart of the title's stylistic and narrative proposition.

Such an approach is not in vain and allows to create rather interesting atmospheres, that borrow from a certain Japanese traditionalism. In a way, it would have been difficult to replicate this impression with a visual novel. However, the production remains generally average and the acting is never brilliant. Some would say that, in this respect, the game succeeds quite well in imitating the dramas that punctuate Japanese television schedules. The issue lies in the implementation of interactivity. Unlike Ace Attorney, where the player has the latitude to explore and where dialogue allows the investigation to progress organically, The Centennial Case places the player in an aggravated state of passivity. During the FMV sequences, it is entirely possible to put the controller down and do nothing: the game does try to force interactivity with some sort of QTE to retrieve clues at various moments, but they are, in any case, given automatically just before the Reasoning Phase. Similarly, some discussions require the player to choose an answer, but this has no influence on the course of events. As such, the game admits that it must be viewed and solved like a film, a fact that does not necessarily speak in its favour.

The Reasoning Phases themselves are a game design disaster. In essence, it's all about matching clues to questions, which leads to the acquisition of hypotheses. These can be used in the final phase of each chapter to confront the culprit. On paper, the idea could work, but its implementation is absurdly cumbersome. The volume of clues given is excessively large, so that it is difficult to get a clear picture, while many of the hypotheses are false leads, more or less stupid. One could consider that it is feasible to discover only the relevant hypotheses so as to save time, but another problem arises. This is because the game keeps its twist for the conclusive phase, and therefore the hypotheses do not usually point directly to the solution. It is therefore difficult to determine which ones are necessary and which ones are superfluous.

As for the mysteries, they are at best artificial and at worst implausible. Chapters 1 and 4 are trivial, as they reveal the solution to the mystery too unsubtly, while chapter 2 makes little sense in the modus operandi. The only exception is Chapter 3, whose nightclub investigation is rather well crafted: it relies on a trope familiar to those used to crime fiction, but the execution is convincing, insofar as the psychological element is well taken into account. The title also tries to renew its gameplay with chapter 5 which, in its escape game logic, is very reminiscent of the Zero Escape series. Once again, the idea is not wrong, but the gameplay remains very clumsy: the hypothesis creation sequences are even more frustrating, while the puzzles struggle to be really engaging.

For the rest, The Centennial Case fails to convince. In its last chapter, Haruka's deductive line is particularly hazy and is based solely on intuition. No formal element allows her to reach the truth and the resolution is thus more disappointing than expected. Indeed, it is not very difficult to identify the final culprit by the second chapter, but the ease with which they concede is frustrating. This is the general impression that remains for every case. While the Reasoning Phase is a particularly long and uninteresting process, the presentation of the findings always seems too quick and never rigorous enough. Of course, lay players will appreciate the twists in each case, but veterans will see the contrivances very easily, spoiling the potential of the title.

The Centennial Case thus offers a rather strange story, which blends shin honkaku a la Kindaichi shounen no jikenbo and Detective Conan, with a curious armchair approach. An almost metatextual reflection on detective fiction and its writing is shaped – the epilogue, of outstanding quality with a real subtlety in the detective writing, makes it clear why the title is in FMV –, but the game didn't manage to carry its idea through. Indeed, Haruka's deductions require the various cases recounted in the books to be absolutely true. This is somewhat a discursive failure, preventing the discussion of the falsification of the detective story and the value of the truth it carries: this theme is addressed much more frontally by Umineko no Naku Koro ni (2007), building on the considerations that surrounded Christie's And Then There Were None (1939).

For all that, The Centennial Case is far from uninteresting. It is representative of Japanese culture's love of detective stories, as well as its historical importance in video game production. A very large number of titles have been released on the various Japanese platforms, but few have been translated. In the best of cases, amateur translations make it possible to discover certain titles: this is the case for Kamaitachi no Yoru (1994). More recently, the localisation of Famicom Detective Club (2021) perhaps prefigures a turning point for the Japanese detective video game and its overseas release. The Centennial Case seems to pursue this objective, while being a resolutely Japanese detective fiction, whether it be in the family corporations, the attraction for past traditions, the debt of blood, the melancholy of existence, the place of women in this society or the weight of immortality. The game thus opens a very interesting window on an important part of Japanese cultural production. If the attempt, because of its gameplay, fails to completely live up to its ambitions, let us hope that it paves the way for other titles.

