Six Shots: Heavy Bullets Is A Roguelike-like FPS

Look at this gorgeous video game

The revolver’s an ideal weapon for a roguelike-like: high damage, low capacity, slow to reload, and awfully cool. Three attributes that make it dovetail nicely with the caution, conservation, and permadeath one expects from the genre, plus a little radness as a treat. Heavy Bullets gives you a chunky revolver and six bullets–and only six. They’re special bullets you can collect and reuse, granted, but that’s little comfort when something’s between you and the five you sprayed in panic.

The roguelike-like FPS arrived on Steam Early Access and the Humble Store yesterday, and seems pretty fun from what I’ve played.

It has a simple goal: reach a security mainframe through eight procedurally-generated levels of malfunctioning defences and hostile creatures. Only the thrifty will survive.

Heavy Bullets runs on bullets and coins. Bullets kill enemies and enemies naturally drop coins. Coins go into vending machines dispensing health, extravagant single-use weapons, mysterious and useful items, and more bullets. You can’t afford everything, so you need to be careful and conservative. Six bullets is plenty if my aim is true, but wouldn’t a spare or two be nice? I think this weird gem I found will help somehow but I know for a fact that toolbox contains bombs, so which do I carry? A backpack may let me hold two items rather than one but it’ll cost all my cash, so can I play smart enough to survive at half-health?

There’s a little persistence through ATMs, where you can deposit and withdraw items and cash between games, or buy life insurance and wills to bequeath a bit to your next life. And you will have a next life. You’re only able to take a few hits, and encountering surprises like blocking enemies or obscuring grass for the first time may catch you out. Like many roguelike-likes, I imagine the first play through to the end will be a slow and tense affair, then you’ll realise it’s actually quite short (about half an hour once you know what you’re doing, I’d guess) and can get down to exploring oddities and generally showing off.

Heavy Bullets is flipping gorgeous too. Look at this video game! Bright colours, flat shading, chunky gradients, and a weird abstract indoor jungle, crumbs! It’s stunning. It sounds pleasant too, with music and bloopy noises from Samurai Gunn composer and top rapper/producer/poet Doseone.

I’ve only had time to play a little, busy bee that I am, so please don’t take this as a Wot I Think, but I do think you may want to consider a look at this video game. Early access discounts bring Heavy Bullets down to £5.94 on Steam and, curiously, £5.49 in the Humble Store. Supposedly it only needs a little balancing, a few bug fixes, and some small bits of content before it’ll officially launch.


  1. Bull0 says:

    Roguelike-like, there’s an irritating term. It’s got guns in it, is it a Doomlike-like?

    Lovely style, though, I might give this a go

    • rustybroomhandle says:

      Jet Set Willy was the ultimate Roguelike-like. Instead of one permadeath though, you got SEVEN permadeaths.

    • geerad says:

      I think that makes it a DoomRL-like-like.

    • LexW1 says:

      I think the issue is that if you just call it a Roguelike, you get SWAT’d by the Internet Gaming Terminology Police, who point out that technically, it isn’t a Roguelike (with varying levels of annoyance/aggression/conviction). Yet if you don’t use the term Roguelike, nobody gets what makes it special, because it’s a simple term which conveys a ton of information. Hence clumsy stuff like Roguelike-like. OH WELL.

      I’m glad to see the basic Roguelike concepts of procedural generation, exploration, permadeath and so on are becoming more popular and widespread, I must say. They’ll probably be the “RPG elements” of the later half of the 2010s, I suspect (“Roguelike elements”).

      • Bull0 says:

        Yeah, pretty sure that’s why it’s come about. It’s silly that if you divert from the “classic” interpretation of roguelike you get jumped on – why is it useful to call something that’s basically exactly the same as Rogue a “roguelike”? Wouldn’t “Rogue remake” or “Rogue clone” be just as effective at that point? Wouldn’t something that’s a bit like Rogue (eg procedural, difficult, permadeath) be a Roguelike?

        So yes, let’s just call it a Roguelike and be done with it. Or an FPS with roguelike mechanics, I like that.

        • geerad says:

          Except those roguelikes AREN’T “basically exactly the same as Rogue” in much the same way as all FPSes aren’t clones of Doom, and all RPGs aren’t clones of D&D. JRPGs are different from CRPGs and within those, a Final Fantasy is a different experience from a Tales game, and Skyrim is different from Ultima Underworld, which is different from Planescape: Torment. But Call of Duty isn’t an RPG even though you gain experience points.

