RTABC - CCM 141 Coins in 0 A Presses (6:09.50) Lord SNEK 2026-09-20 | RTABC - CCM 141 Coins in 0 A Presses (6:09.50)
Shooting Star Deathless Gemskip 100% Lord SNEK 2024-12-20 | This took way too long for a lot of stupid reasons. I missed the 2200 berry once despite doing the level dashless (I just dashed under the berry and it left), I died in 2400 somehow, and I didn't even know the 1300 berry was there for a while. This means I've done 10 berries, 11 berries, and 12 berries deathless, for whatever that's worth. Cool mod, hopefully some more videos soon but you never knowI didn't know about the better exit in 1300 until I was watching back this run lmao, nobody told me. At least it's not what I was doing before in 2300 where I uprighted into the cornerjumps on the right. I did know and practice a much safer strat in 1400 but I guess I felt silly this run that's the only time I've ever done that
Tasonleste tech completion Lord SNEK 2024-09-22 | With this, I've now beaten every level in tasonleste besides tas only! This is definitely a challenging level, but it's not that bad once you get the hang of it, and it's very satisfying with how the run ends up feeling. Setups for this level were mostly found by me, with some help from flyingpenguin (who also got the first completion of this level) on the middle section. I believe this level has 16 frame perfects (or less, I kinda just counted in my head real quick). I was originally going to get this back before I got jank, but got discouraged when I got busy with life and other stuff and penguin got it anyway. But now I'm back!I don't think I'm going to do tas only, at least not soon. Even the mashing present in this level (just 3 consecutive cornerjumps) was maddening to work on for me, and tas only is pretty much solely a mash test. I'm not certain I couldn't get it, but I've been trying to work on my mashing for a long time to no avail, so I'm at least skeptical that I could. And, of course, I wouldn't want to.Video is looped 4 times to put it over 1 minute so it doesn't become a short on youtube
Puzzlemod 100% deathless Lord SNEK 2024-08-10 | Kind of cheating to play my own mod like this but whatever. Please ignore how bad I am at my own puzzles it shows that uhhhhhh it shows that the puzzles are complex and interesting and not that idiot monkey brain push blocks around in circle without thinking. Play puzzlemod! lexaloffle.com/bbs/?tid=141339Join the discord! discord.gg/9Dm3NCS
Adelie 100% Deathless (Penguin Saved!) Lord SNEK 2024-08-10 | Not a super difficult run but may as well upload it. I got a run like 10 minutes before this with 9/10 berries because I just didn't get the second berry (one of the easier ones), so that was fun. Aside from that not too much was difficult, most sections got rather easy with setups and the ones that didn't I gamed at etc. Yes I will be doing a lot of easier mods deathless soon lmaoAdelie: h ttps://meepmoop.itch.io/adelieJoin the discord! h ttps://discord.gg/9Dm3NCSYou have to delete the space manually, youtube thinks allowing me to post links would fundamentally compromise the whole video hosting operation they have going on
Perisher 100% Deathless Lord SNEK 2024-08-09 | This was long overdue. Perisher is not a good game to speedrun (or in this case deathless run since I didn't care about speed). It takes forever to beat some relatively simple levels, stupid small mistakes can lose like 5 minutes, and all the difficulty is concentrated at the end. But perisher is a great game to play casually, and being basically the first mod has earned it a place in many community members' hearts, so I think it deserved a quality deathless run. Still, the sheer number of chokes I had after 2700 was totally uncalled for. I first started playing this like 2 years ago or something before getting bored and moving on; this year I think I got past 2700 like 6 times? Maybe more? It was really ridiculous for some relatively not difficult levels. But it's done now! Yay!Perisher: h ttps://managore.itch.io/perisherInterested in celeste classic? Join the discord! h ttps://discord.gg/9Dm3NCSYou have to delete the space yourself because youtube won't let me have links without showing them a government id and that's like comically dystopian
Why Are You Stealing From Ruins? Deathless 100% Lord SNEK 2024-08-09 | An iconic celeste mod that I decided to do deathless recently. Not too crazy of an achievement, but definitely not trivial either! The end of tighter space, exit, and tightest space in particular killed a lot of runs. At the end of the day though, it could've been a lot worse, and I did this without a practice mod or any serious strat finding (I didn't even do the downdiag dash thing in one way climb until somebody else mentioned it because I didn't want to have to remember what the buffered setup was for that and I was doing fine wallsliding). Might try to fill out more of the decently hard mods deathless soon but also might not! The future is always uncertain!!!Video looped 1 time to prevent it from being a not-video
Madeline and her Trusty Stepstool vs Celeste Mountain True Ending Deathless Lord SNEK 2024-07-27 | I figured I should move on to what actually mattersThe beginning of this mod is sort of just ingeste but harder, the ending gets really cool though and has some difficult stuff going on
