A CHASM of Mediocrity - Procedurally Generated Metroidvanias

[@ingeniousclown]

Two things: First: The game at 1:01 is A Robot Named Fight. It's a "Roguevania". Second thing: I have a video about "Roguevanias", comparing Dead Cells and A Robot Named Fight on this channel. Check it out! ► kzbin.info/www/bejne/gqHGloltfKlpnMk

[@Sparkbomber]

Frankly, I liked Chasm enough to finish it. Its definitely nowhere near as good as true Metroidvania's though and I agree with you on the points you brought to the fore. But there are some things I think you may not have addressed in full. Frankly, the NPC's in Chasm brought CV: Order of Ecclesia to mind. They're... okay in Chasm, but there's so much more that could have been done with them. The procedural generation hides a lot of them in less than optimal places. Second: the controls. I know you play the PS4 version (for all I can see), but the PC Steam version has controls that are not quite as tight as they should be. So you might end up jumping a wee bit short or not quite hitting an enemy when you need it. And getting the weapons you need? Random drops... Another gripe of mine is the bestiary. The sheer amount of randomly chosen numbers of a particular monster you need to slay before getting their basic entry, nvm what they drop, is simply not worth it. No description either and I do miss that. Even if its just a 2 line snippet, I enjoy reading those in say CV:AoS and the lot. What frustrates me, and you as well apparently, is that this game could have been so much more with a bit more care for detail from the devs. Some areas do look great, like the Gardens, but yeah... enemies rarely fit any area.

[@ingeniousclown]

There's a lot I didn't address about CHASM, simply because I didn't find it very important to the true main idea I was trying to present in this video. Plus I think it would have made the video too long for a game that probably doesn't really warrant the length. I played the PC version btw. I just use a PS4 controller for most games that would feel better with it (metroidvanias, platformers, Dark Souls).

[@maxque2841]

Man Im hating on you a lot but honestly it felt more like a super metroid review than the game you put in the title

[@UniversalSketch]

this could work better if the areas were not made to be set in certain zones but instead merge the themes a little to make subtle transitions.

[@MidoriMushrooms]

hey you said you like metroidvanias? You ever play treasure adventure world, and its precursor, a free game called treasure adventure game? I know both games have terrible names but they are basically my favorite games in the genre.

[@samcassidy2441]

You missed another key part of that red tower/ice beam sequence: as you leave the ice beam room, you have to use the ice beam on the same enemy type as those that patrol the red tower. This immediately brings the red tower back into your memory and the correct way to progress is revealed. It's genius.

[@RetroNutcase]

My main problem with Chasm is too many of the rooms are hallways with 1v1 fights one after another. I was hoping to see more variety where you'd have to deal with multiple threats at the same time.

[@lassielyra]

The combat system doesn't seem to lend itself to something like that. Just my perception from the video. It seems like it might be too unfair with several mobs having different attack times. Also long narrow horizontal corridors would be awful to play in if there was no way to manuveour well.

[@nemou4985]

Yeah seeing this gameplay this seems like the most bland part of it all. If there were some rooms with enemies shooting or spamming you... you can even adjust the enemy difficulty for more spammy mobs. But no, its all 1v1 with a monster AI which will never be your match.

[@PacoReer]

Typical medusa head rooms filled with other flying enemies... Good old times

[@aetherarcanist4819]

Don't forget the moving platforms and ruthless spikes greeting you every time you fall to the floor. And don't even get me started at the petrifying medusa heads...

[@Tokito935]

As expected from a sunbro

[@thingies4U]

Those are some damn-nice 16bit style graphics though.

[@tiagodarkpeasant]

tldr the beauty of metroidvania is that it is a linear game, in a non linear world, the progress is pre planned to be good

[@SpiralPegasus]

Perfect summary. 20 minutes of watching slimmed down to 15 seconds of reading.

[@mathunit1]

@@SpiralPegasus Minus the part about why Chasm is mediocre.

[@nowonmetube]

How are Metroidvanias are linar? Ffs!

[@ZanathKariashi]

because they're planned around a set path of upgrades to take. But have some flexibility that creative use of mechanics can allow you to sequence break. And that's before you get into UNINTENDED sequence breaks following glitches or really hard to exploit game features working in concert but in combination timing no one even envisioned at the time.

[@nowonmetube]

Metroidvanias are per definition not linear.

[@MisterMunkki]

To be fair, the castlevania "animation skip when you land" thing is very common in all metroidvania, and something I actually expect them to do. And for good reason, gameplay tends to get very stiff when you can't do that

[@VectorBatesta]

(3 years later...) That's why Hollow Knight makes it's appearance that passthrough the stiffness by making it possible to jump anytime and swing anytime, making the movement to be very dynamic and fun to play.

[@ComradeAlpharius]

"You find an item that lets Chasm crawl" Gold.

[@christianschweda2530]

Had to scroll waaay too far to find this.

[@rlpn6710]

Don't forget that the Morph ball exists.

[@Greywander87]

It's all about how the procedural generation is implemented. You _need_ to be creative about how it's going to work in order to get something that feels cohesive and is fun to play. Mark Brown has a good video analyzing the methods used in Spelunky, which worked well for that game. Doing procedural generation right is hard as balls, much more work than just building the levels yourself, but if you can pull it off you can end up with some amazing games.

[@kice1102]

looks like being creative with code is a lot harder than it is with art. This game is a prime example of that!

[@ThePC007]

@CabinDoor: Why would it be impossible? I have to agree with FuzzzWuzzz here. If you study the aspects of a great (man-made) level and manage to create a machine that generates a level with those same characteristics, then you end up with a level generator that creates levels that are just as good as man-made ones. The problem lies entirely in the fact that doing something is much easier than explaining how to do something, even more so when you have to explain it to a dumb machine that takes everything you say literally. And resorting to namecalling when you have no definitive proof that what you are saying is right (or do you?) doesn't work in your favor either way.

[@VestedUTuber]

+CabinDoor Making a statement and then "supporting" it with an ad hominem attack... do you really want to be doing that?

[@majormalfunction0071]

ThePC007 It's NP-hard, I think. An analytical solution to a problem defined by human intuition is not codeable in the generally, supposing it is not possible to enumerate and assign values to game design philosophy for a potentially unknown / arbitrary game (ie. all mechanics are left undefined and are unknown for the purposes of level generation).

[@amateurtakeongaming4298]

Procedural generation in most games doesn't mean procedurally generated levels. It means procedural sorting of levels

[@fen4554]

Just started playing Hollow Knight. That glimpse at the full map. Holy. Crap.

[@Xetelian]

Same, I am scared to boot it up now.

[@RRRRRRRRR33]

Hollow Knight is greater than Super Metroid, for real. This is the best metroidvania game ever made, don't be scared, it's really rewarding. And if you think how mediocre Samus Returns is.. compared to HK, a game made by 4 people.. it's just insane

[@isodoubIet]

Hollow knight is a good game but it does overstay its welcome by quite a bit.

[@RRRRRRRRR33]

The pacing of HK is not ideal for most people, it start slow.. But I think that slow start is very fair, you have to pay for everything in that game, even the save point markers, teleports, etc.. you have to buy that, lol It's not like in metroid that you have everything marked for free. When you complete the map in any metroid or castlevania, that feels mundane. In HK, that is more special because you earned the completion. And the game is not even that hard, some comparisons with the souls niche kinda make me scared at first because I hate the artificial difficulty. but HK is perfectly balanced, unlike all metroidvanias, you never becomes too powerful after the mid-game, the game is always balanced start to finish.

[@felipeconstantino9004]

R I recommend you play AM2R - metroid 2 remake. It is a fan-made game and, comparing with the 3D version Nintendo released, is much more enjoyable and have more bosses and weapons.

[@helbendt581]

Making a good procedurally generated Metroidvania is possible, but extremely difficult and effort-intensive - you'd need to design a TON of rooms, including rooms and versions of bosses that can be handled with entirely different combinations of power-ups, and enough memorable setpieces and bosses to prevent too much repetition (at LEAST five times as many as one run will ever contain). That's after setting up the generation engine to determine what the power-up sequence is for this iteration of the gameworld, and then which zones get those power-ups, and then how the zones interconnect. After all of this, the engine must also sort these numerous zones in such a fashion that the overall map is coherent, then arrange the rooms within the zones to make those coherent and interesting as well. All told, it'd be easier and more profitable to just design four or five whole game worlds and sell each of them as its own title.

[@_.-.]

"All told, it'd be easier and more profitable to just design four or five whole game worlds and sell each of them as its own title." This is the reason I get pissed off when players pretend a randomly generated map that's guaranteed to be fair and has a nice difficulty curve isn't a big deal. People reeeeeally don't know how much work goes into building an automaton that is able to make human-grade content unsupervised every single time.

[@Clawdragoons]

Procedural generation is effort-intensive in general. I've been working on a Roguelike game called SummonerRL for maybe around five years now, and as an example, I've made around 100 or so handcrafted bosses. For context, the average playthrough of the game can expect to run into an average of 6 or 7 bosses, assuming they choose to fight every single one. Every familiar you can summon has extra abilities you might not see each run, and even then you can only see one or two familiars in a run out of the current seven implemented. The floors each are generated according to unique rules which vary per set of floors, with additional variations on those rules unique to each set of floors. Basically what I'm saying is that procedural generation requires a ton of care to be put into basically every aspect of the game. I wouldn't say that a ProcGen Metroidvania is particularly more difficult than any other sort of ProcGen game, it's just that a lot of those Metroidvanias aren't putting in the effort required, and it's really apparent. To be frank, a lot of games are trying to ride this bandwagon without putting in the necessary effort. And then of course there's some various design elements that are necessary towards making ProcGen worth it. There's a reason that roguelikes and lites tend to be brutally difficult, and extremely mechanically deep - those elements tie extremely well in with the procedural generation and with each other. And of course, the worst ProcGen games will often drop those elements as well, resulting in a game which has randomness but no motivation to engage with it. I am obviously a huge fan of roguelikes and roguelites when they are done well. But too often I see this element of those games being used as a lazy crutch, and it drives me bonkers.