          There’s actually a lot of variety within the traditional roguelike genre: Nethack plays differently from Angband or ADOM or Dungeon Crawl Stone Soup or Brogue or Powder or TOME 4. They’re all unmistakably descendants of Rogue, but playing each of them is a different experience.

          Nethack, for example, has something of a kitchen sink approach. (In fact, it literally contains a kitchen sink.) It has a ridiculous number of monsters and items. It contains both the swords Excalibur and Sting, for example, a Tourist class, and enemies ranging from Medusa, the horsemen of the apocalypse, and mindflayers. There is one dungeon, but it has several forks, including one where you play Sokoban. It’s known for it’s complex item interactions. For example: you can get hints at identifying wands by writing on the ground with them; you can do similarly with rings by dropping them in the sink. If you read a scroll while confused, it does something different. You can wield any item as a weapon, including a cockatrice corpse. (Wear gloves first!) There are a bunch of ways to steal from shops, some even without angering the shopkeeper (which it invented). You have to continually eat food or starve to death.

          TOME 4 is wildly different. It has a bunch of lore tying its world together and the enemies fit together more thematically. Rather than your options mostly being what scrolls and wands and potions you’re carrying, in TOME you spend points in talent trees to learn and improve talents. These talents have cooldowns and usually a resource cost, like stamina or mana. (Some items also give you similar abilities, but not to the extent of Nethack.) It has a world map with a number of dungeons on it, each much shallower than Nethack’s one. Items don’t need to be identified. There are shops, but they’re menu-based, so you can’t steal. You don’t have to eat food.

          Shiren the Wanderer isn’t “basically exactly the same as Rogue”, nor is Dungeons of Dreadmore nor Cardinal Quest. ToeJam and Earl DEFINITELY isn’t! (It’s realtime, supports two players, and you can’t attack enemies normally!) Hoplite is a roguelike for mobile phones with very non-traditional methods of attacking that completely changes how it plays.

          I think there is value in a category that includes these games but excludes Heavy Bullets or Risk of Rain or Spelunky (which is not to say any of those are bad games). I love Spelunky, but it is qualitatively different from the roguelikes listed above.)

          However, I’m not too bothered about exactly what the categories are called. It could be “roguelike” versus “roguelikelike” or “traditional roguelike” versus “hybrid roguelike” or “[something else] roguelike”. I think “roguelite” sounds kind of dumb, but I’d be on board with such games being called “roguish”.

          • Bull0 says:

            The point I was making is people use “Roguelikelike” because “Roguelike” is meant to mean games that share *all* those core tenets with Rogue, rather than just some of them. So FPSes aren’t roguelikes, they have to be roguelikelikes. And I think that whole thing is stupid. Hope that helps.

    • Alice O'Connor says:

      It’s partially a tongue-in-cheek term, of course. A spot of fun.

    • SaddySally says:

      All-in for Ponglike-like-like-like! :tired:

  2. Stormworm says:

    Would Roguelite be a better way to describe these type of games? I saw it being mentioned somewhere and thought it was a good one.

    • SuddenSight says:

      Yech. I mean, it’s fine, probably. Whatever gets the point across.

      I just react with revulsion I anything using the phrase “lite.” Yurch. Sorry, it’s involuntary.

      • geerad says:

        I think it’s kind of gross, too! Also, it’s maybe a little too easy to misread.

    • MojaveMusic says:

      People have come up with all sorts of terms for anything that takes elements from roguelikes and some combination of other genres. I have heard myself “roguelikelike”, “roguelite”, “PDL” (which is short for Procedural Death Labyrinth), and “new-gen roguelike”.

      They’re all relentlessly stupid buzzwords, but because people have the attention of flies, no one will stick around long enough if you just describe something as being “a first person shooter with roguelike and RPG elements”. That’s entirely too long.

  3. frightlever says:

    A revolver with special bullets?

    Outlaw Star?

  4. Michael Fogg says:

    So rogue-lites are a genre that’s currently happening, but the problem I see with it are the algorythms for procedureal level generation. They are just not good enough and produce bland and empty-ish spaces that get boring on the third run or so. That’s been true for almost all of these games since Spelunky started the craze (Spelunky of course managed to avert the problem and its levels are always fun obstacle courses). All the Towers of Guns and Eldritches and even Sirs have subsequently suffered from this problem and failed to realize their full potential. Ironically, what is meant to give those games always unlimited replayability ends up denying it, since the levels get seriously boring so quickly. Unless new titles manage to come up with better level-gen I hope the fad will fade soon.

