ROGUELIKES: "DIFFICULTY" VS "CHAOS"

Kate B's picture

I was trying to figure something out. why is it that I enjoy hard games, but I find so many games that are often called "hard" to not be satisfyingly hard, but instead to be tedious? why do I hate playing these games when in theory they should be fun and rewarding? why am I *good at* some of these games but don't want to actually put the effort in to beat them? Let's take a look.

1. Roguelike structure and what it means

there's two different types of roguelikes and roguelites. the first is one where you can take something permanent from the game loop (like Hades, with its upgrades) and the kind where when you die, you keep nothing whatsoever and lose it all.

I'm not against the second type. But I think it lacks one thing, and I'll try to describe it: i.e, when I lose everything, I lose everything, but my skill and mastery of the game remains in my brain. In the best kind of roguelikes, this means something. it means that on my next attempt, I'm better. I can get things faster, I can beat enemies quicker, I know where to go, etc.

...however, the problem is, most of these roguelikes DON'T work like that.

...If I die, I'm back at the start, with my pea shooter and low health, and I'm doing it again. What I want, now, is a faster route.

(Enter the Gungeon: I spent so long in this first area every single time I died that I got sick of it)

I want a shortcut so I can trade off my higher skill level for getting my upgrades back faster. I want to have a direction I can go where the enemies are stronger, which a new player would struggle with, but a seasoned player can get through with no trouble.

this fixes an issue I have with a lot of these games: I can be good at these games- I can be *really* good at these games- but no matter how good I am, I have to go back to the start, and spend 15 minutes with a pea shooter, having no fun at all, daydreaming about the build I *could* have later, if I don't die.

2. Not dying

There is a problem in a lot of these roguelikes, which is that at the start you have like, maybe 3 or 4 hit points but they're so hard to get back that if you die, you might as well restart.

This puts the entire game's risk into the very first enemy you encounter. if it hits you, you're worse off than you'd be if you restarted, as now you've been given an additonal quest to go find some more health. Get hit, and the game is over. also, you have your very first weapon and that's it. This encounter will be basically identical every time you play the game, without fail.

(spelunky: you get 4 hit points from the start in spelunky. they're so hard to get back that if that first enemy hits, you're now on a quest to find mre health, in addition to the quest you're already doing. you'd be a fool not to just restart!) (nuclear throne: health pickups are a common item drop from enemies, and so losing a bit of health on level 1 usually isn't a problem, and it's not as much of a risk to keep playing.)

3. Builds

once you get a while into the game, a "build" appears. You're stuffed up with upgrades, the numbers are going up, there's "juice", it feels different. It's harder to die. There are way more choices to make moment to moment. You're in the "real" game.

if this part of the game takes a while to get to, I think the game looks like this:

(I call this the "inverted pyramid")

the "inverted pyramid" structure: the start of the game has fewer branches, and the end of the game has more branches. you can make fewer choices at the start, and more at the end! In fact, the game usually ends right at the point that opens up the most!

I propose a different structure to play with:

(I call this the "pyramid")

what if the early game had the most choices, but the later game starts bottlenecking you into fewer, really putting that build to the test and deciding if you live or die?

I thought about it and realised the second structure is more akin to how most puzzle games work. Roguelikes aren't puzzle games, though.

A more likely structure for a fun roguelike could be this:

(the "trellis")

4. Chaos

The problem with the inverted pyramid structure is something I'll try and illustrate here:

(high chaos at the start of the game. You can live or die based on one decision. Low chaos at the end of the game. the difficulty levels out and it's easier to stay alive. the entire game happens at the start.)

I'm going to name this effect "chaos", and mark it as something different to "difficulty". chaos is when a game becomes a bit quantum and unstable.

alternate definitions of chaos: "how easy it is to lose", "how tight the tightrope is", "how quickly it can go wrong". I define it as similar to quantum instability, or schrodinger's cat. when you are in chaos, you are in a cloud where the game could be won or lost at any second.

dying and restarting a lot at the start of a game is something you could call "reshuffling the deck over and over until the game is fun"

This is chaos. it's not challenging, it's just chaos.

another way of doing a roguelike could be this:

(low chaos at the start of the game. You don't die as easily, but as the builds get juicier as the run goes on, the tradeoffs become more drastic, and the game becomes more hard, really putting you to the test.)

achieving this is difficult, mind, because, out of all the examples I went over in my head, a run that is "high chaos" at the end and "low chaos" at the start could easily once again turn into a case of the early game being unfun and the later game being a lot more fun.

IDEA: what if you started with a lot of max hp, which was easier to get back, but your max hp actually *went down* as the game progressed, instead of went up?

At the moment, a lot of roguelikes feel like this:

(high chaos at the start, fewer choices)

I think it could feel good if it felt like this:

(fewer choices at the start, but low chaos.)

or even THIS:

(lower chaos at the start, fewer choices)

these structures trade away the major problem of the start of the game being *both* a low-choice, AND a high-difficulty challenge, by swapping only one of these things around. if you swap both, it becomes un-fun again.