[@Kirainian_Gaming]

See, I love almost everything about Chasm. the fact that I know my map is not the same as someone else's map simply fascinates me. While playing, I seem to have been given a very good seed as most enemy types are spread out, and secrets are evenly dispersed, forcing me to backtrack to finally complete the map(Gosh that's so satisfying.) Not really having played many Castlevania games, I didn't see the copied elements except the spells, which are a total joy to use in Chasm, unlike all Castlevania games I've played. Which the ones I played made me choose which spell I liked most, and gave me a sense of paranoia about seeing another spell on the ground, then proceeding to dodge around it, often times getting hit by the everything else in the area. In terms of hidden connections, going back to my great seed I was given, the secret path from the Gardens to the Catacombs put me directly near a warp point, which was amazing, as the closest one was way too far away to be used effectively. All in all, the game is successful in my opinion. And that's really what makes these discussions fun, hearing someone else's point of view.

[@shelbyherring92]

Yeah, I find it fun too, my only gripe or two really is about the stats and character progression in terms of leveling. It does that thing Harmony of Dissonance does where enemies give less experience based on the difference between the player's level and theirs. And with the damage output being linear (which isn't bad, necessarily) it just makes the combat feel a little bloated at times and the rewards have diminishing returns of a sheer cliff. That said, it is still a fun game and it does have legit platforming moments where I'm having to take a moment and plan my routes to progress or get chests.

[@benjaminbeltran7004]

So, Chasm could have been a good metroidvania but instead is a machine that "creates" a different mediocre metroidvania every time you start a game.

with no female option...fuck it

[@GenesHand]

What kind of fucking complain is that? "SO FUCK METROID FOR DON'T LETING ME PLAY AS A MALE" that sounds right to you?

[@tigerfestivals5137]

@ who cares?

[@thetruemariofan6415]

*Cough cough* Rogue Legacy *cough cough*

[@tigerfestivals5137]

@ Why is that such a deal breaker for you? There are so many reasons to play a game besides the gender of the main character.

[@ThrillingDuck]

Castlevania LoS: Mirror of Fate was not a particularly good game, but the ability to take notes and stick them at any point on the map was a simple yet BRILLIANT addition that should have become a new standard for Metroidvanias. I give Hollow Knight a pass because it handled maps in a unique way in general, but there is no goddamn reason for ANY other Metroidvania to lack this note-sticking feature anymore. Metroidvanias are my favorite 2D genre bar none, but needing to remember the location of every instance of an inaccessible item or area with an obvious solution (i.e. I need to come back when I can crawl, when I can double jump, etc) is an archaic trope that should not be artificially forced on players anymore.

[@tomasparant8901]

UNepic had this feature too and it was cool there as well. I often left notes like "Find x item for ugly guy" or "spikes, LOTS OF THEM. bring y potion"

[@alvedonaren]

Hollow Knight added the ability place map markers in a patch (after you buy them from the map store). There's no text notes, but they come in four different colours so you can use different colours to mean different things.

[@anatolevilbois9210]

Hollow Knight added the ability to place notes in an update. But it wasn't there at the beginning, so I guess my experience with it would have been really different if I played it when it was released

[@EliosMoonElios]

Notes in map is not brilliant is very old school stuff in PC RPG's.

[@ThrillingDuck]

Higher Power I assume you've never watched a game analysis video in your life.

[@WaywardRobot]

Salt and Sanctuary is a good example of a sprawling map that's complex, and La-Mulana is a good example of a tight map that feels large because of its complexity.

[@SrEncon]

I found it weird that he didn't mention S&S since he had made a video about it a couple of years ago.

[@Raziel_Zero]

Salt and Sanctuary is so underappreciated. I find it a great combination of the metroidvania and souls-like genre.

[@alyjos9103]

Salt & Sanctuary is easily a top 5 of the Metroidvania genre in my opinion. I really wish they would implement a DLC

[@louisvictor3473]

This might be tangential, but there is one massive problem with proceduraly generated content that is often overlooked - your brain. It is a fantastic pattern recognition to a point it can recognize a pattern even when it is faint and quite honestly was a mere accident of random chance. And it is also a fantastic generalization and abstraction machine. For example, No man's sky. Basic complaint I heard was "it is different, but really, it is mostly just the same _samenesss_ ". And quite honestly, it is what I feel about Deadcells after the few first hours of play. The thing is, when it is used to generate content that your brain will automatically generalize and abstract, so that only the general pattern (rather than the technically specific details), you spent a whole bunch of time accomplishing absolutely nothing. Actually, less than nothing. With a fixed layout, you can get the player to appreciate the attention to detail in layouting the place and designing at least in some level of consciousness. The layout is entirely random? You lose that and on top of that the player is generalizing your content as "the sewer looking level" and the "the sewer looking level, but in yellow", which isn't a positive trade-off. L4D is still imo one of the best examples of procedural generation. It is reserved exactly to the sort of thing the brain wouldn't immediately generalize and abstract the entirety of it. What enemy comes up next, in which way, when, which weapons you find, etc. Those are things you can't afford to "abstract and generalize" when they're random. Scenery? Simple layout? "Different" species that act in just as many finitely many ways so it is "the same shit as the other cat, but this one is purple with spots and weird tail"? All this stuff that, other than the first couple of minutes/hours observing the visuals will have no real impact on how you play the game? Easy af to become oblivious to the details (aka irrelevant differences).

[@_.-.]

Fucking human brain picks up on patterns even when there aren't actually any. It really is a steep battle to fight against the "sameness". And yeah, L4D hit gold after implementing procedural generation in only invisible game logic. Most people don't even know.

[@OmarMejiasGamertologist]

I agree 100%. A great deal of what I like to call "video game magic" (which is what enables a game to become memorable) is when your ability to notice logical patterns is reduced to a point where your immersion is so boosted that you manage to organically feel what the developers wanted you to feel. For example, the first time I played Resident Evil (PS1) back in its day I was terrified. I had no notion of logical switches (getting an important item triggers something) or efficient enemy placing so everything made me feel as hopeless as Capcom wanted me to feel while playing. Sadly this "magic" is lost most of the time developers try to re-use resources and methods in order to extend the game. Sometimes its purpose makes it work like the Chalice Dungeons on Bloodborne, but most of the time basing the whole game on procedural design makes all the magic to be short lived.

[@cinnamonnoir2487]

I think it's quite telling that dungeon crawling for special loot items, a feature that was included in the original version of Minecraft, was _completely_ ignored by the fanbase in favor of the open exploration and crafting. Procedural generation can create a lot of content to play, but it can't make the result feel rewarding. For a Metroidvania this is an even bigger problem; randomness clashes horribly with an upgrade-based exploration system. You need structure to have a good game of this kind. By the way, people _hugely_ overestimate the amount of backtracking classic Metroidvanias had. I just replayed Metroid Prime, and it's surprising how neatly that game is designed to almost always permit you to reach the next upgrade quickly even if you've left the critical path. Super Metroid was also much more linear than people remember. Long sections of backtracking in both of these games generally means you're fooling around, not that the game is wasting your time. ...Also by the way, was that a SNES Classic controller you had around your neck at the beginning?

[@ingeniousclown]

Yeah it was. That's what I used to capture the Super Metroid footage.

[@TheMetalOverlord]

It's not only Super Metroid that is linear, ALL Metroid(except the first one on NES) are linear if played in the intended way. They become non-linear if peoiple play them with sequence-breaking tricks, when they exist, Directly non-linear metroidvania are extremely rare and the only example i can think of at the moment is Hollow Knight.

[@umbaupause]

Basically, the world itself isn't linear, but the progression is, I guess?

[@TheMetalOverlord]

Well, having a non linear world is a main feature of the metroidvania genre, so if a metroidvania has a lienar world, it's not a metroidvania. lol Anyway, yes, it's the progression that is linear, without sequence breaking you always go to point A to point B to upgrade 1 to point C to upgrade 2 and so on. But there are some metroidvanias that let you go point B to upgrade 2 to point A to point C to upgrade 1 ecc. Those are non linear metroidvanias, and Hollow Knight is one of the few of them.

[@edhc44]

If you know what you're doing in Super Metroid, if you know all your routes and every item location, you hardly are going to need to go to the same place twice. It is pretty amazing when it hits you, and what once was already a great game might become even better, it is a truly remarkable work by *TEAM DEER FORCE* .

[@TinkyTheCat]

Rather than spending all that time developing procedural generation so Chasm could have infinite crappy maps, what if they had hand-crafted more maps than people can generally remember accurately - say, a dozen - and picked a map at random every time you start a game? Might be the best of both worlds. Even if none of the maps are as superb as Super Metroid's, they'd surely be better than whatever is vomited out procedurally, and still offer far more replayability than 99% of gamers could ever ask for.

[@todesziege]

I've been saying this for a long time. Spend the time creating better level design tools, and then design a shitload of levels. Even a mediocre level designer can spit out stuff that will be better than what an algorithm will produce, and while it wouldn't be "endless", you could make so much that it might as well have been. Even 2 or 3 variations would be more than most players would see or remember.

[@phill8273]

That's kinda what Dead Cells does. Although it doesn't have your going back to previous areas. There are branching paths and you can choose which one you want to go to. If you make 10 level 1,2, 3,4,5,6,7, and 8s all each with the necessary things to finish the game you will have a higly replayable game. Just make sure the game memorizes what places you've been to so when you play it again you don't get unlucky and half of the game is already stuff you have done.

[@_.-.]

Yeah no, that's a shit design choice. You really aren't onto anything here

[@Darkmirror7]

It's pretty poor to just disregard any form of criticism by saying "oh where is your XYZ? You have none, your criticism is invalid" Do you use that "argument" for movie critics too? For people who try out foods and rate it? @Adolif4

[@Igneeka]

@Adolif4 You listen to a movie critic (for example) to have others opinion and compare it to yours, some just to know if this is something that might interess them oreven have your point of view challenged or debate in the comments (if it's on youtube), of course you shouldn't just blindly follow someone's opinion but that's not what critics are for Plus you're just disregarding his actual point : instead of deconstructing OP's idea and have an actual point you just pointed out that he never made a videogame which doesn't mean that his idea wasn't good and isn't relevant here

[@skudzer1985]

I just don't understand the allure of procedurally generated metroidvanias. In SotN, everything was always the same; I knew exactly where the library was, the coliseum, the catacombs, and it's still a wonderful game. I never wished it was randomized. The graphics and soundtrack were superb and I can play the game today and have a great time. Maybe I'm just boomin' here.