    • Bull0 says:

      Well, procedural content isn’t all about replayability, although granted it’s often billed as thus. For me there’s a lot more satisfaction in exploring a unique environment that’s been created for me, by systems, than one that’s been hand-crafted and is a carbon copy of the next guy’s.

      There’s also the benefit that you can’t turn to a walkthrough to find where the next thing you need is, etc – there might be hints in guides or a pattern but you’ve still got to physically find it. It’s challenging, I think that’s a big part of the appeal.

    • Professor Paul1290 says:

      It’s not always supposed to give replayability, in a lot of games its more about rewarding the player for learning the game rather than memorizing the game.

      In that regard, something as “simple” (relatively speaking) as randomizing enemy and item positions can have a significant impact on gameplay. Even if you are playing the same level every time, the game becomes less about remembering where the enemies are and instead more about learning how to fight them in general. Sure you can memorize certain patterns in how they are placed, but the benefit of that for the amount of memorization you have to do is drastically reduced.

      In games that don’t do any procedural generation or randomization at all discussion tends to gravitate towards “that enemy hiding in that corner”. When you introduce some kind of procedural generation, even if it’s relatively minor, the discussion shifts away from “the enemy hiding in that corner” and towards “how to fight enemies hiding in the corners”. In a lot of games that makes a pretty big difference.

      • terriv says:

        Agree with this 100%

      • Synesthesia says:

        Exactly this. It’s a big part of the roguelike feel, learning and beating the game, getting actual muscle memory, not just memorizing sections.

    • Michael Fogg says:

      @Bull0, Paul

      I didn’t say anything agains procedural generation in general, in fact I like what it’s trying to achieve, what I don’t like in most cases is the execution, i. e. levels that end up being full of nothing. It takes some real brains to make an algorythm that outputs consistently interesting levels. Minecraft and Spelunky pull it off, but a lot of stuff like Don’t Starve for instance don’t, which leads to to questing if many of indie devs have a clue how to create a level-gen tool that does the job.

      • Bull0 says:

        Oh, for sure, there’s a lot of weak procedural generation goes on. I just thought I’d point out that replayability isn’t the whole idea most of the time. But you’re right, it can still hamstring even the first playthrough of a game if it’s done badly.

  5. Shooop says:

    Ah procedureal level generation. The ultimate crutch for people who can’t do level design for shit.

    • Professor Paul1290 says:

      You don’t always want level design to be the star of your game. Level design means that the level stays the same and in some cases that means memorizing a game rather than learning how to play it, and in some kinds of games that can be game breaking.

      A lot of “traditional rogue-likes” (using this term in this case for clarity) arguably get no replayability benefit from their level generation, but they can’t work without it. If they didn’t generate their levels then they’d become about “save that arrow for that enemy you meet at the bottom of the stairs”, or “grab the potion behind that hidden wall so you can use it on the boss”. It wouldn’t matter how fantastic the level design replacing the generation would be, the game would be mechanically ruined.

      Some strategy games wouldn’t work as well without some form of procedural generation, regardless of how skilled the potential “level design” might be. The original X-Com: UFO Defense would arguably not work with hand-crafted levels. AI War: Fleet Command would not work with hand-crafted maps, it would become a memorization game.
      Again, they don’t really get a replayability benefit from their level generation and what is generated isn’t particularly varied, but the games simply don’t work without it.

      Heck, even simple minor things like random enemy placement can make or break a game. Rainbow Six and SWAT would not work with static enemy placement, they just wouldn’t. They’d literally stop being “tactical shooters” entirely. It doesn’t matter how ingeniously placed the enemies would be, the fact that they would be consistent would be game ruining.

      I appreciate awesome level design as much as anyone else but sometimes I literally don’t want level design, as shocking as that might be to some. Sometimes I want the gameplay mechanics to be able to work as independent of the levels as possible. Sometimes I want to be encouraged to refine how I’m playing rather than figuring out where everything is. Procedural level generation allows for that option.

      • Shooop says:

        Enemy placements aren’t the whole of level design, if that’s what you’re getting at. And if the mechanics of a game can be ruined simply by the level being the same, then you should probably re-examine your mechanics because they must be pretty weak.

