Please, deconstruct me the Rogue-Like.

[jsnlxndrlv]

Because of this video, I only consider two design patterns necessary to produce Roguelike effects:

1. One or more aspects of the space which the player traverses is in some way randomized or dynamically generated, so that subsequent attempts will encounter features of the landscape in a different order, or not at all.

2. The player’s avatar accrues some abilities or statistical benefits to the exclusion of others at the start of play and/or as play progresses, some (or all) of which are lost upon failure or success.

While I’m happy to consider games like Spelunky and Into The Breach a part of this genre, I did play a lot of ADOM in school. I can’t call it a maximalist roguelike anymore because Dwarf Fortress exists, but it’s the first thing I thought of while reading that post by @ferrets. I don’t know how it holds up now that it has graphics and got ported to Steam.

[shrug]

you can attempt to drink a goblin, or throw a health potion (even though you wouldn’t want to

those are my preferred actions to take with both of those items irl just fyi btw

[mauve]

roguelike as a term has ceased to have much meaning but:

• using randomized content to force you to work the game’s rules systematically rather than memorizing specific patterns
• character growth focused on acquisition and consumption of resources rather than pure stats alone
• death causing a restart, emphasizing that the entire game is to be done in one go

many modern rogue"lite" games, such as Rogue Legacy, are designed to be beaten on multiple playthroughs, where you progressively unlock things. these are not to be confused with proper roguelikes.

necrodancer is probably the best modernized example of a roguelike that maintains a lot of what makes the genre compelling without sacrificing it in the name of more traditional rpg unlocking loops. it is not just designed to be played in a single run, it is heavily tuned towards items emphasizing speed vs safety vs “well, this dropped and i hate it but i’m going to use it because i’d be stupid not to,” making it ideal for as a game for multiple playstyles.

a poorly made roguelike often fails to give you strong enough tools, or just boils down to racing stat growths vs the enemies.

[Ymer]

Spelunky was the game that kind of started the roguelite trend, right? Or was Binding of Isaac that really took it off?

[mauve]

i want to make a roguelike.

actually when puzzle game is done, that’s kind of the goal of what i want to do with that engine.

many of the hooks are already there…

[the_kyphosis_gamer]

arcade environmental strategy

[Broco]

It was Spelunky yeah. But Spelunky’s design philosophy (particularly in HD remake) isn’t very representative of most of its successors at all. It’s twitchy, erratically paced between super slow and super fast, deliberately left somewhat glitchy and poorly balanced (Derek Yu has stated that he feels leaving out some polish gives games more character and depth), and focused above all on discovering the path to a secret area by making observations across multiple playthroughs.

It stands alone as a distinctive thing not much like either traditional roguelikes or modern roguelites. (I might say that Spelunky is the Demon’s Souls of its genre.). The only other roguelite that seems particularly influenced by Spelunky itself is Nuclear Throne

[la_ciel]

Just dropping in to remind everybody to play caves of qud asap

perfect maximalist example of the old school roguelike

[Broco]

Also play Brogue, the perfect minimalist example of the old school roguelike

[mauve]

brogue is a roguelike without anything fun left

[Felix]

yeah I never liked brogue, qud is probably the only non-stone soup proper roguelike to really grab me in like a decade

it is visually nice though

[ferrets]

I disagree. I think Brogue is a really engaging tactical game that adheres to a specific style of Roguelike. I think it’s meant to be aimed towards a different, but not lesser, taste for the genre.

Also COQ is dope

[Broco]

Brogue’s fun is first a slow-burning accumulation of small satisfactions as you improve your efficiency. Then it reaches a climactic stage of managing high-stakes, resource-draining late-game encounters and finely assessing risk to maximize score (exactly how deep can you dive with this build, or are you so weak you’ll settle for a vanilla ascension?).

The two distinct skills interplay with each other perfectly, because your exact efficiency over the early game informs the call you’ll make in the late game. The beautiful lighting system adds flavor, making the choice to dive further into the dark feel ominous and the ascent back into the light a gradual relief.

[Broco]

XCom is an interesting case: pretty roguelike in most ways, but also perfectly reasonable/expected to play through it exactly once and with a bit of save scumming

[r-i]

oh.