[@Deadsnake989]

Part of the appeal of metroidvania games is the meaningful sense of exploration. The idea of a procedural generated one is that you always feel that sense of exploration you got the first time you played. That's the theory behind it anyways, not sure it'll ever work in execution though. I think the main reason you get that meaningful sense of exploration from Metroidvania games is because of the hand crafted nature of the maps. The most recent game I played is Death's Gambit. It's not truly a Metroidvania but it shares some similarities, also some similarities with Dark Souls. Replaying it with different builds has been a lot of fun. but Every play through after the first felt like it was missing something, the sense of discovery.

[@RRRRRRRRR33]

The sense of exploration.. Thats kinda vague tbh, many games has this feeling in some capacity. The charm of metroid is related to smart designs and back tracking with the help of items unlocking abilities. Castlevania copied that in SOTN and added cheap rpg elements. So, imo, it's kinda insulting to cite castlevania and metroid in the same sentence as a unique genre, castlevania didn't created anything. SOTN is goddamn great, but it tarnished the legacy of metroid. Guacamelee for example, that is a metroid styled game, not a metroidvania. The rpg elements are so minimum, you have some powerups to make the character stronger, but that's it. The items you collect dictates the progress, not the rpg aspect. And that's the major mistake of SOTN and many rpgs: the character becomes too strong, there's no gameplay balance. At the end of guacamelee 2, I had every skill unlocked + 20k gold.. all that gold for nothing because there was nothing else to buy. Some can argue Metroid suffers from a similar problem, but it's not so severe to a point of breaking the game balance. How many SOTN players you know who completed the whole map and then they could defeat the final boss easily, no matter their skill? Now, Hollow Knight fixed these issues in a glorious way, this game is special, arguably better than super metroid and SOTN itself.

[@DriitzzCabal]

if symphony of the night level was randomized it wouldn't of been as good as it was , you'd have lost sense of progress the same way it affected chasm, Heck if you're stuck in chasm you can't even go to google or whatever to ask for help cause it's not the same seed.

[@handsomebrick]

I think it would work better if the randomness was more simple and targeted; for example, imagine if there was a specific door that's open in half the playthroughs and closed in the other half but always in the same place. So the player will still get that feeling of not knowing what's going to happen, but in a more controlled way that the level can be designed to incorporate.

[@foxymetroid]

Exploration is a big part of Metroidvanias. The appeal is that you find the bosses and power-ups instead of being led to them in a linear fashion. With procedurally generated Metroidvanias, you're given a new map to explore every time you start a new game. That would, in theory, keep the game fresh well past the first playthrough. In practice, Metroidvanias might be one of the harder genres to procedurally generate right. There are far more variables to take into account than, say, a simple sidescroller or a racing game. The inter-connected map makes things harder since everything has to fit right. You don't want too many of the power-ups to be completely skipable by even the most novice player, but you also don't want the player to get hopelessly stuck because they need a specific power-up to reach said power-up (ie needing high-jump boots just to reach the high-jump boots).

[@veggiet2009]

It seems like this game could do with a dependency tree, they probably have one but it seems kinda "off." A dependency tree that would label certain rooms and pathways a "dependent" on a power up, or skill. Once you have that create two separate layout generators, one that crafts a linear path, and one that creates an open space. These two systems would alternate, and randomly place items which would then inform the generators what can be done. Now create a 3rd generator which would essentially craft a linear path backwards which would use the dependency tree to add one way obstacles and then an exit at which point one of the two generators would takeover. Now I have no Idea if this would create a more interesting layout than what CHASM offers, it's just my speculation that this would create a system that would be playable (and unique, at least for the first 2 or 3 playthroughs) I would also incorporate secondary items which would give similar powers, but these would be hidden by the generators, and would not influence the generators. These items, if found, could get you through future obstacles, and in theory could give you a puzzle in which you have successfully collected item A early, which let you bypass a section, but in that section held items B and C which you'll need later on.

[@Battleguild]

Controlled Procedural Generation does flow better than complete Random Procedural Generation.

[@Strider_Shinryu]

This video is pretty on point. Chasm wasn't terrible but it was just underwhelming on pretty much every level. It didn't have the hand crafted feel needed to make it really work as a Metroidvania, but it also didn't take advantage of the potential craziness inherent in a procedurally generated experience. Back when the game was first announced I was under the impression that the entire "dungeon" would be different every time you entered it but because the "dungeon" only generated itself the first time you entered the procedural elements were actually basically unnoticeable (except in how they made the game worse). This alone would have helped the game a lot. Or, another option would have been to have the "dungeon" actually randomize itself to the point where you actually experienced different challenges and gained different powers in a different order (or maybe didn't even get some powers at all) on different playthroughs. This sort of thing is already proven to be a very successful approach as evidenced by the proliferation of "randomizers" for everything from old school Metroidvanias to Zelda games to JRPGs. You don't need the word itself to change every time you play if the way you explore and experience it changes.

[@krele14]

You don't realize how difficult it is to do what you described. Gaining a different skillset during the game would be almost impossible to balance with the procedural generation of maps, as long as the skillset provides meaningful options and map traversal depends on them. That "evidence" of yours only applies to RPG's, and all of those don't ever lock out the player from progressing and traversing the map by not having a certain skill obtained. It's such a fool's errand to undertake.

[@Rendat666]

Well, I wouldn't say it was impossible. You'd have to take out the element of true randomness in the skill progression, setting up 3-5 progression arrays. Plug em into your generation algorithm, using a tag system for certain obstacle/skill pairs...... All for what would still probably be a relatively mediocre game.

[@leigonlord5382]

people have made randomizers for a link to the past, which shares some elements with metroidvanias in how different items are needed to go to certain places. fairly sure some of the metroid games have randomizers.

[@krele14]

@@Rendat666 exactly my point, there's no way to make the transitions feel seamless and keep the procedural generation functional for what they tried to do.

[@_.-.]

Educate yourself. What you ask for is either completely imposible or imposible in a realistic amount of time. And leigonlord, you fucking idiot. Don't you realize that not only most of those randomizers do not guarantee that the result is even doable, but also that they just change were items spawn tops?

[@FrMZTsarmiral]

So many rooms that are just corridors with enemies repeated multiple times... It even copied the least enjoyable aspect of SotN.

[@bythenumbers10]

Wow, can't believe NOBODY said it. Fine, I'll say it, then: ROGUE LEGACY. Enemies designed for particular environments combined with procedurally generated maps. If you make the enemies and environments a little more cohesive, you can certainly re-combobulate the Rogue Legacy elements into something very much like Chasm. EDIT: I didn't say Rogue Legacy IS a metroidvania, I said it has almost all the parts. I'll expand on my last sentence. Scatter Rogue Legacy's classes/equipment/runes in the castle as items, put up some gates that limit exploration and force the finding of certain abilities before exploring further, and you're headed in a decidedly more Metroidvania direction, except unlike Chasm, you already have area-specific monsters and environmental hazards. Almost all of Chasm's gimmicks, but arguably better executed with less "saminess" of hallways and random enemies plopped haphazardly all over the map.

[@YEs69th420]

Rogue Legacy is really more of a roguelike than it is a metroidvania; it works fine there since it isn't trying to marry the two styles, instead just using one to supplement the other. Also why Dead Cells works as well.

[@teratoma.]

@@YEs69th420 there's literally nothing roguelike about rogue legacy besides random generated levels it doesn't even have permadeath, it makes you think it does because of that heir system which is ignorable and those few shrine items that you lose upon death, barely anything

[@AnalyticalReckoner]

Have you played Rogue? It's nothing like Rogue Legacy.

[@kyotheman69]

its not really Metroidvania game it was at core the in its name Rogue, as for Roguelike or Roguelite

[@Changetheling]

In fact the name is because of the story, but obviously most people would misinterpret that.

[@grfrjiglstan]

"Where are the enemies throwing projectiles from a higher platform?" 19:22 There's one.

[@icecatti]

I picked this up on eshop for 10 bucks, which I think is a pretty good value for what it is... I know I'm late to the party. I enjoyed the game, but the backtracking, hallways, shitty save point placement, and incredibly tight platforming sections made it sort of a chore. I would have loved for them to scrap the procedural element entirely. One of my favorite parts of the game is finding Drake's body and reading his journal where he writes about being stalked by monsters. When you exit his camp, you're immediately attacked by a giant wolfman in the tunnel you entered from, presumably the beast that killed Drake. Moments where character's (such as the ones whose journals you were reading throughout the game) arcs culminated... I found those moments to be very rewarding. Go figure-that kind of experience can only be handcrafted. A more intricately designed world that looped back into itself in very meaningful ways would have greatly elevated the experience. At the end of the day, I don't think Chasm needed the procedural level design to be a great game. There is already a big market for these kinds of games and the aesthetic direction of the entire game is already very polished. It's an interesting case study for what went wrong in a game for sure.

[@b1uezer]

I think part of the issue too, is that there's also been a desire for replayability, and procedural generation is seen as (and technically is) a low cost way to add replayability. I would love to play Metroid and Castlevania again with say, a different planet or fortress to trek through, and I think the developers had that sentiment as well. The thing that made those games memorable though is in direct contrast to that: deliberate design. Instead of making a well planned dish you enjoy once or twice, you get to eat a bowl of oatmeal every time you play CHASM. But I don't think they were wrong entirely, I just think that they relied to heavily on their method of creation. If they could generate the map, then handcraft some specifics, you could get the best of both worlds. Of course those, that would be pretty hard to accomplish in itself, randomness that feels deliberate.

[@DPadGamer]

Great video Alex!

[@paradoxx_haha]

Wait what 2 years ago and only 6 likes

[@ShinoSarna]

10:50 - You don't mention that Super Metroid intentionally reminds you of the Red Tower, by goading player to freeze a Ripper enemy using the Ice Beam - the same enemy we saw in the Red Tower. That's big part of what makes this section so cool.