Conclusion

sometimes people will say "git gud" about games that are simply no fun to "git gud" at. games that waste time every time you die but require dying to get the knowledge to progress. Games that put you through chaos instead of difficulty and then turn into an easy game as soon as you figure it out. I think that a roguelike can reset entirely and lose all your progress and still be fun if it aknowledges the fact that at the start of the game, it may be addressing an experienced player, and not a newbie. I think it could make the games more fun and more juicy AND more difficult without spending my entire run in the chaos.

Comments

spinnylights's picture

meowmeowmeow

I'm glad you wrote this, it was an enjoyable read ^^ I think a lot about roguelike design but a lot of what I've read about it feels to me like it doesn't try to really get at the meat of it or something, or is more about just like playing them or w/e as opposed to making new ones…so this was fun to read for me :P

I'm kind of curious—have you played /Brogue/ (https://github.com/tmewett/BrogueCE)? I wonder what you might think of it in the context of this essay. I think maybe it has interesting ways of addressing the design concerns you raise here, but I'm not sure if you would actually agree with that or not.

* It's of the traditional "permadeath" variety where you don't carry anything over between plays, but I actually like that about it—I feel like it makes the game more "arcadey" and keeps it from having any kind of grind. It has enough variety in its item drops and floor layouts and so on that every playthrough is very distinctive, also—it's never really the same game twice.
* It gives you extra kit on the first four floors, and kind of obligates you to have some sort of a build ready by then with its enemy pool, so you typically have the meat of a build going basically at the start. You have to get kind of creative though; there's enough gear available at first to make /some/ sort of viable build, but you have to know how to work with what's available. The game has special "treasure rooms" that offer you a choice of random kit but let you only take one item, and there's usually 1–3 or so somewhere on floors 1–4, so you make important build decisions at the start with those.
* There's no XP; instead you periodically find "enchant scrolls" which you can use to power up any non-consumable piece of gear. This gives you very wide latitude in your build decisions. Some very powerful builds come from enchanting pieces of gear that might not be obvious (like a teleport charm).
* The first two floors are easy enough that you can mostly do them using the game's autopilot, but it's basically extremely hazardous the entire time once you clear floor 2. :P Even floor 2 sometimes presents challenging scenarios involving eels, depending on its layout. A big part of the game in strategic terms is understanding how to scale and add onto your build on each floor to meet the new threats coming on the next floor—that arguably starts on floor 3 and never lets up after that. I wouldn't necessarily say that the late game is easier or harder than the early game per se; they just involve different challenges. (The 14 optional floors below floor 26 are kind of their own thing in this regard though.)
* You continue adding onto your build for the whole game; by around the midgame you need to have enough stuff together to handle a wide variety of different threats, but you basically never have the ideal set of kit for whatever you're doing, so you keep looking out for helpful stuff the whole game. If you've put a lot of enchants into something and find a piece of gear that you'd rather replace it with later in the game, you often find a "commutation altar" in the lower depths that lets you swap the enchantment value on two items, so you can make dramatic changes to your build late in the game that way if you want to.
* Even though it's definitely possible to die in the early game even for an experienced player, I've never had a playthrough I felt like was just unwinnable. When I was more of a beginner, I had deaths that seemed totally random and unfair, but I never really feel that way at this point—I can always point to something I could've done differently.

I'm not sure how much you would consider it difficult vs. chaotic in your terminology. As a general rule, it can go south on any turn and is very unpredictable, and often the scenarios that arise feel very chaotic and wild, but I also feel like you have a wide enough palette of options that you can deal with anything that arises if you recognize what's happening early enough and know what sorts of actions are likely to put you in danger. You can kind of plan for the chaos—if I'm in a scenario where I feel like the chances of me surviving to the next turn are 50/50, I know I've already made several mistakes. I'm not entirely sure what you would make of its gameplay though; maybe it's kind of "low chaos at start, high chaos at the middle to end" in your scheme, but in other ways I'd say it's maybe kind of high chaos at all times past floors 1–2. It kind of depends on what aspects of chaos you emphasize I guess.

Anyway, if I (or more likely Lily and I) ever make a roguelike, I'll think of the points you raised here ^^

P.S. I also think about how like, "hard" can mean a lot of things when applied to a game—sometimes it means the game requires deep knowledge of its mechanics or finely-honed reactions or that sort of thing, sometimes it means that the game requires a huge amount of grinding, sometimes it means that you have to be really lucky to do well, etc. etc. I think like, there can be some overlap between these, like how with a game that's really grindy often you can apply deep knowledge of the game to get through the grind really fast or that sort of thing. Even so though, I've noticed that different people consider really different sorts of games to be "hard," so that like, one person's hard game can easily be another person's tedious game.