Played with BertKnot.

The release of Higurashi no Naku Koro ni (2002-2006) undoubtedly marks a cultural turning point for the visual novel genre and for Japanese animation. Far from being a confidential success, Higurashi stood out as a unique synthesis of the horror and mystery genres within video games. We could go on and explain that the title is part of the moe turn, to the point of imposing a new artistic standard for the medium, but I would be rambling. The imprint left by the series is exceptional and goes hand in hand with that of Umineko no Naku Koro ni (2007-2010); however, the latter's much more complex structure and its strong intertextuality with crime literature make it a more hermetic title. In any case, many works have followed in their wake. More than a decade later, the waves can still be felt in World End Syndrome, which hardly hides its inspiration. Multiple references punctuate the story, both by openly using the titles of Ryukishi07's series in the protagonist's monologues or by playing on familiar aesthetic details (Sacred Land/Golden Land, the cry of the kite).

World End Syndrome is in fact almost exclusively a rehash of previous ideas, which gives a very old-school feel to it. Formally, the story appears to be a very loose rewriting of Ikeru shikabane no shi (1989) by Masaya Yamaguchi: the premise is that of the dead coming back to life and interfering in the society of the living. In the game, this serves as a pretext for the legend of the Yomibito, the dead who come back every hundred years during the summer. Beyond that, the title is meant to be an otome game, whose structure it takes up in its second part. The idea is to complete each route – by dating, more or less, each girl –, in order to get the true ending. This one reveals the mysteries of the city of Mihate and the truth behind the murders. Mechanically, the otome phases seem to be a return to a very archaic past: it's all about going to this place or running into that girl, to increase our stats with her. There's an almost RNG aspect to it that makes the encounters rather limp, as the characters don't really interact with each other during the free phases. They exchange two lines of dialogue before the other character decides to leave. For a game released after the turning point that was Persona 5 (2016), the weakness of these mechanics raises questions about the validity and interest of the title to have chosen it as a core gameplay mechanic.

As I said, those mechanics are extremely minimal and rely only on very hackneyed japanimation tropes: here comes the tomboy, the aristocrat or the candid girl, among others. Of course, these are the same models as Neon Genesis Evangelion (1995), but those have dominated fiction for almost three decades. Similarly, the title falls into cousin incest and the sexualisation of the teacher: World End Syndrome is much more restrained than other visual novels, but it only offers the negative sides of the genre. In fact, if the result is relatively harmless, it is still infuriating. Discursively, the game tries to offer a discussion on a certain metaphysical idea of life and death. These are not the only opposing themes, as there is a contrast between the modernity experienced by Japan – technological progress, entertainment industry and tourism – and the folklore and local traditions coming from shinto philosophy. What is perhaps more regrettable is the game's inability to form a sense of contemplation through its otherwise beautifully composed backgrounds. Yet this is an idea that Japanese fiction has captured well in works set in towns and cities far from major metropolitan centres: Boku no Natsuyasumi (2000) is a striking example, while some of the Ghilbli films openly play on it (Omoide poro poro, 1991; Umi ga kikoeru, 1993). Beyond that, World End Syndrome repeats its August structure ad nauseam over five different routes. The same places are visited for the same purposes, which brings a sharp sense of fatigue. This lack of thematic dialectic in visual novels is nothing new, but it hurts when the game openly cites Umineko, which offered an interesting take on detective writing, in relation to Christie's work.

Here, everything seems to be full of futility. No scene really hits the mark and the characters remain empty shells. The plot unfolds over twenty hours, but one hour, after the prologue, was enough to identify the culprit and to solve the major mysteries. The various twists thus suffer from a real weakness, especially as they are particularly questionable thematically and morally. For all that World End Syndrome attempts a complex metatextuality, it is notable that it lacks the means to fulfil these ambitions, no matter which aspect is examined. The title therefore remains a permanent frustration, in that the basic concept wasn't necessarily bad, but it could only be achieved with a more modern structure, which ditches the idea of a bland male protagonist, to allow for player identification. One could argue that this serves a purpose in the narrative structure, but it remains a failure. Yukino was doing the work of a detective and one can only think that she would have made a better protagonist. World End Syndrome doesn't say much, except that you should live, not lock yourself away in regret, and that you should be surrounded by your friends. Admittedly, these are candid ideas, but fiction has already exhausted these themes in a much more elegant way.