Tasonleste jank completion Lord SNEK 2024-07-26 | Years ago I released a mod with the express purpose of using only vanilla celeste classic mechanics, yet being difficult to beat using tas tools (thus being a sort of puzzle). Over the last couple months penguin and I have been finding strats and completing levels from that mod, and this is the final one to be beaten. This level wasn't supposed to be the hardest in the original challenge, and there's cheese, but there's just a stupid amount of navigating through really tight spike pits. There are 14 inputs in this level that I'm certain are frame perfect, and there's a lot of additional difficult from stuff that technically isn't frame perfect (if you hold right for too long before the walljump to clip at the end you have to adjust your left press timing with basically no time to react to what's happening visually, for example). It's definitely the hardest 1-screen level I've beaten in this game. It was actually much worse, but I'm also a strat finding god which lead to the strat you see at the end where you get some rather precise rem and can just edgejump and updash to the end. Very happy it's over, although I admit it did start to be a bit fun at the end when I really locked in for the last like hour.This took slightly less time of attempts than all of JANK (the full mod) combined! Of course, it's much much more when you consider the practice and strat finding. I have no idea how much time I've spent on this over the last ~3 days, but it's a ridiculous amount for one level. I really wanted to get this before I go on vacation soon and before penguin gets it, so I've put in way too much time and energy into this despite already being totally burned out/tired. Yesterday I nearly fell asleep while doing practice lmao.I might do tech soonish although probably after the vacation I mentioned earlier, and I want to do a full playthrough someday which implies I'd have to beat tas only, but I'm also not certain I'm physically capable of mashing fast enough for it regardless of how much time I put in. We'll see? Not too worried about it though, as cool as a full completion would be it's enough for now to see every level beaten in isolation, which is kind of a full completion if me and penguin lived together and passed the computer back and forth and had completed stuff in a different order.Video is looped so youtube doesn't turn it into a short, and honestly it's just tradition now so I may as well just do it. Also you probably do need to see it 7 times if you think about it.
Tasonleste Precision Completion Lord SNEK 2024-07-09 | This level used to be much harder until I found this spring bounce setup recently, but it's still rather challenging. I took ~2 hours of runs to get this (after getting a lot of practice preparing for a harder setup), but I was also really tired during 90 of those minutes, I don't think it should've taken me nearly that long. Fundamentally, I was just a choke artist at the ending. Getting past the beginning got rather consistent, getting past the middle was annoying since I always either pressed right early or for too long, but the ending was awful for me. It was somewhat consistent in practice, but I got there like 30some times (I don't know exactly, I lost count) before I got it. I actually got to the clip twice in a row one time and both attempts failed. I think this level is a little bit easier than spikes because of the new setup at the spring. Every input is frame perfect except the walljump after the upright dash, but there just aren't that many inputs total (this level has 11 frame perfects, certainly a challenge but far from the highest I've done for a single-screen level)Two levels remain in tasonleste v1.0, jank and tech. Tech seems likely to be about as hard as this level to me, but jank will be very very difficult. Still, we've found a setup that works for jank, so it will happen, it'll just take some time. Who knows how much; I'm going on vacation soon, but I want to at least get one of the levels first lmao
Tasonleste Spikes Completion Lord SNEK 2024-06-16 | I beat it! By far the hardest single screen challenge in celeste classic right now. ~12 frame perfect inputs and a lot of 2f/frame perfect but repeatable stuff. By far the hardest part of this level is getting through the spike at the top left. That alone requires a good approach despite the tight space and a frame perfect right press for only 1 frame (and then you need to release left on the correct frame right after or it doesn't matter that you got through the spike). The rest of the level is pretty much what it looks like. This was my third run to the bottom block, I lost the first one to a 2f tap instead of a 1f tap, lost the second one because I took a long time so I was trying to take a gif since I figured I would leave the 50s gif range soon and accidentally fell off the edge (no really I lost a pb run to that), and then this is the third one. I think I clutched the ending but I wouldn't call it a total fluke, I'd practiced it a bit and was usually pretty good in practice too. This level was actually a little bit fun to me, after I got the beginning inputs down fine (which took less time than I expected) it was sometimes even a bit pleasant. Still, it was pretty common to have periods of awful consistency or just dying over and over again to the beginning, it's so ridiculously frontloaded that it definitely gets to be monotonous. I'm trying to keep better track of my hours for completions; I think I spent a little under 4 hours in the actual game playing (40ish, break to practice, 2:43, reset because that's just too much time, then this run), maybe an hour in practice, and another hour and change tasing setups and asking penguin about stuff. This is definitely a very challenging level, I'll provide more thoughts on that later if I remember (typing this description before leaving for a while).This is the first level in a mod that was designed as a tasing challenge; spikes, tas only, and 2300 have been beaten rta, but there are still three more to go! I'm not sure how viable some of them are, but I'm hopeful one day we can beat the whole mod rta. Not today though, I have other things planned.
