Occassionally Draws Stuff

What really burns my ass about the Celeste modding community being obsessed with execution of janky glitch tech as the sole legitimate measure of difficulty and concomitantly having zero respect for puzzles is that the Super Mario romhacking community is turning out these amazing puzzle levels that are being played competitively at GDQ events and such, in spite of the fact that Super Mario isn't even a puzzle-centric game, while Celeste, which very much is a puzzle-centric game, is basically entirely absent from that scene.

(I guess the one consolation is that it's almost certainly a transitional phase. The Celeste modding community is basically where the Super Mario romhacking community was a decade ago in terms of attitudes toward what it means for a level to be difficult, so there's a good chance they'll grow out of it in another few years. It takes time for a modding scene to mature, and Celeste is only five years old!)

My guess is that since speedtech is still relatively new for a lot of ppl, they find those things engaging. Like, notice the difference between how dream blocks specifically were handled in SJ and SC. A lot of the SC dream block maps came down to dreamhyper spam. No dreamsuper, dhyper redirects, none of that shit. Just plain dhypers. And I remember there being, like, two or three maps that were just that. Meanwhile, in SJ. dhypers aren't treated as novel or difficult enough to be an entire map.

I think tech execution maps are universally enjoyed in the community specifically because it's not hard to make a decent tech execution map, but making a bad puzzle map is very possible. So, you can get this loop of going "well, I don't like puzzle maps, there are no good maps" therefore "I don't like this map because it's a puzzle map." I got that way about jelly maps where I just got thrown in the deep end and was like "fuck I hate this." The community also has a stick up its ass about stuff like cycle maps, which I think it's starting to chill out about. Cycles are another category of map that's hard to make work.

And there are a ton of mechanics that the community just avoids working with because it's hard to get right and therefore unfun. Stuff like Badeline Boss (despite literally being on of the best boss fights in a video game), Feather, Wind, Snowballs, Seekers (Temple Monsters), Core Dash, etc. It's just a ton of unused design space.

Then there's advise that's a good rule of thumb but gets taken too strictly. Stuff like indicators and sight readability are generally a good thing. It fucking sucks banging your head against a wall only to realize the intended solution was just not something you would've thought of, especially in a precision platformer. But at the same time, indicators are horribly misused in SJ. Thinking with Portals is the worst offender for this because the map doesnt feel like it was designed with indicators like that in mind, which sort of ruins the "thinking" part of the map. Indicators are meant to be hints, not solutions, which I think a ton of SJ's maps miss. When I think of indicators, I think of stuff like camera offset triggers and touch switches. Not decals saying when to release the jelly or whether you should dream jump. I think the only map where I liked this style of indication was in Square the Circle where the map felt like an introduction to the concept of mechanics interacting in weird and unintuitive ways.

Also, I want to point out that some of the most beloved maps in the community contain, at least, light puzzle elements. Glyph is the biggest example. But in general, light puzzle elements is a good way to break up gameplay and make maps feel more unique. And "light puzzle elements" could mean anything. I personally like making GM+ maps that interpret the concept as a "do you understand the fundamentals of why tech works to figure out what batshit tech I'm trying to set up?" Stuff like doing a bird ultra into a corner jump, realizing that you can get a boost from those grab trigger temple gates if you superdash into them while upside down, or realizing that with the superdash variant enabled you can perform the set-up for an ultra to increase your dash length while still steering it to reach places that you otherwise wouldn't be able to. That type of stuff is just based on what I personally find appealing tho. There are a ton of other approaches to puzzle elements. Just make something based on what elements in Celeste you enjoyed thinking about then trust the player a little to actually solve the puzzle.