In 1982, Doug Smith, then a student at the University of Washington, began developing a game to pass the time. The result, after some trials and tribulations with the publisher Brøderbund, is Lode Runner, published in 1983 for the Apple II. A forerunner in the puzzle platformer genre, the title may have been inspired by BurgerTime (1982), with which it shares the idea of a defenceless protagonist who deals with enemies by clever use of the platforms. Lode Runner thus features a hero who has to collect piles of gold, in a sort of underground network, whose different levels are full of ropes and ladders. Because the protagonist cannot attack, he has to get rid of the enemies either by go around them or by making them fall into holes. Those are created by digging into the various blocks of the levels. Enemies are dispatched if the block respawns while the enemy is still in the hole. Such a system encourages a delicate and prophylactic approach, in order to trap enemies at the right time. It's often a case of 'divide and conquer', so that one only has to deal with one or two enemies at a time: moreover, some situations can be more elegantly resolved by not killing the enemies, but locking them in a particular place, after abusing their pathfinding.

The basic concept is solid, which explains the success of the arcade title, as well as the consequent number of sequels and ports on different consoles and computers. The NES version tries as best it can to translate the arcade concept, but technical limitations have forced some structural modifications. First of all, the levels are vertically contracted, since the NES lacks two lines: this makes the levels marginally easier. Secondly, of the 150 levels available in the arcade, only the first fifty have been ported. Nevertheless, the title retains its level creator. Indeed, as soon as the prototype game was up and running, Smith shared the level editor with the people around him; the story goes that he also shared it with the kids in his neighbourhood to finish the 150 levels required by Brøderbund. As such, the level editor is a pioneer for the NES, since it predates the ones from Excitebike (November 1984) and Mach Rider (1985).

There are a few things that illustrate the antiquity of the title, notably the handling of hitboxes and collisions. For example, it is noticeable that the digging animation is cancelled, if an enemy gets too close to the square being dug, resulting in a quick death. This forces the player to anticipate the arrival of enemies from a distance: the problem is that the NES' horizontal scrolling only works when the avatar is very close to the edge of the screen. As such, it's easy to be surprised by an enemy one didn't know was so close to the protagonist – this is even exacerbated by the respawn of enemies at a random location. The gameplay is still stiff and enemy movement can be sometimes erratic and incomprehensible, especially when the character is one line below them. These problems are not critical and it is possible to adapt to them, but they somewhat pose unnecessary complications. Lode Runner is still an interesting title – if only for its pioneering nature –, never technically bad, but it is hard to see the point of playing this particular version nowadays.

The concept of Arkanoid is well known as it is directly based on Breakout (1976), which illustrates the timeless and intuitive nature of this genre in the world of video games. More broadly, the title is in line with the Pong-like games that flourished in the arcades of the 1970s and 1980s. Indeed the necessary technique is rather simple: a few pixels to form a bar, a pixel that bounces at an angle dictated by the part of the racket touched and bricks that disappear on impact. The concept is crystal clear and simple. Perhaps too simple for the 1980s. The development of more ambitious titles, which leave more room for player development – think of all the upgrades that shoot'em ups offer during a game –, pushed Taito to refresh the Breakout concept for its own title.

Several powers can thus be collected: a ball splitter, lasers, a sticky racket to be able to aim at the relaunch or a ball slowdown. To justify these additions, Arkanoid features a space invasion plot, so that the paddle is not just a bunch of decontextualised pixels, but a spaceship, if that makes any sense. The game doesn't bother with any further explanation, until the final stage where a moai stands before us: this is the only real combat in the game, if we can use that qualifier. The target is massive and dodging the few projectiles is an easy matter. Fatally, therein lies Arkanoid's problem: if it tries to bring novelty, it remains marginal or does only the bare minimum to keep the audience entertained, but is never transcendent. For example, lasers are a welcome addition to counterbalance the low degree of latitude on the bounce angle – a characteristic feature of the NES, remember the frustration of not being able to change finely the angle of the shot in Golf. In Taito's defence, this brick-breaking concept is really simple. Simplistic even, making it hard to build on. Perhaps this is why later concepts often remain weakly innovative: Motion-Twin's AlphaBounce (2007) or Wizorb (2011) also struggle to innovate. In fact, Arkanoid remains a title that occupies a few minutes, but is never a memorable or even enjoyable experience.