1:18.10 gem 100% Lord SNEK 2024-05-28 | First sub 2 100% pace gem! This was basically everything you could want in a gem, and fortunately it put me on sub 2 pace with a significant amount to spare. I can get a 1:17 in theory but I'm not really worried about that, I need to just keep getting runs like this until one of them gets to summit. Number 1 of who knows how many. Will be practicing postgem more now for sure, postgem strats are so hard.
2:04.17 100% Lord SNEK 2024-05-26 | This run is really good! As a proof of concept. There is so much time to be saved off of this and I am not happy with this as a final run. Still, it's good to have a time I can compare to/have as my pb that isn't some super old 2:08. Timesaves here should be easy to see, I wimped out on 2800+2900, pre-old site was horrible both rngwise and skillwise (my fingers are too tired to play 200 right now lol, I actually stopped doing runs for the night because my consistency dropped to basically 0%, my fingers simply can't move fast enough for that), and 1700 was just the worst. These strats can get me a mid 2:00, give or take a bit if my movement's better or worse than I realize, and a bit less if I incorporate 2900 or 2800. Anyway my brain is too tired to write stuff just look at the run what else do I have to say. I'll update my splits to this run later probably comparing against the 2:08 is demoralizing whenever I miss a berry I continue the run just so if it finishes below a 2:08 I can use those splits instead of keeping these stupid ones. Well no longer! Maybe!
40.37 old site 100% Lord SNEK 2024-05-22 | Pretty sweet run but not unbeatable (sub 40 soon?). Just want to upload this so I can have it on hand whenever, making it public because now other people can have it on hand whenever.No this is not Old Site (the mod), Old Site (the mod), or Old Site (the other mod), this is old site vanilla celeste classic aka 1200m sorry if this caused confusion we really gotta stop releasing mods called simply Old Site. Or not this is actually pretty funny.
1:20.87 100% gem Lord SNEK 2024-05-20 | Playing around with the viability of sub 2 hundo. This run is ~.3 behind the world record (2:01) into 2200m, and I do about the same strats as the world record postgem, with probably a bit lost because my hundo movement is not good. Still, that makes this run seem closer than it is; postgem is INCREDIBLY hard and there are quite a few tricks you have to get right. I will probably either beat this gem time or give up on hundo soon so don't expect this upload to be relevant for long. The biggest timesave here would be good 900 rng (~.4 in my testing, but I also just did the backup slowly lol), but I also have some pretty substantial lapses of movement (notably 100, 1200, 1400, there'll probably be more as I get more comfortable with hundo runs). I will also probably implement more strats if I end up going for sub 2; 1900 clip saves a lot despite being garbage, and 1700 has some really juicy timesave that I just can't find a good setup forSplits are comparing against my 2:08.33 pb, but that run is pretty old and I'm doing strats that are very excessive for just beating that. I mostly just like having splits to compare against, I expect that a deathless run would beat that pb by quite a bit
32.70 old site Lord SNEK 2024-03-30 | Video title. A pretty quick old site, pb by 3 frames. I did this in part because I felt like it was weird I didn't have a super fast old site, in part because I wanted to test out some new stuff in 400/with my setup, and in part because I misremembered being a frame or two off the any% wr with my 3 segment sob (with this improvement it's now 1f off which is annoying lmao). This was never supposed to be a serious attempt at any% (I mean look at me trying to do a run afterwards lmao I was not prepared at all), I just wanted to improve my old site time, and I did that! This run can be improved a fair bit by doing a faster 700 strat, doing 800 better, or just not being super slow in 900 (in my defense this is at least partially rng so I have a scapegoat there). However, I'm probably not going to, at least for a while; if I want to improve my 3 segment sob midgame is now probably my worst segment and I don't plan on doing actual any% runs soonThe splits in this video are, like, totally wrong. Just ignore them. This is part of why I was saying I wanted to test my setup, I didn't notice they were wrong until the old site time wasn't adding up at all.