Another thing I think is important to mention, the mentality behind playtesting in Celeste's community is not helpful for this issue. I think in order to see proper results for playtesting something with even minor puzzle elements, you have to see the playtesting VODs. You have to take note of what the playtester's initial approach to each screen was. How long did it take for them to realize they're doing it wrong? Were they able to brute force it? Instead, the standard method for playtesting in Celeste's community is basically to write everything in a Google Docs. SJ also had the specific thing where playtesters could opt out of playtesting puzzle maps, which I think was a mistake. The term "puzzle map" isn't a complete categorization. I personally think Dropzle is the only puzzle map in SJ, Gift from the Stars was more of psuedo-puzzle map because while it definitely prioritized puzzle solving over platforming, it still had elements of platforming in there. Then ask where the line between puzzle and not-puzzle is. Like, is Glyph a puzzle map? I don't think anyone would deny the puzzle elements, but I think, at least, a noninsignificant number of people would say that it isn't a puzzle map. I personally chose to opt out of puzzle maps because my response to puzzles is brute force. Like, if I playtested Dropzle, I literally would have nothing of value to say (that's constructive anyway) because I solved half the rooms through trial and error and the other half by making the pixel perfect jump. But it also means that playtesters who opted not to playtest puzzle maps likely saw light puzzles in the maps that they did playtest and were turned off by it, which could have contributed to the decreased value of puzzles. I also cannot speak to other ppl's experiences, but it felt like the maps I was being asked to playtest, difficulty wise, were right above my comfort range. If that holds true for other playtesters, it's possible that contributed to the conceptual difficulty being nerfed, since the mechanical difficulty was too high and unquestioned. This is more an issue with how SJ specifically was organized. I think in the future, playtesters should've been given a mix of maps on the edge of their comfort zone, around that level, and a bit below. Like, I think I'd have been able to provide useful feedback for Intermediate and Advanced maps (all of my playtesting was yellow - red expert).

I also want to quickly defend the tech-based difficulty system currently in place. From the perspective of organizing a large collab like SC and SJ, which are the two main instances of this system, this system does a lot to help manage the difficulty curve of these collabs in spite of them having dozens to hundreds of contributors. While I agree that outside of Beginner to Intermediate (where the difference is no tech to yes tech), tech is not a solid system of judging a map's difficulty. Especially starting in Advanced lobby, you can very clearly see maps that are by all means a map of a previous difficulty but with a minor amount of tech outside of that lobby. Probably the two biggest offenders are Lethal Laser Laboratory (Adv) and Hypnagogia (Exp) which both felt like Intermediate maps but with Adv/Exp tech thrown in. But the reason why this system is good for managing the difficulty curve is that it creates a pretty clean pattern where the player can learn tech in the gym as an intermission without overloading themselves within that gym. Which I think can be seen with how tech placements were altered between SC and SJ. In SJ, for example, Corner Jumps, and Puffer Boosts got moved down while stuff like Theovater got moved up. Personally, I think in terms of tech execution small Chained Ultras and Demohypers could easily go into Expert without causing any issues, but that's neither here nor there. The advantage of this system is that it prevents the tutorials in the collab from being overwhelming and helps manage the difficulty curve because it standardizes the tech used per lobby. I think tech execution is also just the only objectively measurable factor to a level's difficulty, which is why it's so heavily leaned on in large collabs. Smaller projects shouldn't feel constrained to these collab's definitions or difficulty ratings tho. Like, take base game, for example, where would you categorize 7C and 9a? Or what about Monika's D sides? I personally would stick them both as Red Intermediate, but a lot of people argue that they'd count as Green-Yellow Adv. My point is that these levels don't conform to the modding community's understanding of difficulty because they're not based around SC's and SJ's models. Also, I think a major reason these models exist is because Celeste mods are almost entirely stuff that's harder than base game, even amongst people who don't make mods because they thought base game was too easy. I think part of that's because base game gives the player a flavor of speedtech at its highest levels, so the common entry point into modding is players who are able to wallbounce and wavedash. So, really the difficulty schemes represent something to lines of Beginner, for players who beat the main story but didn't do much else; Intermediate, for players who beat something from 7b to 9a and are looking for easier spectrum gameplay; Advanced, for players who have beaten 9a and are looking for harder gameplay; Expert, for players who are comfortable in the modding scene and may still desire challeng; Grandmaster, for players who want to be eviscerated. At least, that's my interpretation of the categorizations. But different people are going to have different takes on it.