Stupidly cruel in its difficulty, Athena appears to be a bastard descendant of Ghosts 'n Goblins. This is a first try for SNK, which was rather used to creating top down shooters, such as Ikari Warriors, Vanguard or Alpha Mission, to name only a few. The switch to a 2D platformer brings its share of challenges that obviously posed major issues, at least for the NES port. The concept is not uninteresting though. The title features Athena, a disillusioned princess, who decides to open a "door that must not be opened". It transports her into a fantasy world populated by enemies. The game thus is divided into seven different stages, each represented by a particular environment (forest, cave, sea, etc.), as well as a final one that acts as a boss rush. In the vein of Ghosts 'n Goblins, our character starts out with no equipment and must build up an arsenal by killing enemies and breaking blocks that may contain upgrades. Each weapon follows its own upgrade tree, which specialises Athena towards a particular archetype. In addition to weapons, armour pieces and various other items that aid progression must also be found.

All of this sounds rather nice, but the problem is in the execution. Not unlike King Arthur, Athena is extremely rigid and refuses to attack and move at the same time, making any dodging extremely complex. This is anchored in a screen readability that is, in general, very poor. It often happens that we accidentally pick up a weapon dropped by an enemy, which sends us backwards in our progression, because the sprite is messily overlaid on the decor. There is also a general lack of feedback, especially on the final boss, where there is no indication that you hit your target or not. The game is very cryptic about how to progress: for the snowy world, it is necessary to equip Athena with a bow, the only weapon capable of hurting the boss; however, there is no hint about this, and the fact that an unenhanced bow is unable to break the blocks puts off the idea of equipping yourself with one. Similarly, in the labyrinth level, it is necessary to discover the statue of the goddess to obtain the Harp of Protection. But this is never made clear and it will be common to pick up a fake harp, which punctuates the level to trick the neophyte and downgrade their arsenal – a notable change from the arcade version. This arc of flaws adds to the already high difficulty of the title – the lack of invulnerability frames punishes the slightest mistake very heavily –, and makes it a particularly unpleasant experience, even though there is real potential and ambition.

Indeed, the scenery is particularly numerous and generally quite shimmering, which gives each stage an appeal. Of course, the whole thing remains a potpourri of western fantasy, which does not hesitate in mixing Greek ruins with submerged ziggurats or more technological spaces. The idea was certainly not to create a coherent universe, anyway, insofar as the title seeks above all to appeal to a male audience, with the character of Athena. If the American cover is set in the aesthetics of pulp films, the Japanese boxart literally features a teenage girl in a bikini. Athena was indeed published at a turning point of Japanese video games, where female anime characters were more liberally used – it was notably the case with Tenshitachi no Gogo (1985). We can thusly find the artistic style that flourished during the PC-88 era and that coincided with the success of Mamoru Oshii's adaptation of Urusei Yatsura (1981). As such, Athena appears to be a pragmatic and opportunistic attempt by SNK to expand its catalogue with other genres. If the attempt was unsuccessful here, due to a complete lack of quality control – the absence of the proper ending screen shows the little concern for quality –, the following years saw a diversification of genres, with fighting games (Street Smart, Gang Wars, Fatal Fury), RPGs (Crystalis) and numerous sports and simulation games. As for action-platformers, SNK published a sequel to Athena, called Psycho Soldier, but without real success. The formula was mostly reinvented with Blue's Journey (1991) and Top Hunter: Roddy & Cathy (1994), which foreshadowed the success of Metal Slug. But that's another story.

