Let’s make a time loop game. First, we need to establish a mystery, something that’ll really play into the strength of the format, with something new to discover each loop. Since we at Arkane have mastered the magical assassin concept, we’ll blend the ideas, and have players discover how to assassinate a list of targets across a repeating day.

But how do we prevent players from just lucking into a solution, going to the right places and beating the game in two hours? Dishonored was already criticized for being short, and if even 1% of players beat the game in one run, we’ll never hear the end of it. So, we’ll have to force some repetition: some necessary codes will be mutually-exclusive, so players will have to loop at least a few times before they’re able to unlock the ending. We’ll author a linear sequence of events that will guide the player and pace the experience.

What about players who get tired of the repetition though? It won’t take long for people to get tired of repeatedly fetching their favorite weapons. To solve that, we could have players preserve their loadout between runs… but that would mean that we need to add a little more depth to it, so they don’t just gather everything once and stop caring. The weapons could have randomized bonuses like a looter-shooter, and collectible trinket buffs as well. Adding in character buffs and loot rarity would ensure that there’s always something new to find each run.

Of course, that will work well with the invasion-based multiplayer. Everyone will be fighting a unique opponent, which is great. We can also kill two birds with one stone by limiting the amount of powers players can equip at one time, further emphasizing unique approaches and making gunfights easy to follow. Speaking of limitations however, there will need to be some sacrifices in the realm of map design, since having a one-on-one fight across sprawling maps with load zones would be a nightmare, especially if hiding on rooftops and turning invisible is on the table. So, we won’t have events progress in real time, just in a single time-of-day per mission, because we won’t know how long those encounters may last. It also wouldn’t be good to lock weapons and buffs behind the multiplayer system, because that would let expert assassins steamroll new players. As a final failsafe, we’ll include an option to only play single player, in case it devolves into an invisible sniper camp fest.

Great. This design makes sense from front to back. We’ve walked through all the decisions and how they fit with all the others. We’ll have a time loop game where… players preserve everything from loop to loop, with no time pressure to navigate a linear sequence of events. We’ll prevent players from being bored with excessive repetition by… having them farm currency and random items. They’ll do that until they feel comfortable with tackling the big challenges and handling multiplayer invasions, because losing to an invader resets all the progress on your current loop. You’ll only ever do it when you’re not trying to focus on completing the story, since multiplayer has no benefits compared to isolating yourself in single player.

Hold on, how did this happen? We made decisions that made perfect sense; why is everything so wrong? Why do all our systems work against themselves? I guess it’s because we started with some good ideas, like the time loop assassin stuff and spy-versus-spy multiplayer invasions, but then immediately focused on how to sterilize those core concepts for people who aren’t interested. We made a time loop game and then removed all the time pressure! We took the magical powers and intricate maps we’re great at creating, and saddled them enough limitations to where they're worse than our old games! We made those sacrifices so the multiplayer would work, and then disincentivized engaging with it, killing the point and the playerbase in one shot! Next time we try this, we gotta keep it simple. Focus on what we think is cool and commit to it. Start from scratch. Ok.

Let’s make a time loop game.

The last time I played this, around six years ago, the representative moment in Fallout 2 seemed to be killing the Hubologists as part of the Shi questline. The tiny, underleveled Tom Cruise sprite exploding into gore as his attending hooded cultists float taunts about how they'll kill you in the name of L. Ron Hubbard. The outrage-seeking, the easy satire, the inarticulate disaffection for any and all trends in contemporary social life, and the childlike joy in explicit violence all made the game seem not only like a perfect relic of the attitudes of the turn of the millennium, but of the milieu of American game development at the time. It's a moment at which the line between professional game development and say, flash games about torturing George Bush is shown to have been terribly thin.

On the last run, what stood out was the converse sense in which this is the progenitor of the post-millenial open-world game, remarkable in just how fully-formed the genre already is. The tight structuring of the first game is almost fully abandoned alongside the time limit: the only thing really keeping the player from the end sequence at any given time is the fact that it's on the opposite end of the map from the starting position, and the accompanying confidence that the player will want to consume all the content with which she's presented along the way. The notion of all this being in service to a main quest isn't totally dispensed with the way it is in the fully realized form of the genre, but it's treated as something of a running joke.

Jokes are really the stock and trade of Fallout 2, and range from Monkey Island-adjacent clownishness at best to something like an adult animated comedy at worst. A few stray lines genuinely work for me this time: the option to call a woman homophobic for rejecting you and a dialogue option reminiscent of Disco Elysium which only pops up if you've got meth in your inventory to offer a starving child.

