so it turns out that Dungeon Crawl Stone Soup is extremely good. my previous experience with roguelikes has been POWDER (I was briefly kind of obsessed with it lol) and Angband (my first impression was “oh wow they took all the fun stuff out of POWDER”). but my first impression of DCSS is “oh wow this is like POWDER except ridiculously user-friendly and with all kinds of QoL features and what might be the best tutorial I’ve ever seen in a game”. it’s even open source

my new criteria for whether something counts as a roguelike is if it has a scroll of identify, which itself needs to be identified

(this is a joke but I’m ngl I think identifying items is an important part of what makes a roguelike a roguelike. in fact I think it’s actually more important than the game being grid-based or turn-based. like Binding of Isaac feels like it’s right on the borderline of being a roguelike to me)

@kasdeya There are remarkably divergent opinions on what makes a roguelike for a genre that has its criterion of inclusion right in the title.

(though tbh, that's probably part of the problem. Even ignoring the fact that most players of modern roguelikes probably aren't aware Rogue is an actual video game and thus have no frame of reference, like Rogue in what ways? It really is an open question.)

Video game genres are fun cus they tend to be named rather straightforwardly after either gameplay features, place of origin or in relationship to canonical games and you'd think that would mean sorting games into categories is easy and reasonably objective, but in practice you just sort of run into every single edge case scenario at once.

@kasdeya I have said on SEVERAL occasions now that when people say a game is a "roguelike" they actually mean it plays like Binding of Isaac and not that it plays like Rogue.

To me, the borderline of being a "Rogue (1980)-like" is Spelunky.

@nycki I’m ngl I don’t think people even mean that it plays like Binding of Isaac anymore. I think the word “roguelike” has been flanderized to the point that it means “procedurally-generated levels and permadeath” and everything else has been lost. like what the fuck does Hades have in common with even Binding of Isaac other than that there’s (kinda?) permadeath, and a little bit of random level generation?

@kasdeya I haven't played Hades but doesn't it have the whole "random powerups that combine in funny ways" thing? I think that's part of an isaaclike, right?

@kasdeya "roguelike" just means "arcade" I guess

@nycki it has random powerups but they only combine in ways that you’d see in a really bland RPG - like you can increase your crit chance and also increase your crit damage which is technically a synergy but like c’mon lol. they are random though which is kinda Isaac-like

@kasdeya gonna start referring to Mario Party 3 as "my favorite roguelike"

@kasdeya can someone please make a roguelike hack of Mario Party 3 like they did for Super Mario 64 -> BAZR, except that instead of randomizing the whole game it just lets you skip to any Story Mode mission like how you can play the decks in Balatro in any order

@nycki @kasdeya it has random room selection and random powerups but the level of randomness is "okay i might have to adapt my usual strategy slightly because i didn't get the optimal boon" rather than "oh this just made that strategy impossible, time for the backup one"

basically it gives a lot of ways to bias the randomness in the direction you want, particularly as you unlock more of the story-gated items

@kasdeya we think the main criteria for us is "the randomness is able to put you in unwinnable situations, and this happens pretty regularly"

if the player has too much influence on the randomness, or the severity of random effects is small enough that it can basically always be adapted to in a run, that's not really a roguelike to us, that's just keeping the game interesting

like roguelikes are a genre where we sit down to play and expect to do ten runs and maybe only one is viable to take further than the start

@tempest @kasdeya this is the same criteria rowan uses. it’s particularly why balatro is a favorite despite a deckbuilding roguelite about gambling. it has been pretty overwhelmingly proven that the vast majority of runs are winnable and there have even been some tests done to find unwinnable seeds. the only known unwinnable seeds are actually known bugged ones that make the random generator goof