Another thing I want to point out for SJ specifically. The cracked difficulty meter really should've been implemented into every lobby. I think SC got away with it only being in GM, but SJ had so many maps that were just cracked difficulty for their lobbies. Gift from the Stars is a cracked beginner map, Deep Blue was a cracked intermediate map, The Tower XVI was a cracked advanced map, Psychokinetic was a cracked expert map. They're all maps where their difficulties somewhat overlap with the next lobbies. And they're all maps where their difficulty goes beyond their tech grades. Gift requires the player to be super intentional with their dash, which is a characteristic that I think starts to become relevant starting Intermediate; Deep Blue requires the player to manage a lot of things at once rather quickly thanks to the jellies and zip movers; the Tower XVI required a lot of finnicky precision to its movements on top of being very puzzle leaning; and Psychokinetic has a ridiculous input density that's not normal for Expert. I think listing these and maybe a few others depending on the lobby as being cracked would've helped since it would've indicated that these maps are really difficult but in a way that makes them more than just their tech grade.

As a whole, difficulty is hard to properly judge. So, I think the tech-based model is effective for large projects like SC and SJ specifically because it standardizes things a little bit for each lobby. But it's not a fullproof model and it absolutely isn't applicable to smaller projects. I'm not sure if the complaints about tech to difficulty are directed at SJ or just Celeste mods in general. I haven't really been keeping up with smaller projects recently. But I'd generally recommend people mapping to consider tech as more of a rough guideline to what type of player your map is going to court rather than a definition towards a map's difficulty. That's probably a better usage of these metrics.

I want to be smart and be seen as smart by others, its one of the only things I latched onto as a pup that could make me feel like I had something that neurotypical people didn't. It's at best a silly fantasy, but since I'm white I don't really know if I can have a productive relationship with "intelligence" in the first place. Much to think about.

Maybe you'll find this helpful? But I think the concept of intelligence sort of dissipates as you get older. Because everyone works by their own metrics and specialties. Like, I couldn't tell you the first think about biology, that doesn't make me less intelligent than a dedicated biologist. Just means they specialize in something that I don't. And applies to shit that I'm good with too. It's all just experience. Even within the same profession, you wouldn't look at two brilliant artists and go "yep that one's smarter," without coming across as extremely reductionist. It's sort of hard to have a good relationship with a false concept.

I think the main thing that made me dismissive of the concept is that I couldn't define a coherent metric for it. Even as an abstract concept. Is Intelligence a measurement of the amount of information someone knows? One's propensity to learn and accept new information? Is it a collection of general knowledge or is it specialized knowledge? What information constitutes one's intellect? Is it a static variable or one that's in constant flux? Like, I've generally been above average at math in elementary school, but then I hit a brick wall upon reaching higher levels of it, which involved invoking a different set of skills. So, what does that say about my intelligence on the matter? Another potential metric for the concept is one's propensity to think. But the act of thinking, itself, isn't a measurable concept either. Like, it's the closest approximation to how I'll use the words "smart" and "dumb." Basically as ways of saying "you gave me a lot to think about" or "I don't think you've put in enough thought." But there's a more accurate set of words for that: thoughtfulness, nuance, etc. For the sake of comparison, I'd argue that wisdom is an abstract concept, but still a real one. It's the measurement of one's life-experience. It is abstract because it cannot be objectively measured, but its components are commonly understood.

Also, when I coin the term false concept, I'm just saying it's a concept people invoke and attach meaning to despite the concept itself being rather meaningless. When I say an abstract concept, I mean it's a concept that lacks quantifiability and yet still holds meaning.