I missed this thread because I was having laptop troubles all day.

(and was kinda avoiding sb-proper for a few days after a bit of an emotional spat regarding roguelikes, actually.)

regret-index can’t help but ramble. if nothing else, nobody really brought up *****, so she shouldn’t be too bitter.

• the only defining matter of the genre permadeath + procgen + adaptable build decisions, because that provides the defining dynamic of scrambling up one’s options and opposition so as to emphasize learning both. most forms of permadeath subversion are dilutionary means of permitting grinding or a sharp lack of confidence in newcomers being able to learn systems / drops. “roguelite” labels as not being grid / turn-based means compressing genre-hybrids like Downwell, FTL, and Nuclear Throne together into one label, which is completely useless as a reference descriptor. while amongst the knowledgeable there’s some use in ‘roguelites’ being the non-permadeath games and dividing that up into genre-hybrid labels (and even more horrifyingly, identifying games as having roguelike and roguelite modes), the confusion there with people applying it to other games doesn’t feel any useful for regular discussion.

• procgen can’t be just randomization because blind floorwork and (even modular) enemy dumping makes all levels essentially become same when not looking at them from above- there’s no decision in moving left or right in a pure maze, no real meaning to what side of a box cave one is on. sectional contrast, in generation refinement or in fitting in variable scaling handmade pieces, are wellworn and easy (if constantly neglected) means to give floors senses of identity and to break up player decisions some more with a wider true possibility space. optional branches and floor themes are also useful in the same fashion, of course, but only with such above knowledge. hearing Tangledeep failed this matter was greatly disappointing, personally, but that game just doesn’t want to even be a roguelike.

• identity minigames and X clocks are decades-old unexplored-dead-ends of design beyond making the earlygame livelier and ideally providing some vague space of gambling assessment. there’s some good space in set-based content existences- Hero Trap is one of my favourite 7DRLs, I guess.

• lifespan and learning room are both rarely assessed matters despite their relevance and misunderstandings. I’ve got bad experiences with contrarian experts, after all. how well one is taught to avoid mistakes and optimize builds are very rarely coherent matters without a community discussing tips back and forth- as much as they can feel like randomly hopeless attrition, it’s worth noting a very few people technically do streak Brogue and FTL. modern roguelikes usually benefit from restrained sufficient scopes and runs rather than being simulators if they’re not at the peak of that format like Qud, to be honest.

• Brogue was amazing and respectable for its time (seriously, in total: conscious and active aesthetics, mechanical-tutorial vaults, combat-chances-summations, simplified / mechanical extremes, dropping xp) but is sharply overtuned against consistent success. I do rather like the fear the final floors always have, at least. Necrodancer’s defining premise and item roster is great for speed versus safety being a constant conscious matter at all levels, and I’d also argue that its enemy roster + level design is top notch in supporting that with enemy swarms, but hitting the skill wall + drops wall in leaderboard-climbing for high-level play plus a somewhat weak final zone in the dlc has weaned me off continuing to play it. Qud uses what classical maximalist roguelikes should have been like to build a sense of an alien world far more alive than most any other takes, is what I’ve gathered, and the focus on writing is extremely refreshing in a medium that falters at that.

I really need to get around to streaming random roguelike play some time, and also figure out how to make this something personally important less of a means of dissolution. I might be turning things around soon…

[Broco]

While “roguelite” is a lot looser than most genre labels, doesn’t that make it more useful rather than less? If not as a discussion framework, then as a brilliant branding gimmick and jumping off point for popularizing indie innovation. For players who’ve enjoyed one “roguelite”, they can then explore the rest of the non-genre, which promises vaguely the same kind of experience but without the boring, cookie cutter mechanical replication that playing only one genre typically means.

[Tulpa]

Brogue and Shiren DS are the only two roguelikes that I have played and liked

[ferrets]

Have you played the Vita update to Shiren? It adds a lot of cool stuff to the base gameplay of the DS version

[Tulpa]

I haven’t! I should take a look at it.

[Broco]

Try Necrodancer too. Similar-quality artwork, charm and complexity level, and hews more closely to the traditional grid and inventory format than any other roguelite.