Brogue and Shiren are good games
I don’t think people who play roguelikes are fools, I do think that playing a roguelike is on the same level as the way I compulsively play picross. They’re compulsion games and the way the design trends of roguelikes have been reified in modern games has been an overwhelming net negative.
I’ll agree here. The things that modern games learned from roguelikes are “random good” and “stupid hard good” without any of the good bits.
Actually I view roguelikes in sort of the same light you view Sekiro: the goal is to get good enough that the mooks on the first 5 floors pose no threat and can be steamrolled. It’s satisfying getting very good at the first bits because it narrows the game down to a series of interesting decisions (and luck) that could have a major impact later.
Anyway I’m not a roguelike expert but I bristle at these sweeping statements. I understand your position a bit better now
As a dilletante of many genres I find roguelikes have the healthiest output of learning for the first few dozen hours out of any of them. And that is to say, I think they have some of the healthiest return on play out of any genre for a standard player experience (a straight clear).
If we’re talking about the edge of roguelikes played to mastery – played for months and dozens to hundreds of hours – I think they flirt with but don’t fall into what I’d consider grindy patterns. Instead, I think they’re comparable to multiplayer and competitive aesthetics.
Players repeat similar actions, in similar contexts, dozens of times, hoping for a combination of luck and incremental advances in their skill to reach new plateaus of ability.
As this is not a direct time->progress output, I don’t think it qualifies as a bare progression loop, nor do I think it’s a good fit under our softer definition of ‘grindy’ (I’ve been defining it as: game systems gating content behind repetitive play with a low skill component).
Now, like with multiplayer games, players can easily be hooked on a roguelike long past the point of enjoyment. I think this is a bigger discussion about what we normally consider healthier game components (skill and execution) turning unhealthy at high contexts. And we need to be careful to separate it from the healthy competitive side that many people enjoy, in becoming true subject matter experts, in organized competition, and the like.
(to the RTS point: I don’t think this is grindy, and they didn’t use progression loops, but they were intensely compelling in the way many multiplayer games are)
The other assertion is that pre-PSX RPGs cannot be called grindy. Well, they definitely defined progression loops. Did they misuse them? I think a case can be made that the better examples didn’t; that the slow pace is more a result of a slow content pacing and repetitive, simplistic battle systems are a side effect. But the 8-bit JRPGs are usually egregious in this; lacking strong tactical choices in battle, and a good clip of narrative, the vast majority of them consist of dozens of hours of samey battles with the promise of slightly different tilesets. Grind was significantly lower in the 16-bit and early 3d eras as the production values of Dragon Quest and Final Fantasy shaped the genre.
I like the doom roguelike, that’s a fun time.
My favorite other than Shiren is probably Powder, not for its complexity but for its ability to be played with a dpad and 6 buttons, i.e. on a GBA. Very traditional roguelike otherwise
can you be specific here? Can you clarify why leveling up and collecting loot is bad-wrong in a jrpg but good in a roguelike?
I actually consider this far worse than JRPG-ish time -> progress because this is literally relying on gambling mechanics to exploit player’s time. It’s Skinnerian ‘semi-random rewards for consistent inputs’ design. One of the most insidious compulsion loops in games.
I don’t think the random reward schedule of loot is all that important to a run’s success up to a certain high skill goal. The more important consequence of randomized player inventory is forcing the player to consider their loadout in new and different contexts; to unpack assumptions from other games and try sub-optimal possibilities.
It’s great! I naturally love working under pressure so this is just about my favorite thing.
Now, it’s hardly fair to build your critique of random reward schedules and hold up Dragon Quest, the mainstream game with such a deep and true love of gambling.
I love working under pressure too, but I can think of two roguelikes (which I named) where I have ever been required to work under sub-optimal conditions to try to succeed. Most of the genre is trapped in this awful awful place of one-true-strategy-ism and wiki-scumming. That is to say, they are pseudo random in a way that they’re either trivial to solve or require an exacting script to follow. Both of those options preclude interesting sub-optimal decisions. When skillful play is isomorphic with scripted play, how can it be interesting?
This is good context; I deliberately avoid wiki play or reading strategies from people less cool than SB. I want to figure it out myself, that’s the entire reason I play.
If a game boils down to this at any but the highest levels of play, the game has a real problem. At advanced skill, it’s extremely difficult for any game to avoid compressing down; usually the designers don’t even play it at that level! In this regard I forgive most games the way I can find most games extremely compelling without standing up to a lifetime of study.
Interesting to think about the strategy trading as part of the cultural environment these games foster; it’s really great in a friends context, a playground context, and to some extent the streamer context (I think the longterm Spelunky and Binding of Isaac streamers would say it’s more interesting than playing multiplayer games over a long time). From a Word-of-God context it’s as poisonous as any other.
related: I refuse to read about Magic: The Gathering meta before it’s two years out of date
when I have 20 options and 19 of them are traps that only idiots would choose, I consider it to be bad design. This was infamously the case with character building in 3rd edition D&D. Insofar as system mastery comes from trial and error figuring out which choices are universally bad until you are left with the one good choice, I am not interested. What I want from certain games (let’s call them dynamic puzzlebox games): every strategy is viable if you can figure out how to apply it, there’s no optimal solution and there are no ‘traps’ to make you feel stupid and people who have ‘mastered’ the systems feel smart. What I get in most roguelikes, randomly generated king’s quest puzzles as game design.