Originally published in 1983 for arcade machines, Elevator Action precedes the turn of the 1990s espionage fiction and, in the context of video games, the revolution of the spy genre in the late 1980s. The time is not yet ripe for stealth mechanics (excluding the seminal Castle Wolfenstein, Metal Gear has just been released when Elevator Action is ported to the NES) and the title borrows liberally from pulp fiction: amusingly, the Japanese boxart is deliberately comic, while the US cover is more realistic, somewhat evoking Harry Palmer.

The protagonist Otto starts at the top of a thirty-storey building and he has to climb it down while stealing the secret plans, the location of which is indicated by the red doors. While he is carrying out his mission, enemy agents come out of the blue doors and try to kill the spy immediately. There is no real stealth, as you are generally spotted at the beginning of a level, the enemies thusly chasing you directly. The core of the gameplay therefore lies in the ability to choose a particular timing to enter the red rooms and elevators, while getting rid of the enemies in any way you can – bullets, high-kicks or more creative approaches. The concept is quite simple and easy to grasp, but it's hard not to feel a certain boredom by the end of the first level. The title only increases its difficulty by speeding up the enemies' reflexes; sure, the lift patterns are changed, but the environments remain the same.

In fact, such an approach even creates frustration. The game never provides a map of the building – even if only with a vertical scrolling at the beginning of a level – so it is impossible to construct an ideal route. It's not uncommon to take a lift that leads to a dead end, forcing you to turn around and suffer through the endless spawn of enemies. Also, some actions respond in a sub-optimal way: to crouch you have to press down, but the character does not get up immediately. For that, you have to press up, which wastes time when the goal is to jump directly. The imput detection seems to strict when taking the stairs, causing some chaotic and unwanted fights. This arc of shortcomings leads to serious problems when the action gets messy on screen and signals a general lack of quality of life. Elevator Action isn't bad, but it never stands out in its execution, which remains in the average range of NES ports. The potential is however there and gave rise to a second, more polished iteration, Elevator Action Returns, in 1994.

A year after the release of Ninja Hayate, Taito published Kage no Densetsu (1995) on arcade machines. The gameplay is quite different with the former being an interactive film – not unlike Dragon's Lair (1983), the inspiration is crystal clear –, while The Legend of Kage appears to be a very classic action platformer. The ambitions are certainly not very high for this title, which is confirmed by the length of a game: about twenty minutes, at most. You play as Kage, a shinobu in a fictitious, misidentified Japan: his goal is to free Princess Kiri, captured by a renegade samurai, Yoshirou Yukigusa. The manual for the NES version suggests that the action takes place at the end of the Edo period, but it's hard to see this kind of plot fitting into the bakumatsu timeline. Instead, the game seems to draw on Japan's feudal folklore. This is a general trait of Japanese fiction towards shinobu and samurai stories, which it perceives with a certain poetic licence.

Moreover, The Legend of Kage is a very formal title: it consists of five repeating levels, each with its own particular objective. You have to get rid of a monk in the forest, before infiltrating a castle. This requires killing a certain number of enemies or simply reaching the end of the level. The game attempts to play on verticality, capitalising on Kage's jumping height, but the execution leaves a bit to be desired, as air control is almost non-existent. Most of the action is thereby about preemptively getting rid of opponents with our shurikens, unlimited in their number. It is possible to parry adverse attacks – except fireballs – with our own blade, but its exetremely short range requires sharp reflexes. The gameplay loop is particularly short, which prevents the player from feeling the slightest fatigue. As soon as one level is completed, environment changes. When you save the princess, the game loops from the forest, and all the colours change to mimic the passage of the seasons, from summer to autumn and then winter. This is not an unusual touch in games from the 1980s, but it is still a welcome addition for the sake of visual diversity. As such, the game is rather charming and, albeit simplistic, an effective representation of the fantasy of medieval Japan.

It's hard to say much more than that, as the title is in the underbelly of the productions of this era. The gameplay never shows a real spark of genius, but is decently implemented. In some respects, it's a shame that The Legend of Kage isn't a bit more similar to Ninja Hayate, which offered more creative and comical scenes. Thus, Kage no Densetsu acts as a neglected little brother, without ever being bad.