[@Lightmagician60]

the game just seems copy and pasted alot. and it's not even Procedural Generation's fault. if the tools the Procedural Generation is given all look the same then no shit it will look very copy and pasted. if you give a Procedural Generation lets say 30 different rooms and the whole game is 400 rooms. you will see those 30 rooms about 13 times each. Ideology and creativity only go so far when you can't do basic counting. it would be like trying to fill a Museum with that can hold 200 Paintings with only 60. you would give some funny looks at seeing the same painting 3-4 times the easiest solution is to actually make a non-randomized Metroidvania. and then make a Procedural Generation that can rearrange those rooms with the rule of not repeating rooms.. hell a real "replayable experience" would be to make more Rooms then the Game could use. this way you can go playthrough(s) without seeing some rooms i don't understand why Devs try to use Procedural Generation as a way to "have the game make content for us" doing that just looks lazy. meanwhile Procedural Generation could be used to put more content into a game then it actual has room for

[@fshiruba]

Maybe use procedural generation for connecting areas, but hand-crafted rooms for bosses and puzzles/items

[@renookami4651]

Having way more rooms than a single playtrought will handle is a nive way to compensate or "brush the problem under the carpet" and can work, but if most rooms lack the ability to be interesting to begin with because of the dev's lack of vision...You can feed a hundred of them and the problem will remain anyway. Adaptive rooms are more likely to be interesting. Ais need to be fed alot of info to do anything not generic anyway and here's where the main problem with most devs using level generation comes from. A lack of rules and specifications that lead to boring rooms because "we need to go trought everything without upgrades, just in case the important ones only appear at the end" Landmarks and rules are a great key factor that outweight "pull the lever to get stairs to appear" puzzles. Let say you have a linear start, obviously no abilities needed, then room 10 is always a branching path and room 12 always need to go trought it. Now, the second path always need no ability to go trought, but need the same to get back, and ends up in a room that gives the said ability. Players will recognize which ability it is on next playtroughts, but the leveldesign itself will change depending on abilities order. Next rooms will have more tools to make their design with each new abilities. There is the same number of tools for a set part of the game, but different combinations lead to different designs each time. To avoid the "backracking should not be getting a worse ability" problem, one should just set priority rules. Like "triple jump" never appear before "double jump" (and not two skills of the same family appear in a row?), meaning that you keep the sense of progresion no matter the order. Of course failsafe rules have to be implemented aswell to balance the AIs descisions, like using all lvl1 abilities before laying down upgrades, or always having main abilities (crawl/multijump) appear after a threshold if not already done. That ensure varied leveldesign start to appear not too far in the story. The risk of having "double jump" and "crawl/slide" as the last two ability because RNG, and so a completly flat leveldesign for 95% of the game isnt something you should ever gamble on as a dev. Another problem is that many devs using this for Metrovanias use it like for Roguelikes, even knowing the two are different. In one case there is an interconnected world, in the other you can afford to go linear, with branching paths only leading to different rule set. Extra rules have to be taken in account to ensure physical cohesion of the random worlds. Like "1st encounter of two rooms from different biomes, the last spawned one requiere ability from the other biome to proceed", "3rd encounter, both rooms will have their back wall become breakable", "5ft encounter open/delete back wall" and so on. As shortcuts goes, the easiest rule is to get the recieving room to appear alligned vertically on the closest biome, like metroïd's elevators. Or to use failsafe rules to avoid the shortcut to go wherever feel wierd on a map. Conserning ennemies, some games generate them individually, meaning you get mostly 1vs1 fights that feel as repetitive as if you were playing a turn based RPG, or with little to no logic as why they appear in this order. Ennemies have to follow rules aswell, not just "be there". A good exemple is castlevania's lancers, who attack up/down on the next platform. Alone they are not that dangerous, but when one is above/under you when you fight another of these, you are glad too being able to use backdash. Same goes here, ennemie types should appear in leveldesigns revalent to the challenge they are supposed to represent. Extra rules should exist for a group with good synergy to appear in more and more rooms as the game go on. That also mean if a group have good synergy on a certain combination of design rule that don't appear in this playtrought, you won't encounter them but instead another group. In all the theorical rules I mentioned there are common themes. While incertainity is the key factor that motivate the use of random generation, players should be able to recognize "landmarks" must it be a blocked path telling them what the next ability is (or the one after is a rule allow a chain to be made like for metroïd's red tower). And while there are ennemies in almost all rooms of the game, they should stay revalent to their biome (except for reccuring ennemies linked to the story) and more importantly even for the flow, to the current gameplay. There is a lot of consider when making a game. When you randomize things, you get this amount and basically square it in every possibilities that may occur, just for the game to not feel bland. Not even good designs, balance or anything. Then you add the extra rules and pray for your game to not be broken or unbalanced again, rebalance things if needed, and by the time you're finished you realise you could create at least another game using the base engine and the extra time you lost making a wacky level generator.

[@bioniclink3026]

What you said actually happened to Super Metroid, there are several different randomizers for that, Item randos, where just the items are randomized, giving a better replay experience already, but there are also door randos, where you can change up the doors in an area (To a point), and you can add item randomization to that as well! Doing these are really fun, albeit sometimes annoying when you forget an item somewhere, but that's the fun of it!

[@AuroraBorealis170]

minecraft is a good example of procedural generation done right

[@BenoHourglass]

@Sergio Ramirez the reason why it works in Minecraft is that the procedural generation isn't the point of the game, the quest for resources to make builds is the reason for the game.

[@kooyah6420]

I think there's a good inverse example of the loot problem you described: Around the game's midway point, there's a noticeable spike in difficulty once you reach the Keep. In this zone you'll find the Arena, which has 4 challenges where you face enemies of increasing toughness, and is arguably the hardest part of the game before the final zone. Since it's not just a stray from the main path but a long journey from the nearest save point, I think most players would assume it's a side challenge for experienced players that they should return to later in the game. Then they'll carve out every available nook of the world map, only to find weapons that became obsolete hours ago instead of a utility that'd help them progress. It turns out what you're supposed to do is 100% the arena so you can get diving gear (why a scuba mask is the final reward in a gladiator arena eludes me). And considering that the weapon and armor you're rewarded with for the first 3 challenges was pretty good, I felt cheated when my reward for the final challenge was an arbitrary progression key that's only used twice to go under a pool of water. This is where the rewards should've been inverted-exploration should've been rewarded with items that enable further exploration, and combat items should've been on an optional path instead of being notable landmarks that only become accessible long after said weapons would've been useful.

even if the game is dull (haven't played it) I must say some of those enemies looks great. Specialy that lighting casting wizzard/emperor palpatine guy. The protagonist, on the other hand, doesn't looks interesting enough.

[@mrfandangojimmies7117]

I feel like the protag never got updated from the 2013 sprites while everything else kept advancing and becoming more detailed. Protag-kun's arms look like giant meat clubs

no female option no = crappy game = no play

[@mrfandangojimmies7117]

My man, what are you doing with your life. Nobody's going to fall for bait that obvious.

[@mrfandangojimmies7117]

My dude, please. Put in *some* kind of effort. This is just dumb.

Do NOT feed the trolls, please.

[@sajohnson09]

Why does chasm have that afterglow when he makes his first move after stopping like alucard? I'm not too sure that's entirely explained in Sotn, but I know it's more believable that a supernatural vampire would have it than some random street dude.

[@gabrieldemyst3541]

Ever heard of "references" or "shoutouts" in video games? Apparently not.

[@sajohnson09]

Gabriel de Myst I'm guessing you didn't watch the video, there are plenty of "shoutouts" and "references" to symphony of the night, that doesn't change anything about this particular thing seeming out of place.

[@zalmute423]

@@sajohnson09 that glow also happens in mega man x and with non vampires in castlevania.

[@sajohnson09]

@@zalmute423 Again though, supernatural beings in Castlevania and robots with crazy ass powers from the future in mega man x/zero's case.

[@elspethtirel]

On the point about generic enemies: I've been playing Castlevania: Order of Ecclesia lately, and it has these really cool rooms in its prison island area. See, the area has knights wielding lances all throughout it, who can attack you from range, which the game makes you learn to deal with early on. Then, later, you get to some optional rooms that have a ground floor and a single, long platform above it; the latter is just high enough that you can't jump to it, with an extra, lower platform on the opposite end of the room that you can use to reach it. Those knights with lances are on both the floor and the platform. As you make your way across the room fighting knights on the ground, you begin to see (or get hit by) the knights on the platform above, who are using the long reach of their lances to thrust down, through the platform, at you. Suddenly, due to the unique architecture and enemy placement of the room, a familiar enemy has become, for the moment, not a physical enemy, but an obstacle to be platformed past and fought around. This continues until you can fight your way to the end of the room, use the lower platform there, get onto the raised platform, and kill the knights you had previously been out of your reach. It's an absolutely brilliant use of room architecture and enemy placement, and a procedurally generated game simply can't replicate that level of thoughtful game design.

[@MrCrompz]

I am actually working on a procedurally generated metroidvania at the moment (oh dear!) but this video has actually been very helpful in finding out what went wrong in a previous attempt, to at least try to make a few of these issues better. One thing I have done is made sure no one room appears more than once, as I think that can be very damaging.

[@CapnNapalm]

I think that indie developers should take a break from procedural gereration for a while. Sure, it gives you "infinite" content, but it doesn't make your game infinitely replayable. rougelikes often overstay their welcome because they encourage playing until you're bored of the experience as a whole

[@afistfulofpimples1745]

I don't.

[@CapnNapalm]

@@afistfulofpimples1745 cool.

[@StewNWT]

BoatHouse yup. I see procedurally generated, I lose all interest

[@wylie2835]

Not at all. Look at Unexplored. It's PCG produces the greatest dungeons ive ever seen. Beats out any attempt at handcrafting them. It wouldn't have happened without indie devs. AAA devs can't innovate like Unexplored did. It's too risky. They will instead copy his cyclic generation method now that it is proven to generate extremely engaging content.

[@OscarOfGwinnett]

16:55 Okay, seriously though? They even copied the sword companion from Symphony of the Night?? This isn't even a game; it's literally just a mishmash of influences...

[@protolegacy9184]

I truly enjoyed the game despite it's flaws. It has great pixel art,solid control and it has plenty of enemies to defeat and discover their drops. I did fell the whips felt best in the game. Having the distance felt safer and not having to the the jump attack with the sword was nice.The trade off was that the enemies took a couple more hits or so. Then I would get a good drop item only to find it stronger than most in most if not all the treasure chests in the game. I believe its an above average game and can be benefit and made even better with some patches like having luck mean more for drop rates. I think fans of the genre will enjoy it just don't expect and evolution in the gameplay . Its just more of what you like.