The character of combat encounters in this game is really determined by the way armor and health progress more steadily than damage output, which plateaus around the middle of the game unless one uses perks to squeeze out an extra attack or two per round. The high-lethality casino shootouts which characterized the midgame for me were enormously fun in their demands that the player think carefully about range, action economy, and cover. The early game's typified by low-stakes slapfights with rats, and the later game is more of a flat DPS race between bullet sponges, but that sweet spot makes the combat engine seem like it has genuine potential as a TRPG.

Playing this game in a four-day stupor made it easier to see that the game gets worse as it goes on, that the relative focus and sense of tone informing the early areas falls apart around NCR and San Francisco. Navarro and the Oil Rig, in particular, while themed impeccably and nicely built up to over the course of the game, are remarkably empty as dungeons. This time, as with every other time I've played the game, I spoke to the scientist who releases the virus before talking to the one in the reactor room, because the former is placed before the latter, and had to spend twenty minutes looking for a bomb so that I could still trigger the escape sequence.

The totally disparate nature of the world it creates, made up of stereotypes and references and the conflicting aesthetic and thematic preferences of its authors, retains it appeal for me. It's like a constellation of what it meant to be a socially awkward man in the late-90s, like a Rifts campaign frozen in amber. It's commonplace to present New Vegas as the more authentic sequel to Fallout 2 than Bethesda's attempt, but each game's making an effort to pare down an incredibly eclectic product into something like a brand.

The last time I played this, around six years ago, the representative moment in Fallout 2 seemed to be killing the Hubologists as part of the Shi questline. The tiny, underleveled Tom Cruise sprite exploding into gore as his attending hooded cultists float taunts about how they'll kill you in the name of L. Ron Hubbard. The outrage-seeking, the easy satire, the inarticulate disaffection for any and all trends in contemporary social life, and the childlike joy in explicit violence all made the game seem not only like a perfect relic of the attitudes of the turn of the millennium, but of the milieu of American game development at the time. It's a moment at which the line between professional game development and say, flash games about torturing George Bush is shown to have been terribly thin.

On the last run, what stood out was the converse sense in which this is the progenitor of the post-millenial open-world game, remarkable in just how fully-formed the genre already is. The tight structuring of the first game is almost fully abandoned alongside the time limit: the only thing really keeping the player from the end sequence at any given time is the fact that it's on the opposite end of the map from the starting position, and the accompanying confidence that the player will want to consume all the content with which she's presented along the way. The notion of all this being in service to a main quest isn't totally dispensed with the way it is in the fully realized form of the genre, but it's treated as something of a running joke.

Jokes are really the stock and trade of Fallout 2, and range from Monkey Island-adjacent clownishness at best to something like an adult animated comedy at worst. A few stray lines genuinely work for me this time: the option to call a woman homophobic for rejecting you and a dialogue option reminiscent of Disco Elysium which only pops up if you've got meth in your inventory to offer a starving child.

The character of combat encounters in this game is really determined by the way armor and health progress more steadily than damage output, which plateaus around the middle of the game unless one uses perks to squeeze out an extra attack or two per round. The high-lethality casino shootouts which characterized the midgame for me were enormously fun in their demands that the player think carefully about range, action economy, and cover. The early game's typified by low-stakes slapfights with rats, and the later game is more of a flat DPS race between bullet sponges, but that sweet spot makes the combat engine seem like it has genuine potential as a TRPG.

Playing this game in a four-day stupor made it easier to see that the game gets worse as it goes on, that the relative focus and sense of tone informing the early areas falls apart around NCR and San Francisco. Navarro and the Oil Rig, in particular, while themed impeccably and nicely built up to over the course of the game, are remarkably empty as dungeons. This time, as with every other time I've played the game, I spoke to the scientist who releases the virus before talking to the one in the reactor room, because the former is placed before the latter, and had to spend twenty minutes looking for a bomb so that I could still trigger the escape sequence.

The totally disparate nature of the world it creates, made up of stereotypes and references and the conflicting aesthetic and thematic preferences of its authors, retains it appeal for me. It's like a constellation of what it meant to be a socially awkward man in the late-90s, like a Rifts campaign frozen in amber. It's commonplace to present New Vegas as the more authentic sequel to Fallout 2 than Bethesda's attempt, but each game's making an effort to pare down an incredibly eclectic product into something like a brand.