The genesis of Metroid is relatively well known and takes place at the time when the Famicom saw an increase in storage capacity for games. Two directions were taken: the first is well known, it is The Legend of Zelda (1986), with the Dream Team in charge of the project – Miyamoto, Tezuka, Nakago, to name but only a few. A few months after, Metroid was also intended to be a non-linear title in its progression, but for the action-platformer genre. The team was different, as it was led by Satoru Okada and Yoshio Sakamoto, who had won their spurs on ports of arcade titles.

The scenario is more elaborate than the majority of the titles of the time and is backed with an extensive lore in the manual: we play as Samus Aran, who has to find the Metroids, stolen by the space pirates. These are holed up on the planet Zebes, which serves as the main exploration site. If the influence of The Legend of Zelda, in the non-linear construction of the world and the quest for objects to facilitate exploration, is obvious, Metroid innovates by its claustrophobic aspect. The cramped environments force anxiety, which the music often compliments. Brinstar, as the first locale, has heroic flights in its melody, while the bass in Kraid's Hideout creates anxiety with throbbing arpeggios and an ostinato that mimics a heartbeat. The soundtrack clearly plays on this carnal side, especially in Ridley's Hideout or in the secret rooms: strangeness is invited by chaos or silence, which take hold of the throat. The influence of Alien (1979) is undeniable and documented.

It is this mysterious, almost frightening side that accompanies the exploration of Zebes. Our initial, very limited arsenal prevents us from moving around as we please. The initial inability to attack ground-crawling enemies makes some sequences more complex than they seem, and it stays a perennial problem. Missiles are an effective expedient against more powerful enemies, but their number is limited, pushing a conservative approach when using them. Only the Screw Attack sets us free, but this is well hidden in a corner of Norfair. For the majority of the game, exploration thus remains a breathless challenge. This exploration, while intended to be organic, can nevertheless be broken down into several sequences – more or less interchangeable. From Brinstar, three zones radiate out and allow us to acquire upgrades. It is possible to explore Brinstar and then Kraid's Hideout; then Norfair and Ridley's Hideout, before tackling Tourian. This last area is the only one that is conditional on exploring other areas, as it is pegged to the death of Kraid and Ridley. Otherwise, it is possible to explore freely, although some passages require a certain amount of creative skill to progress without key items (Ice Beam, High Boots). We can already see the emerging potential of Metroid in this opus, especially the bomb jumping – though it's a bit different from the way it is done in Super Metroid, in that the game sometimes doesn't recognise the B button imput, when you want to drop a bomb in the air.

Beyond these qualities, Metroid is still a NES game and is limited by the technology of its time. Interviews have pointed out that storage was at a premium, so assets had to be reused to compress the size of the game. In practice, this results in rooms that are sometimes very similar, if not absolutely identical. Unfortunately, this makes exploration less iconic and, much more so than in Super Metroid, it seems essential to rely on an external map, as it is very easy to confuse one corridor with another. The lack of eight-way directional shooting feels limiting, especially as Samus can't redirect her gun in the air, leaving her sometimes defenceless against some fast-moving enemies. The game also lacks a bit of visual cueing to guide us through its world. Tourian requires possession of the Ice Beam to freeze the Metroids, but the title is never really clear on this point – admittedly, it places one in Norfair, in a more or less convenient location after facing Ridley, but this is somewhat inelegant. Similarly, some of the secret passages are viciously hidden and require, similar to The Legend of Zelda, the player to bomb every block to ensure it's not crumbly – the situation is even worse with the fake lava lakes. These little hiccups can detract from the quality of the exploration, especially as the difficulty remains generally high.

A quick word on the representation of gender: depending on the speed at which the game is completed, the ending screen may reveal that Samus is a woman; exceptional speed even shows her in a bikini. The inspiration is again Alien, with Ripley. Nevertheless, it is obvious that this choice is still situated in a very sexist environment, to the point where a non-negligible part of the public did not admit that Samus was a woman – Western magazines not helping on this point, notably because of the confusion surrounding the password 'JUSTIN BAILEY'. Fatally, the subversive side of Samus eroded quickly – perhaps less so for the manga? This is illustrated by her sexualisation in Zero Suit Samus, as well as in the themes of Metroid: Other M (2010). In any case, Metroid remains an essential game, even if it has not aged very well and Zero Mission (2004) is far more easily recommended.