[@Herptroid]

The balance on the weapons was way off and really hurt the game in my opinion. The whip's hitbox made every other weapon (save the best sword) obsolete. Why use a risky dagger which kills enemies in 4 hits when the safe whip kills them in 5?

[@pokaie]

Great Job on the Video!!

[@thehaloring90]

I'm glad that I stumbled upon this video. I have a lot of problems with rogue-likes and otherwise "randomly generated" games, obviously one of the biggest issues being bad level design or rather lack there of. This is actually the first time I've gotten confirmation that my views aren't the ravings of a madman, and I think this video explains the issues of random design very well.

[@niconeitorrr]

"an item that lets chasm crawl" you got me completely unprepared with that one, gg, well played

[@CarbonatedTurtle]

Your criticisms of this game are definitely valid, but I still found it to be really fun, despite agreeing with a lot of what you said here. I don't really mind all of the standard stuff that's been reused in a million games before, because they're things that typically draw me to a game in the first place. Exploration, leveling up, equipping new items that improve my stats and abilities. Despite the cons, I still think all the pros this game has make it worth playing.

[@Crazelord91]

I'm glad someone is speaking about this. I get so dissapointed every time I see an interesting Indie game, until I see the inevitable "Metroidvania" and "Procedurally Generated". Though I'm not the biggest fan of Metroidvanias, I can appreciate ones made with intent vs designed randomness.

[@chriswahl1337]

The video was so relevant to my new game dev project that I am eternally grateful. I'm certain the ideas that were organically sparked by this video will make my game better and help establish its core experience. Thank you so much for this video.

[@twilightiger]

In a genre where you're competing with the sheer unbridled creativity of Cave Story and the love hate relationship you'll get with the puzzles in La-Mulana procedural generation just seems like it misses the point of what make a game memorable and worth becoming invested in because if you're going to be exploring a world that someone else has built turning it into a corridor simulator mixed with a poorly disguised game of chutes and ladders means that players aren't experiencing a new world every time they play the game but rather a bland and boring rehashing of the same things they've already grown tired of within the first fifteen minutes of play. Personally, I think I was more surprised by how instantly forgettable Chasm was when compared to a number of other metroidvanias given the sheer number of Mario Maker levels that prove the adage, if enemy placement punishes the player for playing the game, that's bad design. If the level itself punishes the player for not playing the game the way the designer intended, that's bad design.

[@kyotheman69]

clearly the developers didn't bother consider this and didn't do research to see why other metoridvania type games work, this was clearly failed experiment or just developer just lacking in creativity

[@feelinghealing3890]

How about having parameters for interconnectivity at the world generation? -Are paths between areas physically close? --Horizontally % --Vertically % --Average distance ---- -Are [Type]-Items hidden in [Type] secret rooms? ---- -What generation mechanism will be used? --labyrinth % --Voronoi % --[example] % ---- -What room types will be prevalent? --Hallways % --Small % --Medium % --Big % --Special % --Hidden % Also, I dont know much about this game, but if it is not there, A few visual/audio variables would be neat. -Retro -Medieval -Futuristic -MeatPunk -Blocky

[@PlaguedByEarth]

As someone that loves procedural generation and the potential it has, it is disappointing that failures like this do nothing but discredit the notion that proceduraly generated worlds can ever match the quality of designed worlds. Thank you for not using this as an excuse to impress that idea on others. You were quite fair to procedural generation, not giving it as much credit as it deserves, but giving it more credit than I see given on average. I hope that failures like this do not discourage others from innovating with procedural generation in the future, but instead teach them through the mistakes it highlighted. At the end I don't think that failures like CHASM should be despised for not living up to their aspirations, but should be treated as stepping stones to a goal that once reached will be more all the more satisfying for keeping the hope of their promise alive.

[@MalekitGJ]

as the video mentioned, this is more a fault on the developer rather than the procedural generation code maaaaaaaaaaaaybe the guys at BitKid should have run more simulations before releasing it, and improving the code

[@KnakuanaRka]

Yeah, the issue is more with the limited resources of the creator, and the poor design of the individual rooms, than the base idea of procedurally generating your game world. I Hope groups that can do more work can improve its reputation.

[@_.-.]

Uneducated back-seat codding is what pisses me off about this the most, to be honest.

[@SebastianSkadisson]

Great video, i hope devs will learn that procedural generation will sacrifice a lot of their gameplay ideas and will dumb down the ones that actually make it, everytime. It's great for rogue-like games or rogue-lites and random loot grinders, where you only visit a map or area once, but metroidvanias, where you have to backtravel and traversal is one of the key points of your gameplay, it's always harming the experience, especially for technically focused players. Unless we are able to implement machine learning into enemy AI we will not be able to have enemies that react to a procedurally generated environment in a challanging and interesting manner. Once machine learning makes it into an enemy the moment they are loaded in, and them being then able to use interesting traversal mechanics to profit off of the environment or the way the player has to react to the environment at hand, we will be in a whole different ballpark with procedural metroidvanias and procedural generation in games in general.

[@chriswalker3375]

Chasm is one of the best games I played in to the last 20 years!

[@coreydeluna5258]

I'm really enjoying it as I wait for bloodstained

[@uba754]

Someone hasn't played Hollow Knight then.

[@mr.randomguys7629]

Derek Gruber I can’t wait for silksong

[@Gatitasecsii]

First thing I saw and hated was the orc exploding death animation. Why did it explode? was it a robot?

[@acidwizard6528]

Just wanna say, I've been playing video games since 1987 and Super Metroid for the SNES is my all time favorite game. I appreciate how you've analyzed and put into words what makes it so great. An absolute masterpiece and genius creation.

[@aurumarma5711]

Theres a reason Randomizer runs are so big. Chasm's goal is an excellent one, that I think could truly be innovative in the Metroid-vania genre. However a game based around this idea would be hard to make great by even a large studio. Chasm was Good in all places, but never really great anywhere. I look forward to seeing what the studio behind it does next, because I did like Chasm. However I feel a smaller vision, and a more refined game, would have a better reception.

[@Myrdin90]

I agree. The studio has gained alot of experience and im eagerly waiting for what they create next.

[@AsmageddonPrince]

I think it mostly failed to devs not being sure where to go with it. The level generation itself could certainly have been managed better.

[@DragoEpyon]

Randomizers are different things though. With a randomizer the level design and enemies stay the same the only thing that changes is the position of powerups. The fact that it's possible to beat some of these games while getting the powerups out of order is a testament to how tight they're programmed. What this game tries to do simply does not work and it never will.

[@efilwv1635]

“Hollow Knight didn’t do anything new or exciting” 😆

[@pollux-84]

I played and really liked Chasm alot. I also played games like SotN, Metroid, Axiom Verge etc. My only complaint about Chasm was that almost all of the good armor type items give a bonus to Int and not Str, Vit or Def like you think they would. I wanted to not use mana sub weapons at all and be more like a heavy armored melee type. Also different equipment is not visible on the character. I did not have other complaints and I think most enemies actually do fit into their environments. Bees, frogs and plants in a jungle area. Kobolds and goblins in the cave. Knights and armored skeletons in the castle and so on. Btw, what is the name of the game in the beginning -the one with the guy in a red cape swinging a sword? Looks a bit like Hyper Light Drifter but it ain't. Thanks for this video and your opinion of Chasm!

[@MagnificantSasquatch]

I’ll actually take a contrary point. A procedural Metroidvania _cannot_ work, and the reasons why you yourself give. What makes a Metroidvania successful is the flow and structure of the adventure. How you progress, the options available to you, how you acquire these means and, all the while, well-balanced and original rooms and pathways presented to make each area feel distinct and unique, so that merely the act of exploration as you look for the next key event or item is rewarding in and of itself. A collaboration of creative minds weaving an ongoing, memorable journey is absolutely key to this. You cannot program originality into an algorythm, all you can do is give it a list of pre-created assets to slap together in a way that prioritizes being functional and nothing else. A roguelike doesn’t require creativity; it’s just a dungeon crawl. An adventure does.

[@z-beeblebrox]

In fact, as far as large-scale procedural generation goes, not only can it not work well, it *shouldn't* Of all the game styles out there, medtroidvanias are probably the least suited to procedural generation, because the entire point is that their maps are bespoke creations tailor made to interlock perfectly with a very particular set of complex progression paths and art direction. It's the designer clothing of video games. You *want* it to be made by hand.

[@lollard]

Dead Cells works. Though, to be fair, it's only loosely connected to the genre, and emphasizes combat over exploration. It still works in my opinion.

[@thesilverblueman]

dead cells is more of a roguelite than metroidvania you don't really care about the areas you're more focused on pure combat and making your next run easier while in metroidvania games dying matters and if an areas to hard you can try backtrack to an unexplored area to look for something that will help you

[@foxymetroid]

It can, in theory, work. It just takes an incredible amount of time, energy, and genius. It would probably be easier to go the Four Swords (GBA) route. In that game, the individual levels are split into two parts. There are X number of part A and X number of part B. Each part is painstakingly crafted like normal levels, but which part A and part B you get in a playthrough is randomized.

[@wurst1284]

It can work and you only need to change the genre.

[@Harvestersz]

The one thing that stood out to me in this video was the fact that every time there was a clip of CSOTN I instantly recognized where it was in the map. I'm not sure how many years have passed since I played that game ... but I find it amazing that I still remember those things.

[@Gundurr]

uhhh, its *pronounced Shazam*

[@ingeniousclown]

CRAP!

[@Gundurr]

@@ingeniousclown lol

[@zeozen]

kzbin.info/www/bejne/Zpi1c3xngJ6ofpo

[@guiteshima]

10:41 I believe it is worth mentioning how Brinstar (the green area where you fight Kraid) also has a TON of puzzles that use the Speed Booster, heck, it even has a room that functions EXACTLY like the room in Norfair (and even a hallway similar to the one you use to get back to Norfair's elevator from Bubble Mountains) and since you can go directly to Brinstar from the Red Tower (if you manage to grab the Wave Beam before heading back to the Red Tower.) chances are those other rooms that are similar will still be on your mind too!

[@BostonBugsy]

I'm playing this now and absolutely love it. Procedurally generated does not affect this game at all. If you want to start over with the same map you were just playing you can input a number, "seed". At the moment I am over 10 hours in and at 70% explored. I've done a decent amount of grinding which I enjoy, I feel the difficulty is perfect. Fantastic and highly underrated game!