Yeah, I totally agree with that. I think we’ve got different context and I haven’t played a broad enough swathe of roguelikes at a high enough skill level to feel like that’s a genre trait. And indeed, most of my thinking about the Form of roguelike is drawn from Shiren, Brogue, and Spelunky, colored by the mutations of the modern action interpretations and their various successes and failures.
damn, that was a tear
I love this clubhouse
dragon quest games are good because they are full of cute monsters and animals to look at
roguelikes are bad because they typically don’t really have anything to look at in them
As the guy who has played MK11 more or less daily since release and spends all his time tapping X so AI Noob Saibot can clear towers to earn currency so I can reseal all the chests in the Krypt and open them up all over again in hopes of getting a color variant of a dagger or something I’ve already got.
That’s not really the best way to play that game.
Also it turns out I’m awful at it when towers require you to actually play.
the only allure more powerful than The Grind is Dressing Up All My Ninja Boys
Roguelikes are dear to my heart, I’ve never beaten any of them or felt the need to, the suboptimal weird choices and finicky bits are the whole reason they’re good, and you guys writing these huge nerdy-ass essays about how playing 10000 hours of games made you too smart to enjoy them are driving me fucking nuts. But I can’t ignore any of you. Normally I like your posts
I play the subset of roguelikes that abstract all that into pure ASCII. 
Most of the older rogue-likes are stuffed with cruft, such that trying to learn them on your own without having a spoiler handy is adding years to a successful win and running into many gotchas. There was a person on Usenet that took 4 years to win a Nethack game just like that. These were games that were intended to be played on mainframes and discussing them fanatically with other players.
Angband as played by experienced players was not intended to be a grindfest, slowly accumulating resources as safe as possible before proceeding to the next floor, of which there are 100. The intent was to determine how deep one could dive without certain resistances due to threats that could now spawn and skipping past the boring stuff in-between. Still has the wide problem space that necessitates knowing combinations but again, came from that era of intended shared community knowledge.
Sil instead condenses what Angband was as a Tolkien inspired roguelike into a much more tightly woven set of mechanics, 20 levels deep and an aim to allow players to gain experience without having to kill everything. This would mean learning the game from scratch is absolutely possible within a typical timeframe. It is much shorter than the other two I touched on and there’s an anti-grind mechanism in place that pushes you closer to the confrontation with Morgoth the longer you stay hovering outside of that depth.
Play Dungeonmans if you want to avoid traps and mastery and length. Similarly Tangledeep.
Play Towerclimb. MORE people play Towerclimb!
See I got this all sorted out, I just don’t play any games with roguelike elements and hence I don’t have to worry about any of this stuff.
what I’m really trying to express is that I’m too stupid to enjoy roguelikes and I don’t get a kick out of failing 10000 times because I didn’t read the developer’s mind. The RLs I do enjoy are the ones where I can be an idiot and still get by
The point is that neither of us are stupid because we enjoy different things about roguelikes. We can be stupid for other reasons.
Races in DCSS are essentially difficulty settings (they’re labeled “Simple”, “Intermediate”, and “Advanced” in the menu); most good players don’t play MiBe over and over because they’re fairly easy once you understand the game and there are more interesting and more challenging races to play once you’ve mastered the basics. (Also I don’t think they’re nearly as trivial as you make them out to be; I just queried my games and I’ve only won 50% of them online, and I’ve certainly killed many more offline when I was learning the game.)
The tournament I’m in currently forces each player to play a randomly selected race & background combination each week. You only get one shot each week (well technically two if you die before XL 5). You get points for completing certain milestones, worshiping certain gods, and each week has unique challenges that get you additional bonus points. I think it’s a fun way to play DCSS that forces you to learn a bunch of different strategies and playstyles, and play in a careful and conservative manner since you only have one try.
I do think Brogue is much better in terms of mechanical transparency and simplicity, but DCSS does try to surface a lot of information in the game and has been getting simpler over the years (a main complaint from older players that have fallen off is that the devs remove and simplify too much). The design goals explicitly call for skill based gameplay, meaningful decisions, avoidance of grinding, and playability without need for spoilers. You can certainly argue that they don’t achieve all those goals perfectly, but it’s closer than a lot of other games I play. I don’t think any of the devs would say it’s a perfect game, but they’ve fixed a lot of the low hanging fruit and most of what’s left are hard problems without obvious solutions.
You mentioned the ‘o’ key, which I think is a good example of a hard problem. If a character is doing particularly well some parts of the game could be boring, so the ‘o’ key is an attempt to automate exploration to reduce the number of trivial decisions the player has to make. Ideally the ‘o’ key reduces the time between interesting situations, so that the player has to make meaningful choices more frequently.
Certainly there could be other solutions to this problem: they could make floors smaller to reduce the tedium, but then that affects the player’s ability to run away or avoid threats. They’d also need to increase the experience every monster gives, re-balance how many enemies spawn per floor, potentially resize all the handmade vaults, maybe change spell ranges, how far sound travels (you wouldn’t want to wake every monster onthe whole floor since it’s smaller) and it’s not entirely clear that a better game would come out the other end. A lot of mechanics that currently result in interesting situations would have to change and be re-balanced to support that.
Maybe it’s a shitty stop-gap solution, but I don’t think it’s designed to be compulsive. I think the same game without the ‘o’ key is certainly a worse game because you’re spending more time in the boring parts.
As for randomization: I think it’s largely to force the player into new situations they haven’t been in before. Good players can win many games in a row, so I don’t really think it can be a gambling compulsion at a high level. You could argue worse players might get some gambling thrill out of it, but since ultimately player skill has a much larger effect on winning than getting good gear, their time would be better spent trying to learn the game, rather than mindlessly mashing and trying to get lucky.