I personally find talking about concepts and philosophies to be rather annoying because language is fundamentally imprecise. I dont think thats a fault of English specifically. I think it's more that language is how a person translates their thoughts into a format understood by their peers, and I think my autism just gives me a heightened sense of awareness to its function as a translator. Every word carries a degree of nuance and air for interpretation that makes communicating ideals, especially with non-autistic ppl, tedious. That's also why I end up using a lot of big words and coining a lot of terms; feels more precise.

Trying to explain to my friend this concept

this is a good step to take, but enlightenment is realizing that you wouldn't like something if you didn't believe it was good on some level, even things that seem superficially "bad" at first.

That's a pretty lame arguement in my opinion. Like, you have to scrape the bottom of the barrel before you find art with zero good qualities. But there are plenty of instances where the bad outweighs the good and you still enjoy it (and vice versa).

Sonic is early 90s cool and Shadow is early 00s cool, they should have just kept introducing new guys for each new era of cool and have them all coexist in a big dated hodgepodge. What would the early 2010s hedgehog be like?

For some reason my imagination immediately jumps to an eternally disappointed grey Hedgehog who simply says "you are cringe" with a straight face.

There aren’t many video game related hills I’m willing to die on, but dammit, modding out the neutral wall jump exploit does make Celeste a better speed game. I will literally fight you in a parking lot over this.

Gotta agree with this take. Personally, I have a few neutral techs that I think are funny (mainly neutral off lava blocks lol), and the interaction with wallbounce is theoretically interesting. Also Neutral + Theo Crystal I think is harmless to potentially interesting, mainly because it's not too limiting on Theo Tech. But those are overall niche instances where Neutrals in mods are either harmless or feel good.

As a whole, they tend to fucking demolish the concept of stamina management. Maps designed around the concept basically can't use vanilla-only mechanics without running into the dilemma of neutral spam. There are modded entities and techniques that lend themselves to stamina management without requiring the disabling of Neutrals.

Here are some potential examples I thought of.

While these are all potential work arounds for Neutrals, it doesn't exactly negate the core issue that Neutrals are a badly designed mechanic that greatly limit the options modders have for making stamina or the grab button a relevant mechanic. You can feel that the most in stamina management maps, but it's also noticeable in maps like Glyph that would just fall apart with Neutrals. Also to be clear, this list isn't me saying "do this instead of banning Neutrals." I'm more highlighting the types of work arounds that are needed to make Neutrals play nice with the grab button and/or the concept of stamina. Some of these solutions come down to overhauling Madeline's moveset, a few aren't applicable without significantly altering the level's mechanics, and the rest are just not enough if you're trying to make a good wall segment in a level.

I don't think I've really stressed properly how limiting they can be to design spaces simply because they make one of Madeline's three action buttons and a major mechanic in the game functionally obsolete. Which basically withers an entire branch of potential gameplay, especially for maps that aren't strictly dedicated to it. It's also why a lot of climb sections of modded levels fucking suck. You just have to dramatically overhaul the game to make a modded climbing section engaging.

There's also definitely a push and pull in the community about this since a lot of modded players really dislike Neutrals being disabled because it feels like the mapper is limiting their movement options, while also Neutrals are such an oppressive mechanic to design around that disabling Neutrals is often the most effective way to make certain styles of maps functional.

Also, I want to explain why similar techs that let you cheese levels aren't nearly as toxic. Mainly: Hyper Jank and Hitbox Jank.

Hyper Jank

I'm referring to all tech related to dash-jump. Hypers, Wavedash, Supers, Reverse-Wavedashing, Ultras, Dreamhypers, etc. These techs can demolish levels that aren't mindful of their existence. Most notably levels seeking to emulate base game (as well as a lot of the base game). Like Neutrals, they're basically standard mod tech. They basically allow the player to travel the distance of a dash without needing to actually dash, most variations barely if at all require specific set-ups and in some ways they limit options for design space just as badly if not worse than Neutrals do. It's also possible to disable them, which is a measure I've almost never seen taken.