Falling With Style (Hard Celeste Classic 2 Mod) Lord SNEK 2024-03-13 | This is a completion video of a short mod I made recently. It's only one screen long, but that one screen features several difficult tricks with items and precise tech. Overall, most of the level is very precise and punishing to mistakes, but it's not long enough for me to feel like there's too much real difficulty; I did a lot of practice and expected a longer grind, but I got this run after about an hour of serious attempts, and it didn't feel like much of a fluke. I would recommend playing this, since once you know what you're doing it feels satisfying to pull off correctly and it can help you learn some cc2 tricks if you're just getting into it (I basically play no cc2, my only rta background before this was one laggy run of any% from a while ago and a few ccbtt runs)Video is looped because youtube often tries to make videos under a minute into YouTube Shorts™️ and I would prefer to have them not show up like that
Solanum 59.60 Lord SNEK 2024-02-07 | Glad I got a wr run but I gotta say it wasn't all that great lol. 200 in particular was just really bad, and I was overly safe in 700 and 1100-end just because I wanted a completed run even if I lost frames at the end. I think rta you could totally 58, it would be pretty hard but it's also a really fun mod to speedrun so I could totally see somebody wanting to go for it
Actuate 39 Lord SNEK 2024-02-02 | Yeah I missed a lot of stuff in my previous run lol. This one was a lot better stratwise; movementwise, the beginning was pretty slow, but I think the second half was pretty good (wind chill was rough but honestly this is probably the fastest I've ever done it lmao). The biggest timesave I've really thought about is a cornerjump after the crumble blocks in icicle path, but I decided against it because I wasn't sure it would even save a second barrier (I don't think it would have; this seems to be a pretty high 39). I'm sure you could do 38 or even 37 by cleaning up movement/adding some slightly more ham stuff (there's probably a disgusting clip somewhere in wind chill that saves over a minute), but I don't think I really want to go for it, a high 39 is satisfying to end on and I don't want to have to care about wind chill optimizing unless somebody can find a strat better than the walljump skip I do rn
Actuate 42 Lord SNEK 2024-02-02 | Seems like a low 42 and I found a significant timesave while watching this back so 41 or lower soon if I get time hopefully. Also youtube sucks I had to reupload this like 3 times because of dumb problems
Spiral Completion Lord SNEK 2024-01-22 | Double video so youtube doesn't automatically make it a short. First fully recorded completion of a celeste classic challenge mod made and beaten by flyingpenguin.
S+ Completion Lord SNEK 2023-11-17 | Video loops twice because youtube hates fun and tries to make everything a short :(
Abandoned Normal Route 100% Deathless Lord SNEK 2023-11-05 | Pretty much the run I was hoping for. Mainly just played it because it's first alphabetically lol
Scrolleste Brutal Completion Lord SNEK 2023-10-20 | I think this is the first time somebody's done this maybe, although I wouldn't be surprised at all if that's not true. Lost at least 5 runs to 3000 lol. The entire run is an autoscroller that does not need to be done fast except for:-1600 needs a vaguely fast strat, you can't just take the spring without at least having good movement-2400/2500 needs to be at least decently fast-2600 needs to be *fast* and you need to do the beginning of 2700 well. There's a few frames of leeway here, but just a few frames.This run did all of these sections with way more time to spare than normal, madeline isn't even like that close to the lava. Idk it's just kind of doing 26 levels in a row without messing up, then one big choke point at 2700, then 3 more levels without messing up. I haven't ever really practiced this mod, although I did do completions on the first cart without difficulty selection and the easy hundo run I recently uploaded, so I wasn't going in completely blind. Plus I've tased up until like 2300 that probably counts for something. I'm sure people could do much harder difficulties than easy for hundo, again I didn't practice at all and I have several places here that are faster than my easy run. Might not do it soon though, good luck if somebody else wants to do that
Scrolleste 100% completion Lord SNEK 2023-10-17 | Thought this would be worth uploading. It's mostly just a deathless autoscroller; the rubber banding on the lava is so extreme that it doesn't matter if you lose time as long as it's not like right before or during 500 1900 2900 or 3000, which means it's pretty much just playing vanilla with a few levels that are forced to be ham. That being said, I like deathless runs, so I found this fun, and there were some places where I could do fast/swag stuff risk free that feels fun to hit. Might maybe go for some harder scrolleste completions soon but I think I'd actually practice if I wanted to do those, up until this point I've practiced neither scrolleste savestates nor vanilla levels with the intent of transferring a strat to scrolleste, which makes progress pretty fun but is definitely not viable for an actually hard scrolleste completion