        To me, the goal of level design should be make a game world that works with the gameplay, not independently of it. Otherwise what’s the game boil down to?


        You either get lucky the random number generator makes you a map that you can get through or you don’t. You don’t lose the game because of your lack of skill playing it, but because the game decided you couldn’t win that time.

        Some random elements in a game can definitely help, but throwing the entirety of level design into a random number generator is a new level of pure laziness.

        • Nixitur says:

          If your level generator is capable of deciding that “you couldn’t win”, then you should probably re-examine your generation procedure. That’s the programmer’s fault; it’s not inherent to procedural generation.

          For example, I remember a Nethack player (Nethack being a game which is all about procedural level generation) who went on to beat Nethack with every single character class in a row. He didn’t restart until he managed a specific run, he just did it with no “failed runs” in-between.

          I think that’s enough evidence that there is very little luck involved in beating Nethack.

          Also, lazy? There is nothing “lazy” about writing a procedural level generator. Sure, there would possibly be less work involved in design or art, but there’d be so much more work involved in programming.

          A level generator is not a “crutch” if you can’t do level design. In fact, if you can’t do level design, your level generator is going to be shit, too.

          To write a passable level generator, you need to be fantastic at level design. After all, you need to be able to put all those intuitive measures of “This level is good.” or “This level is bad.” into numbers. You need to know exactly what makes levels good.

          If you’re designing a level by hand, you can get quite far by intuition, play-testing, tweaking the level and play-testing again. If you feel that a specific part has too few enemies or is too long, you can just change that specific part and you’re good.

          With a level generator, that is far more difficult. If you play one of the generated levels and you feel that a specific part has too few enemies or is too long, you need to know exactly what makes that “specific part” special. You need to put what makes that “specific part” special into numbers. You need to know how the generator could detect or just plain avoid parts like that. And that’s not enough, you also need to tell it exactly how it should tweak the level.

          So, really, there is nothing lazy about procedural generation.

        • Professor Paul1290 says:

          Just about any game that depends on the player learning the rules of the game rather than memorizing the content can be put at a disadvantage if the levels stay the same.

          Needing to have levels have some kind of random element is not a sign of weak mechanics.
          If your levels are procedurally generated then you usually can’t rely on level design to carry your game, your game has to carry itself on its mechanics.

          “Good level design”, while a positive thing overall, can often act as a crutch to make up for weak game mechanics just as story or graphics can. It allows the developer to make up for edge cases where the gameplay might not hold up well.

          Also, if you can’t cope with “luck” then you are not really “skilled”. As someone who plays a lot of tactical shooters and air combat sims and has instructed others on how to better play such games, I believe if you can’t manage risks and threats coming from unexpected circumstances then your tactics aren’t flexible enough.

    • Bull0 says:

      Yeah, nope. Why nope? Because most of the games I can think of with procedurally generated levels would be fucking diabolically bad if they had normal, designed levels. Think Minecraft but with one map instead of infinite maps? Doesn’t even bear thinking about. But you should have to. For saying that.

    • Baines says:

      If you can’t do level design, then your procedural level design will probably be weak itself.

  6. Shadowcat says:

    It does look gorgeous and fun, but I’m not buying the “it’s virtually done” line — why would anyone start selling something as an unfinished game if they could sell it as a finished game very shortly afterwards?

    Please do a Wot I Think once it’s actually finished, though?

  7. Several says:

    My enthusiasm for the aesthetic keeps getting jostled by a swelling suspicion that circle-strafing might be more of the gameplay than I’d like. Please, strangers, speak of your experiences!

    On a related note, has anyone touched “Fancy Skulls”? It might fit in with this burgeoning gang of stylish FPS’ers-with-procedural-bits that are for-purchase while still-in-dev. Aesthetically, I’d say it’s more “Duchamp versus old wall-paper” than HB’s “Enter the Void versus retro-3D,” with the pacing being more deliberate than frenetic. It seems to do a respectable job inviting less circle-strafing. There’s rather a slow rate of fire, tiny enemy weak-points, corners to hide behind, and some enemies that don’t really care where you are—they just go about amply decorating the room with doses of fatality. As such, it feels like there’s A) less need for constantly crabbing around and B) increased odds of side-stepping into permadeath. Armament’s a little underwhelming, though.