[@kektus4203]

1:50 rare image of Maynard James Keenan having a breakdown over NOT writing the new album (colourised)

[@ASAndrew]

It seems trendy to pick on Chasm, but there aren't too many games in recent memory I was quite as excited to play. The game is beautiful, the controls are smooth and familiar, and the puzzles are fulfilling to solve. Sure, it does have some of the problems you mentioned. But it was still a fun game to play.

[@cheater00]

I've liked your video and it made me subscribe. i like your point that the auto generated world in chasm doesn't work in the way you want it to. if you are technically inclined, you could try making a world generator in a programming language. even if you don't know one right now, this is actually a great beginner project. you could make your world a string with 10000x10000 characters. x is a wall, D is a door, etc. write down some rules that you'd want to be in place and try to implement them. for example: find items in order of increasing power; if i find a new item that unlocks new places, the main game path that needs this item is nearby; room layout meanders and stays compact rather than go in long runs and sprawl. this sort of thing doesn't have to result in a game but just generating map layouts could be lots of fun for you and could let you learn a lot about metroid style games!

[@justice7788]

I wish someone loved me the way you love metroidvania games ]:

[@Mordewolt]

12:45 you can always disable the fast travel for one reason or another. Prey did it a few times, one of which was a complete shutdown that made you revisit every single area and open up both the tram and the ship exits, a rehearse of your spatial memory of sorts, with all the monsters being respawned, naturally making you lean towards more efficient paths from one ship section to another.

[@furyberserk]

The backdash isn't useless. It's purely defensive. Shantae's backdash IS useless.

[@ivaldi13]

Shantae's backdash at least is handy enough for some evasion against ScorpGals in Shantae and the Pirate's Curse, not that you should need it. It also helps with rapidly changing direction. It has a bit of speedrunning assistance in Shantae: Risky's Revenge. Still, yeah, it's not very useful. I think that it's mostly there as an homage to Castlevania games.

[@furyberserk]

@@ivaldi13 I'm mostly talking about Half Genie Hero, and in the ninja mode, it's the best it has ever been, but the normal is terrible. It's the same speed ad her run to possibly slower and you even have to buy it. It was better in every other game.

[@RadishAcceptable]

You can backdash, jump cancel, land backdash to move at about 1.4 speed.

[@Gale_Storm_]

@@furyberserk Actually, Risky's Revenge still has the best backdash. It's like ninjas, but slightly faster...to the point to where you can traverse areas before enemies have a chance to spawn. Ninja's would be great, if it weren't for the fact that the default running speed is just as fast.

[@HiinoXydable]

Hello, this was a very thought-provoking video so thank you for that. In a comment, you said "I know they exist, but randomizers are not the same as randomly generating an entire map." I have to ask you: there is an item randomizer for A Link To The Past (i know, not a vania) that can also randomize entrances (every house/cave/dungeon/etc entrance maps to a different one), and the overworld stays the same. The result ends up being a more-or-less-static overworld/map structure with shuffled/randomized rooms and details, a design that does away with the idea of generating seemingly-new content, in favor of requiring the player to figure out how the map interconnects (which CAN generate a lot of backtracking...) and how to solve challenges using only their current arsenal (which CAN get a stubborn player to stay in a room for a long time trying to solve a puzzle that they don't have the right tools for yet). That design still works really well because it assumes the player knows the game and its various tricks (which translates to them knowing which challenges they can overcome with their current items), not to mention it's more meant to be used for races and speedruns than for regular completion. Nonetheless, do you think this could be a good compromise for a randomly generated vania map? Maybe assuming we can make the generator smart enough to avoid hurdles such as huge backtracking sequences?

[@eac-ox2ly]

Man, you made me start playing Environmental Station Alpha. Amazing game! Thanks.

[@Metallikite]

You COMPLETELY nailed it. Amazing. Thanks so much for making this. I made it 70% through the game before admitting to myself that I was bored and I couldn't be bothered to remember these vague, dull areas that I was required to back-track to.

[@FeiTheVillain]

That's your problem. Other people enjoyed the game.

[@Metallikite]

Thanks for reminding me of this disappointing game 4 years later 😂 Cheers.

[@FeiTheVillain]

@@Metallikite Still your opinion. Taken as a grain of sand.

[@Garinovitch]

Just a side note, Battleborn was better than Overwatch, it's just that Gearbox doesn't have the same cultural impact as Blizzard and Battleborn's marketing was god awful.

[@foxymetroid]

I know. Overwatch didn't even have a story mode. Battleborn's was... okay, but it was something.

[@randomguy6679]

No story mode is better than a shitty one. Yeah, i said it.

[@Garinovitch]

@@randomguy6679 That's your opinion and it's fine you didn't like it. I personally like the story and I was glad to have single player content because I never really liked PvP. Some of the most fun I've had on Overwatch was the PvE events I spent with my friends.

[@randomguy6679]

@Kevin Lee overwatch took itself too seriously? Are you kidding me?

[@randomguy6679]

@Kevin Lee Battleborn is also the one that utilized toilet humor and brodude humor. So of course it's gonna be the one that takes itself a little less seriously. But to say that Overwatch takes itself too seriously is like saying Citizen Kane is a slapstick comedy.

[@PKKing544]

Yo, the transition at 1:00 was really nice

[@afistfulofpimples1745]

No it wasn't

[@shayoko6]

good voice is willing to discuss games not mainstream seems to be intelligent these are the kind of people that are truly worth following.

[@mariahanover9335]

at 9:00 you're literally describing axiom verge. Every new item you'd have to trek across this insanely sprawling map with no interconnectedness to find the one door in 10 your item opened.

[@despawn7663]

Ha High Quality is subjective and not guaranteed. I feel the procedural generation could have a home within the Metroidvania worlds but it would need to be a hybrid approach. Perhaps dedicating only specific "parts" to be procedural. Perhaps the Mines would always be random or maybe specific sequences say (long runs thru a corridor to another area might look a bit different). Perhaps the static set pieces would be more numerous. I believe atomic verge did something quite unique, hiding some sort of distortion levels where you could gain a powerup, that was seeded randomly each play thru. Having to listen to the audio queue that this warp existed. To this day I still couldn't get 100% on my first play thru. There seems to be plenty of room for innovation in between the lines of the core level design. Or perhaps just place it as the last bonus dungeon at the end for infinite replayability in a difficult challenge world. Great video.

[@amrmohamed4308]

Eric Pescatore Which is exactly what dead cells did and it turned out to be an amazing game. I agree that if done right, procedurally generated elements can help make metroidvania games better.

[@ingeniousclown]

Wish lemme stop you there, Dead Cells is absolutely not a metroidvania, no matter what the devs say. Not even a little bit. Great game though

[@amrmohamed4308]

ingeniousclown Gaming I guess it's not exactly a metroidvania but you can still feel that it was inspired by them. If progression wasn't through levels and the biomes were interconnected then it would easily pass as a metroidvania.

[@cloverdeer4015]

I picked up chasm because I had nothing to play, and I enjoyed quite a lot. The only thing I didn’t like was it’s difficultly especially in the Keep area. But that’s what I get for playing on hard.

[@TheCaboose568]

If I could say: Making a rogue like or procedurally generated game instantly makes it a weaker core game. It takes away more than it adds, theoretically infinite content but at the end of the day you'll see the same pieces which are all slapped together with at best a bit of logic. Level creation or cleverly placed secrets, even a deep progression isn't possible. For metroidvanias you lose a coherent world which logically makes sense and can hold non-standard rooms (Typically ones holding NPCs, special visual scenes, events, so on) and you're stripped down to the basic gameplay with no variance. Closest you'll come are boss rooms, teleporters to secrets which are disjointed, and intermediary floors which act as the resting rooms or save points. It's the critical flaw of random generation, it doesn't have a human touch, any real level of detail, a sense of progression, anything. But I'm old and crazy, after all infinite content. (Though I'd argue you'd get infinite content by rolling a 10 sided die to determine which finger to skewer with a toothpick too)

[@TheMetalOverlord]

Onestly, it all depend on the game. The Binding of Isaac is completely random generated at every run, but then it has an insane repalyability(i have more than 2000 hours on Rebirth alone) and it's one of the most succesful indie game ever, it has 2 main game and a total of 4 big DLC and even after 7 years from it's release it still getting supported by the developers and still receive new fresh content. But yea, making a procedurally generated right is really hard an too many games failed at that really ahrd, while some came out ok but became boring after too few hours.

[@tigerfestivals5137]

@@TheMetalOverlord i was just gonna mention issac. The randomness is really a big part of why that game is fun. Plus unlocking new stuff when you hit certain milestones was a good incentive to keep playing.

[@edgarector]

TheMetalOverlord I think Isaac's strenght was not in its level design,all the rooms were extremely simple,it was all about the 1000 items that combine with each other,giving you totally crazy builds each time. Unlike Dead Cells,for example,where every build feels almost the same. Man,this game was a dissapointment!

[@edgarector]

Viking the Mad Couldn't agree more with what you said!

[@TheMetalOverlord]

Yea, Isaac is the only game where achievements are a REAL achievements, and every new one that you reach will improve the gameplay of the next runs. Isaac strength is that it reward all the hard effort you put into the game. But the randomness is still a big part of the game, both in items and levels. If you end up playing the same few rooms every run it can become boring really fast. But Isaac has so many rooms tha i can find new ones even after years of playing.

[@chimarvarmidium1425]

a possible fix seems to be to only do the mixing and matching on new game +. like, design a traditional metroidvania with pre planned progression and then later mix and match rooms in a biome so that the layouts are different but the progression is still the same. so for example, you would still earn item A, B, and C in rooms 1 3 and 5 but in your second run the overall room order would be 1, 6, 3, 2, 5, 4 instead of 1, 2, 3, 4, 5, 6. or in other words you would maintain "anchor" points that will be hit at the same relative time but the inbetween parts will be crazy. a good setup for this would be that the new game + is an inaccurate retelling of what the player did in the first run or some sort of mindscrew/purgatory situation. Obviously devs always have their reasons for doing something and its very easy to "couch develop". they likely set out with the goal to write as procedural a game as possible and that shits super hard to do and props to devs from what they accomplished, even taking into account this video.