But, Hyper Jank works because even though it limits some options for level design, it also has a lot of nuances and variations, there's also a lot of flexibility in how modders can implement the concept. It adds at least as much as it removes from the game. Plus there are already built-in checks to keep Hyper Jank from taking over levels in the same way Neutrals do simply because you need to dash to Hyper and Celeste is primarily built around its dash. As a result, Hyper Jank is really healthy for the modding scene.

Hitbox Jank

I'm primarily referring to any tech built around the concepts of exploiting spike hitboxes and Celeste's leniency mechanics to perform tech that does not look physically possible. Stuff like Corner Jumps, Demo Dash, Spike Jumps, etc. These things can really break a level if done consistantly. Even GM stuff can be demolished by it. They require practically no set-up even when set-ups both exist and can be performed consistantly with practice.

The main reason these never demolished the modding scene is pretty simple tho; they're barely RTA viable, so there's no need for levels to really consider them. Also building around them is really simple. The only exception that comes to mind is Dropzle (Strawberry Jam) where I think landing the pixel perfects just ended up being way easier than the creator bargained for, especially when compared to the actual level, which resulted in a lot of players, experienced and inexperienced, just skipping rooms by going for the pixel perfects instead.

As a result, they're really fun speed techs to work with since it's often "do I want to solve this screen or do a stupidly precise thing to bypass the intended challenge?" And I it's a major reason why TASes are so fun to watch.

okay celeste community: i wanna get into celeste mods and i have minors experience with normie celeste difficulty--i have beat most of the C sides, beat the Core but not Farewell yet, and i been teaching myself some small advanced techniques like ones taught in the game or was easy for me to figure out myself--but im definitely not near the level of advanced hardcore celeste players.

i wanna try to get better at advanced moves and i have heard ppl mentions specific mods to train them but i dont know how to find them or even to find what all the possible actions Are. ive tried looking it up before and got overwhelmed by the amount of similar technical terms im not familiar with and no organization between beginner-advanced and advanced-advanced techniques.

so my question: does anyone have any specific mod/map recommendations that are good for learning specific techniques in celeste, or a concise and clear list/explanatory video for most techniques for someone who doesnt know most of them or how to do them?

Main Mods:

These two mods are basically exact descriptions of what you asked for. But yeah, if someone mentioned a specific mod that they learned and trained tech on, it's probably one of these two.

Strawberry Jam - it's a series of quick maps and it has gyms and a library to teach you the Celeste tech. Levels are divided into five difficulties ranging from A side to God Help You. And it manages to have a comfortable difficulty curve. But the main thing I think would interest you are the Gyms. Which are areas that will explain specific tech to you and give you a chance to practice them.

Spring Collab - the mod Strawberry Jam was primarily based on. Similar formatt. If you want a quick comparison, SJ has a lot more polish, better tech tutorials, a smoother difficulty curve, and better visuals, SC is more experimental, vintage, and is still really impressive. I think either mod is fits the exact description of what you're looking for.

Other Big Mods:

Mods that are largely recognized in the community and worth checking out.

Glyph - one of the best mods in my opinion. Its difficulty is a small step above Farewell but nothing unmanageable. And thats only really from the perspective of tech execution since Farewell and Glyph for the most part express their difficulty a bit differently from each other. Main thing is that Glyph is good if youre comfortable with Wavedash (dash jump, taught in Farewell) and Wallbounce (dash-jump off walls, taught in 7B). The map had a really satisfying pacing, it's gorgeous, and it has a fun custom mechanic. The map doesnt ask you to do any tech not taught in Vanilla. Would honestly recommend to anyone who can beat base game.

Into the Jungle - a mod that doesn't really do a ton of tech. Lots of custom mechanics. Several levels. Low difficulty.