Awake any% 54.16 Lord SNEK 2023-08-26 | Not too bad of a run; I've had visible 17.8xs into trials, but all in all I don't think this was that bad. Also missed a cj in trial 1 but I don't really know what makes me get that and I wasn't worrying about it too much so it was just a lil treat if I happen to get it. 53 is definitely possible with my strats for sure, and I'm sure 52 could be done rta with better strats (like clean clipping in 300 to skip the extra walljump, or clipping in certain places during trials). No idea about lower than that, but people tend to find a way. Pretty happy with this run still; my original goal was to just get a 54 but I kept choking in stupid ways so my movement got better to the point where I almost got a 53 lol
Celeste Tech Training Farewell Lord SNEK 2023-07-13 | Better challenge than ihop ngl. Not my first time beating this
Maude Spoon Ending Lord SNEK 2023-07-01 | Actual medieval torture methodIn Maude, there are 6 endings. You can beat the game with the general route shown here, but casually and without a lot of extra stuff I'll get to later (bad ending), you can exit using hidden blocks in the level with the sign that teleports you to 5 minute timeout (per ending), you can exit under the level with the rodrigo. sign (good ending), and you can exit 800m without hitting a spike (roundelie ending). In the bad ending, there's a sarcastic penguin that makes fun of you for how difficult it must've been to have to go to timeout for a whole entire 5 minutes. If you beat the bad ending deathless, however (including 5 minute timeout, using the strat shown in this video), the penguin will repent and offer to share a spoonful of ice cream with you. This is the 5th ending (ice cream). There is however, a 6th ending, which is what I show in this video. If you read the text in the first level up to a certain point (the o in rock needs to be visible) then leave the text so the rock doesn't fall and kill madeline, that increments a counter 1 time. If the counter is equal to 69 or 420 when you enter 200m, adelie will leave behind a comically oversized spoon when they varnish. After you touch the spoon, madeline will brandish the spoon whenever you crouch. If you then get the ice cream ending and show adelie the comically oversized spoon after they offer you a spoonful of your ice cream, you see the dialogue shown in this video. This means that the route is to enter 100m, read the sign text up until "falling ro", then jump away from the sign so it resets and dash back into it. In my setup, after doing this, I incremented a number in a text file on my phone by 1. Then, the sign will get back up to that point, and you do it again. Repeat this until the phone counter reaches 69, then beat the rest of the mod deathless, including 5 minute timeout which requires 2/3 (depending on how you count them) frame perfect dashes to not die. Note that the "death" in 800m is not actually a death, those spikes are just programmed to send you to the next level if you would otherwise die to them. The random pause was me dropping my phone on my keyboard, which thankfully did not damage the phone because I do all my runs lying down on carpeted ground.00:00 - previous attempt and setup00:33 - run starts 13:30 - last text iteration19:30 - timeout ends
Celeste Classic Any% 1:39.80 Lord SNEK 2023-06-30 | This is a run I've been grinding for for quite some time (it was a stretch goal of mine back when I had like 2+, minute times, although obviously not an immediate concern then lol) and I'm extremely happy to have finally gotten it. At the time I got this, I'm the 5th person to have ever gotten sub 1:40 in celeste classic, and I'm the 3rd fastest person in the world. I probably won't try to beat this; I certainly won't for a while. My only real reason for pushing myself to this time is that I've wanted sub 1:40 for so long and I absolutely hate the idea of ending with a 1:40.xx since it's so close to such an important time barrier. I won't say I can't come back; there are a *lot* of timesaves to be had over this run (900 rng, 1400 was complete garbage, a lot of other small things) and I'm not doing every strat I could be doing. However, if I do come back, it will be far in the future, and with much less intensity than this grind. For now, at least, I'm happy with what I have, and I look forward to being able to work on mods and tases without having to worry about my vanilla pb. The run I compared against was a 1:41.80; this was not my previous pb, it's just a run that I've gotten used to comparing against, so I know how good it is to save certain amounts of time, and it's close enough on most levels to at least not be a terrible comparison. If I ever do go for a lower time, I'll probably use this run's splits, especially because I love comparing against runs with bad 900 rng.