[@austincrist7581]

Game at 1:09? Also, Super Metroid Randomizer IS a thing.

[@JerichoBluff]

"A Robot named Fight"

[@Daesan655]

Super Metroid and Zelda Randomizers are rom hacks made by people who have put thousands of hours into the original game and as such know how to program the randomizer. Even then randomizer runs are really only for people that know the original game well anyway.

[@schar1024]

Three things about the hidden items. They either could make it, that equipment is solely dropped by mobs and hide stat enhancements and gold, which should always be usefull. Or they could add leveling items, that increases stats and apperance by the playtime you allready got. As example you start the game and in the next room is a wooden sword in a chest. If you leave that chest alone and come back later it would contain a steel sword. Third possibility is a counter to items. So you always get the next better item in this cathegory. So you found a stick earlier (which is level one spear item), then the next spear item you will find is the flintstone spear (which is level 2 of spear items). But for the last way of "leveling" items they NEED to implement, that you can buy the next level of the item in the stores. If you wouldn't do that, then you could reach a point where the enemies are way too strong and leveling up alone wouldn't be enough to pass this stage. In this case you would need to walk back to the first areas and search them all over to get better items.

[@trondordoesstuff]

You know, when you think about, Terraria is *almost* a metroidvania. Okay, I'm going to leave now.

[@itsyaboi1245]

Fucking 200iq opinion over here

[@kfcnyancat]

I think it gives off the feeling of one pretty well without really being one.

[@rickpgriffin]

Mark Brown's video on Spelunky highlights how you could make a procedural generated game much more involving and feel less repetitive, so here's what I'd say you need to be able to do 1) The game should probably be shorter than most metroidvanias, HOWEVER I believe you can mitigate this by randomizing what equipment you DO get each run. Let's say you get five critical items each run, and for instance one of them is a soft movement item you get early (wall kick, double jump, air dash etc) and one is a super movement item you get late (levitate, flight, wall climb, grappling hook etc) and the generated levels are aware of what items will eventually be available. All of these items exist on a hierarchy, specifically naming what could potentially overrule an earlier item so the later item doesn't accidentally become an early one. 1a) I don't see this often enough in randomly generated games: you can make each game have a "theme" even if it's minor, just to make it unique from other runs. Maybe this game all health upgrades are only purchasable in the shop. Or ALL non-critical items must be bought in the one hub shop and all secrets just lead to cash stores. Or perhaps the ceiling is always covered with spikes. Or tons of blocks are breakable. Or this run has 10x the usual number of secret rooms. Or perhaps, since you have lava-swimming suit in this run, the entire run will be volcano-themed and this will be reflected in the dialogue. 2) The EASIEST way to make this game work would be a hub-and-spoke system. Generate a hub with clear landmarks that let you see which exits require some access option you don't have. Delineate spokes ahead of time, giving them specifically defined ranges so they don't spill over into one another when attempting to generate the paths. 2a) Technically you don't need a hub system but it really helps; in any case the hub doesn't need to be that big, and perhaps you could make a later spoke branch off an earlier spoke instead of the hub (still using a landmark feature to help make it memorable) 3) Generate a critical path to the first item. Ideally, this path will loop on itself and lead back to the hub, but this isn't strictly necessary as you could include a fast-travel to the end of the spoke anyway. 3a) The critical path COULD create a situation where you cannot return to the hub until you complete it. However you do it, you should make it so that exiting the hub efficiently requires the use of the critical item. 3b) You must keep in mind critical items need a use for horizontal-platforming as well as vertical platforming (both ascending and descending) in different ways, so that you can create item-required rooms that go in any direction you need. 3c) For varieties sake you don't ALWAYS need the pathing the item provides; you could have keys! Ideally the critical item works AS as key but you could make it simpler than that for some of the paths. 4) Do this for each area in turn. Each area's room possibilities opens up to different toolsets depending on what you have available, such as (air dash + high jump gives you different generation possibilities than wall jump * grappling hook, etc). In addition, you can use DIFFERENT TILE SETS FOR EACH SPOKE TO INCREASE THEIR DISTINCTIVENESS. (You could make one match the hub style even!) 5) Generate secret rooms with non-critical items, shops, tunnels, bonus bosses etc. The secret room's power level is equal to how late you can access it. Also, generate connecting branches between spokes, with a caveat: the path must be one-way from higher order to lower order of spoke, OR it must require the use of the higher-order ability at minimum. 6) You MUST BE ABLE TO COME UP WITH LOTS OF VARIED ROOM STYLES, HUB STYLES AND PATHING STYLES. If your game has "flight" as a late possible ability, then your "flight required" rooms will need lots of open spaces and tall towers, possibly some of the biggest in the game! Depending on your method of flight, it might change up the play style a lot. Some of your players may not even see these kinds of rooms! That's fine, if you're making something with a lot more replayability then you need rare content.

[@ingeniousclown]

I love 1 and 6. A Robot Named Fight does these decently well, albeit in a smaller, roguelite-flavored package. I hate the idea of hub-and-spoke, because unless it's VERY well-disguised it's practically just a sneaky level select system. IMO the "holy grail" of procedural MV world design avoids an explicit hub. Out of all these, I think number 6 might be the most critical. I mentioned it in the video, but I think if CHASM had follow #6 and had far more interesting and varied rooms (that you will NEVER see 100% of in a single playthrough) then it could have been a lot more fun and interesting.

[@NoeLPZC]

One thing I would add (and it can kinda break the hub and spoke a little) is that the player needs to pass the lock before they get the key, and they need to have passed it fairly recently in order for them to remember it. The problem with the hub and spoke is that you can go down one spoke before checking all the (locked) spokes, so when you find the key to open another spoke you don't know where to go. Ideally the entrance to the next spoke would be along the spoke you're currently exploring. Super Metroid solves this fantastically by having the first part of the game as a "world tour". You get a directed loop around the world, getting a look at all the different locks, then it takes you back to the start with a small handful of new keys. Since you would've seen several locks along the way you know where to go next. You can do this fractally too, and have each new unlocked zone after that begin with a tour and end with a game of memory, either leading to keys for the next area or the next layer of zones. It would be a good idea to have each 'game loop' (tour -> memory) cross into several different zones as well to keep the environment from being too stale (but don't overdo it or it'll be overwhelming).

[@edgarector]

I guess,the new indie trend is procedurally generated metroidvanias. Procedural generation is something that is not doing it for me at all,because everything feels souless.The design of the map doesn't have human logic behind it,that's why you can't really relate to it.There are always places that don't make human sense and you just feel you are in the hands of a computer that stirs all the level assets and pukes them. Dead Cells,as popular as it is,annoyed me so much that,at one point I just stopped playing it at all!To be honest,it got old for me really fast.I just felt a little decieved by the developers,because they were forcing me to play this game for,I don't know,80 hours,over a very,very thin content.The game has,probably,10-12 zones and 4 bosses,that's it.And you are forced to grind cells in order to unlock better drops,in order to have a chance at higher difficulties. You are forced to go through each zone(biome) literally hundreds of times in order to unlock those drop rate percentages.And to make matters more annoying,you can only fill those drop bars after you defeat a boss. By the time I was fighting the Concierge for the 20th time,I was like "Why,the hell,am I even doing this?!" Also,although there are quite a lot of weapons and items in the game,all runs,kind of,feel the same.All weapons have only one animation and the elemental effects all did just damage over time. It doesn't matter if you choose bleeding,burning,poisoning,it was the same thing,only with different colours.The procedural generation was so shallow that,honestly,I would've preffered if that game had traditional approach,with fixed maps,like Guacamelee or Outland.

[@PlaguedByEarth]

A badly designed game that doesn't use procedural generation feels the same way. Don't blame the technique, blame the artisan. Remember that computers and ai are our tools and our children, extensions of ourselves. There is nothing more human. We'll get there some day, but bad games like CHASM only serve to discredit complicated ideas which haven't even come close to fruition.

[@Genesix6]

The concierge is easy as hek, I don't see how it is hard on the other hand if it the last boss then yeah it is hard

[@edgarector]

phillip gianto It's not that it's hard,it's trivial.I am talking about the unbearable repetition.

[@Genesix6]

@@edgarector I can see that, but those type of games are not metroidvania to be exact, their are called roguevania. Based on your comment I don't think you are a fan of roguevania.

[@edgarector]

phillip gianto Maybe you are right,maybe I am not.But then again,if it's made comletently and has deep enough content I am gonna put a lot of hours into it.I am a fan of Binding if Isaac(although not rxactly rogurvania),so no,it's not the genre,it's the way the game is developed. Dead Cells is actually a very shallow and repetitive game imo.

[@Baron_Blue_Max]

I've never seen a game review with Jazz playing in the background and now I never want to see one again without it

[@SkySaito]

The KZbin algorithm proposed me this vid. I will say That was a good recommendation, you got a new subscriber mate ' v'

[@marksmod]

Neural Networks is the key. For a room, generate a set of button presses with a NN which is trained on button-press data from various platform games. Have a second NN generate room elements around the path which the button presses describe in such a way that the player can go from point A to B, have a third NN judge the hardness of the generated terrain and reward/punish the second NN.