Conqueror's Peak - it's a campaign with a similar difficulty to Glyph. It's got the weakest visuals out of anything I've recommended but the gameplay was really fun. Features a lot of custom mechanics.

A Few Less Major Mods that I'd Recommend:

A series of maps and campaigns that I personally like that I think fall roughly within the skill range you described.

Far Away - if I remember properly, it's around early to mid B Side difficulty. It's just a nice map with a gimmick that I really enjoy.

Path of Hope - An old deltarune themed map that's around the same difficulty as 7a. It's always stood out in my mind as a fun beginner map.

Heather's Utopia - the hardest map I'm gonna recommend because it asks you to do Celeste tech upside down. I really enjoyed this map when I played it.

Escapism - a short map using the Feather mechanic. Stands out for its visuals and abstract story.

Cryoshock - feels almost like a base-game but with wavedash. Uses custom mechanics. Very blue. I had a ton of fun with the gameplay when I playtested for this.

Anubi - I had a hard with visual clarity (foreground and background bled together). But, it's style of gameplay really stood out to me. It's a lot of wallbounces, but it's done in a really satisfying feeling way.

There are definitely more maps out there that slipped my mind, but this should be more than enough to get you started. All of them are available on Gamebanana. If you want a tutorial on how to download mods, here you go:

https://m.youtube.com/watch?v=h-f2unIoafA

"if you don't like, don't watch" sorry but i don't respect the opinions of people who only consume art they like

i promise i’m not trying to be a troll or inflamatory, i just want to know: what is the point of engaging with something you know you don’t or won’t like?

how can i believe you when you say something is a 10/10 when you don't know how to articulate a 3/10? its just like broadening your artistic horizons, you also need to broaden the bounds of your taste to be a good critic

I totally agree with this. Like, I think my abilities to critique grew the most when consuming pieces of art that I disliked but couldn't quite put my finger on why it was bad.

Hi. I’ve endorsed the game Celeste plenty of times, but now I want to beg everyone who’s played Celeste to play a specific mod for Celeste: the Strawberry Jam Collab.

It’s amazing in every regard, and accomplishes something no other similar project ever has, which I’ve written an article about on my actual website: http://cowmuffins.net/strawberryjam.html

It’s more directed towards developers and other artists, and is also a bit messy, but it’s still a thing I wrote, so here you go.

Strawberry Jam handles difficulty in a very different fashion to Spring Collab. Like, the variance in difficulty within each lobby is so much greater, to the point where the extremities intermingle with each other a little & even overlap in some cases. The red advanced maps are basically expert and vice versa. I think this might be preferable to SC's problem where every new lobby felt like you were wandering somewhere you shouldn't be, SJ has you enter each new lobby decently well prepared for at least the easier stuff.

This comes with some new problems though, like do you remember that part in the video where I talked about how the community corelates speed tech with difficulty & how this creates artificial segregation of mechanics between lobbies? Well, in SJ, the difference between red advanced and green expert is so minimal that the only concrete difference is tech usage. The expert lobby has a lot of maps that are basically just middle of the road advanced maps with the occasional ultra or dream hyper, and also a couple maps that by all accounts should be in grandmaster but they technically never have you do more than one ultra in a row (which would make it a chained ultra) so it only uses "expert tech" i guess.

This tech-focused difficulty categorization is a years-long uncorrected error on the part of the celeste modding community, honestly. Like they must understand that tech use on its own can't denote difficulty, or else they'd have put summit down-side in the beginner lobby.

Tbf, it's a functional short-hand, especially with so many ppl making maps at the same time. And putting a bunch of tech in one lobby isn't great because players will want to learn the tech before doing stuff in the lobby. I don't really agree with how every tech was handled and I'd prefer if we nudged away from tech = difficulty (like, it's probably worth noting which tech got moved to a different lobby between SC and SJ). But yeah, having more experimentation and discussion about what elements actually make up difficulty definitely wouldn't hurt.