[@RayTheProducer]

This is an extremely unfair assessment of a wonderful game crafted with love and passion, and one that continues to improve. It just seems like you have an axe to grind. Sure Chasm's not perfect, but singling it out for a 20 minute rant seems a bit excessive. Especially when even genre luminaries like Axiom Verge and Hollow Knight suffer from their own sets of problems (where to go next, some areas better than others, backtracking, repetition etc.) I found Chasm to be a lovely, relaxing, intriguing and nostalgic 20 hour adventure. It brought back warm and fuzzy 16 bit memories of playing games like Black Tiger and of course, SOTN. The art style, the outstanding soundtrack, the accessibility and replay-ability, the variety of enemies, all make this a solid 8.5/10 game. It has a feel of a long lost SNES game. There's a sense of mystery to it, and it nails the atmosphere, even if it is a bit on the staid side. And with the new arcade mode, it's actually fun to just jump in for some randomized hacking and slashing. As far as your point about the procedural generation, I've seen some of your other videos and man, you definitely have OCD about this issue. Sure, the process can be hit and miss but again, I am not sure why you felt so strongly about Chasm's procedural generation that you decided to dump on it for 20 minutes. The procedural generation is there. It doesn't hurt the game. If they'd put it there and not told you, you probably wouldn't have noticed. Maybe it's a marketing mistake on their part to put that front and center, but so what? That's marketing. And it's kind of a cute concept: hand crafted rooms (and they are hand crafted) that are shuffled around slightly each time. Your central theme that procedural generation in a metroidvania can't work is pedantic at best (and HUH? at worst), when even the revered Super Metroid has a randomizer. The recent Bloodstained has one too. I've played a lot of Chasm, and the procedural generation/randomization has never bothered me. Sometimes a room isn't where it was, and other times it's exactly where it was. Do I care? Or even notice? When 99% of games have the same layout each time? It's not like they just copy pasted a bunch of rooms created in a vacuum and let the randomizer do the rest. They created a cohesive interconnected world, and then added the twist. It's not even the game's selling point. It's just A point. Also, your comments on boring rooms is highly subjective: every game of this ilk will have a few dull areas. And while there were levels I liked more than others, this has been the case in pretty much every metroidvania, or side scroller I've ever played. Again, why dump on this particular video game? I don't mind a critical review, but you're not really furthering any sort of useful narrative with this sort of video. Unless your goal was to come off as hyper anal retentive. In which case, congrats? Now, your What Makes a Good Metroidvania video is infinitely more useful. More of that please. And readers, please go out and get Chasm if you like retro platformers, hack and slash games, and especially Metroidvanias. You won't regret it.

[@Dargonhuman]

After reading the comments and watching the video, it occurs to me that one possible way to find a balance between good level design, meaningful exploration and decent randomization would be to not randomly generate a new map on every playthrough but instead hand-design several variations of each section for the algorithm to choose from, then randomize where each key item and cross connection appears across the entire map. That way each section is still cohesive and fun but maintains the sense of mystery, especially when you don't know which items will be in which sections. At the same time there's far less reliance on the algorithm to generate a good overall map from scratch. It's an idea that still has problems - chiefly, development time would be bloated up to account for designing the map variants, and it wouldn't take long for a dedicated player to encounter every variant multiple times. There would also be a massive headache trying to balance all of the different key items so that the algorithm doesn't place an overpowered end game item right at the beginning, but given everything being discussed it could be an acceptable compromise between the carefully hand-crafted experiences of traditional Metroidvania games and the sense of random discovery in procedurally generated games.

[@Luos_83]

Have to fully agree with this video. I loved Chasm's looks, and it had so much potential. Waited many years to play it, only to come across the same fricking left/right platform jumping room (you mentioned) too many times. Just cant play the game anymore, if I ever need to jump around in those rooms again.. I'll pull out the last few hairs I have.

[@TheAlison1456]

What you said about shuffling the rooms in a Metroidvania is actually a fantastic idea. Keyword being Shuffle - there's already a default set path, but you can shuffle it if you already played it.

[@theroseexperience]

As soon as I hear "Procedurally Generated" I'm out as a gamer. I want levels crafted with optimum playability, story and gameplay in mind, not some random generated level that means nothing.

[@megafanzach]

You should play splunkey.

[@theroseexperience]

@@megafanzach No, I'll hate it, I'm out immediately with Rogue Like/Lite and procedurally generated.

[@bzking24]

I'm playing this on Switch and I got no complaints at all, I love Castlevania and pixel art so this is awesome for me I love it lol

[@DokuDoki]

What non-roguelite game does the procedurally generated world right?

[@ericpschoon3922]

Minecraft?

[@ingeniousclown]

I'd agree, Minecraft feels great. But the challenges that it has as a sandbox game are a lot different than what a Metroidvania has to solve. Being a sandbox, it doesn't necessarily have to restrict things and, even if it does, it's easy enough to do.

[@ericpschoon3922]

ingeniousclown Gaming Yeah exactly, due to how open Minecraft is creating a world by hand for it would be limiting and ultimately a waste of time

[@nex7402]

I think to answer this properly defining what you believe as a non-rougelite would help X-X

[@nex7402]

@NeverFinished 3Digits I've played it myself and it feels like it's missing something. it feels like I'm seeing the same thing again. I like that different power ups help you do the same thing that's cool. just a little more map variety and I think it can be amazing

[@tims6046]

Alot of older games like Super Metroid actually have hacked roms or cartridges with Randomizers on them now. The only issue with these Randomizers is sometimes they accidentally randomize an item you need to pass an area past the area.

[@angeldude101]

While Super Metroid probably doesn't care due to the number of sequence breaks, a randomizer that doesn't rely on them should have a dependency graph that specifies what items are required to access certain areas and what items are picked up in a given area. For something like a metroidvania, it would probably help to minimize the graph traversal and make sure that there's a path without too much backtracking. Procedurally generating a roguelike, randomizer, or metroidvania essentially comes down to a graph problem.

[@Ereson]

Fantastic video, I was trying to figure out why Chasm looked pretty underwhelming as a Metroidvania after trying to watch various speedrun's of it. I think the environment design is part of the problem as well. The area's don't feel distinct enough visually, every area is some kind of variant of "Cave", so the environments get boring to look at very quickly.

[@juliohinojosa6923]

Hmm isn't chasm synonymous to cave? Or at least highly related

[@Torqegood]

If they just marked the map to remind you it might help with the issue, or let you mark the map.

[@Psilocervine]

This has been in my recommendations for a few days and I was avoiding it because these sorts of things can be super cynical. I'm glad I gave it a chance, however, as this is a great breakdown of the game's flaws and how we need to focus more on context driven procedural design if we're going to keep going forward with it as a concept. You can actually see this sort of thing in a lot of game design tools that rely on procedural systems for content creation (The building creator the AC games have used in recent iterations come to mind). I do think there is hope for procedural content going forth, but making the systems that allow for it to create memorable experiences is going to require just as much effort, if not more, as directly hand crafting the experience.

[@Myrdin90]

Slay the spire is a great game that uses procedural generation, or maybe randomization, to it's advantage. Similarly to FTL, a beloved game. All the encounters are hand crafted, but the way to the boss is semi-randomized and the loot you find is also semi-randomized depending on the tower leve you are in

[@cullenlatham2366]

i say without playing it but it sounds like the best way to make a game like this work would be to design a set number of rooms, likely more than what actually appears in a single run (although having 2 full sets to the point of no repeats other than powerups may be a bit hard to attain) with a formula that forces thematic areas (it checks to see how many discovered rooms of that style you have found, if below a certain threshold, the room you enter matches theme, if above the threshold, there is an increasing chance to change areas. every area theme has a limit of adjacent theme types, and the formula can only load a theme if it is adjacent to a room with a matching theme or the theme to be loaded has yet to be discovered). I know absolutely nothing about coding to confirm or deny it as a possibility, though. heres some random numbers as an example: 20 forest rooms, 13 of which will appear in any run, and can only connect to mountain or cave, but not both. after 7 forest rooms appear, the chances to discover a new theme increase by a set % every time you enter a new room until either you discover all 13 rooms that will appear in the run, forcing the 14th to be a different theme, or a new theme is loaded by a successful probability roll. Due to only connecting with one area, must be first area player explores. meanwhile, mountain can only connect to lava, church, and cave as return only, requiring either cave power up, or an undiscovered theme upon discovering region. cannot be last to discover. Again, purely example numbers, but i hope it illustrates my idea.

[@italolinarez9978]

besides all their faults, i really enjoy chasm gameplay

[@OmegaMouse]

Procedurally generated Metroidvania might be possible with A.I. someday.

[@ribbon_dye]

It sounds like the problem isn't quite the generation, but rather the lack of effort put into it. I've played several games which do a similar concept, and will use one in particular as an example. Dungreed is an action roguelike, but every room seems to be designed individually with enemies chosen specifically to fit the rooms. Each has its own way it can connect to other rooms, sometimes with having a particular exit blocked off if its not connected to another room in that spot. But overall, everything is hand crafted to offer a personal experience. The problem with Chasm sounds like they took the easiest route and let the random generation handle everything, rather than make individual rooms for it to pick and choose how to slot everything together.

[@IamSalvatore]

The game took a long while to come out, so that seems a bit off to me. But I do think you are on to something, just that it has more to do with leaving the game to cook for too long. So aspects of the game ended up getting forgotten partway through the creation process, leaving us with a game that because it took too long to come out, possibly because of creative indecision and the challenges of coding procedural generation, that we are left with the release version but they forgot about certain core parts in the rush to finally be done with it.

[@marcelob.678]

16:15 How do you acquire that jump in SOTN? And why would you use it when Alucard can just turn into a bat and fly away?

[@ingeniousclown]

You usually get it before you get the bat upgrade. Also it's much faster to use than the bat so it's much more convenient a lot of the time

[@marcelob.678]

@@ingeniousclown rip. Ive played the game twice (once one ps1 and then on psp) and never found it. Guess I'll have a third playthru now. Thanks for letting me know.

[@CHURCHISAWESUM]

Try A Robot Named Fight for a well made procedurally generated Metroidalike with its own twists. I really love it and keep coming back to it. It's difficult but not in an unfair way. Always my own mistakes send me back and dying isnt frustrating when every death brings new stuff into the next game (for a while). Ive been playing more than 10 hours and I'm still finding new stuff. Also the whole game was made by one man and marketed by his wife, very impressive

[@chadschmaltz9790]

I'm only a few hours into it and the only thing I wish is that there were more enemy types. It seems like there are only 10 different enemy types with several being just slight palate change of already seen enemies. I've only had 1 successful run so I may be speaking too soon.

[@CHURCHISAWESUM]

I think there are like 20 different enemy types, to be honest I find more new stuff when I die than when I'm winning, lol. The other day I got spawned in a forest away from the robot city for the first time, really cool to have different starting stages imo The forest also had some unique enemy types which was cool

[@chadschmaltz9790]

@@CHURCHISAWESUMThanks! That sounds awesome, I look forward to seeing it.

[@StringsAndSouls]

A Robot Named Fight is the biggest hidden gem I found this year. Such a good game should have more recognition :c

[@radical_rat]

I think about the best way to reach the "ideal" of a procedurally generated Metroidvania is stuff like the item randomizer ROM hack for Super Metroid. You get the cohesive world and memorable landmarks of a handmade world, but the path you take through it and the challenges along the way are going to be new and fresh every time, and you're forced to approach the game with a degree of creativity to adapt to the distribution.