[TAS] Celeste Classic Any% Glitchless 1:44.13 Lord SNEK 2023-06-29 | This tas completes the game glitchless, meaning it does not walljump off of, completely pass through, or stand on spikes.Long ramblings about why I used these rules for "glitchless" and why I think a glitchless category warrants this sort of demonstration: pastebin.com/DCc6bVzG
[TAS] Caroline True Ending - 2:21.93 Lord SNEK 2023-06-24 | Frame advance is heavily encouraged when viewing this tasThis tas aims to beat Caroline's true ending as fast as possible, which involves beating the game with all the cores and then beating 2 secret levels that can be accessed after the normal end of the game. As part of beating the true ending, the tas also completes the normal ending 100% with a time of 2:11.80. The recording was made by simply recording what comes out of the tas tool, which is sometimes not faithful to the game visually and especially audibly (this is why the music at the end is so weird). My setup for recording sound is also not great, and this is the reason for the crackling heard at the beginning as well as the washed out sound throughout. I don't really know how to improve on this right now in an easy way, but I may look into it more in the future. Caroline is a portal demake based off of celeste classic's engine made by meep moop with some levels by myself. You can play it free here: meepmoop.itch.io/caroline. Huge thanks to Gonengazit for developing celia, the pico 8 tas tool, and thanks to Mirkwood for adding in mouse support and player 2 controls that are needed to tas Caroline.
1:04.47 gem Lord SNEK 2023-06-01 | This was a really good run, got basically everything as fast as can be expected. Obviously there's always room for improvement, but ngl if I end up getting a better gem before I get 1:39 I'll be pretty mad lol. It's honestly kinda cracked how good meep and feetballer are, very happy to even be within a second of feetballer's gem pb lol
Fuji any% in 1:02.930 [TAS] Lord SNEK 2023-02-21 | This is a tool assisted speedrun of fuji lexaloffle.com/bbs/?tid=42977 that grabs the gem (which I believe is the fastest route through the game) in under 1:03 made entirely by me. This is an improvement of a bit over 2.5 seconds from rta, and the first tas made for this game. Even though the timesave over rta is relatively low, that's mainly because the rta record for the game is relatively well optimized, and the game is very short. Individual splits for levels are not currently readily available because fuji uses a different timing system than the tas tool is designed to handle. The audio desyncs on dashes are just a product of how the tas tool handles sound.Side note: in the gem room, there's a glitch that saves 2f where you can skip the freeze frames of a dash by grabbing the orb, which sets the freeze timer to another number. This glitch does not exist in vanilla, and it's actually a recently discovered glitch in evercore, the mod base used to make this mod. Without this glitch, this tas may have been a 1:03. All the rest of the mechanics are very similar to vanilla celeste.
Intermediate Lobby Heartside Golden Lord SNEK 2023-02-10 | Came back to mess with this a bit more. Sorry for the frame drops, they will happen again. I had a really fun time trying to trim this video every video trimmer mysteriously just stopped working. ffmpeg just output a 0 byte file and everything I tried online didn't work. After literally like an hour and a half of waiting online video cutter finally responded... with a 1.3 GB video. I was going to upload that as the video since nothing seemed to work, so I waited out the downlaod, and right as soon as I did that, videocandy, which was stuck at 14% for ~2 hours straight, just randomly worked with a normal size video. This is the 1.3 gb one since I already went through all the effort of getting this to work and I'm not entirely sure the other one will work. If anybody knows why this was all so bad, let me know, because I've never experienced anything like this before, and I assume it's not internet because that's mostly fine, but everything fails so surely it's on my end? I don't know it's super weird.This wasn't that bad to get; it took me a while, but only because I'm still trying to avoid practicing very much (I find it more fun to go in with only a little bit of knowledge about what is coming up). Flag 2 was pain and the end just never worked for me, to the point where I was consistently getting to flag 2 by the end. I think for the advanced lobby I might need to actually start practicing goldens, but a radical concept might be to just stop doing goldens and just beat the mod normally so I can move on to another mod. Who knows? Anyway this isn't that cool a thing but I thought it would be an easy upload so I just decided to record as a little progress marker.
Foreverred Trueskip 100% Deathless (First Ever) Lord SNEK 2023-01-29 | This run completes the hardest route in Foreverred 100% with 0 deaths. I have now completed every route through Everred and Foreverred deathless 100%. Until now, nobody had beaten Foreverred Trueskip deathless at all. In fact, all the remotely close runs were by me and also got all the berries. If you want to play Foreverred, you can find it here, (although I'd recommend you don't play Trueskip at first) lexaloffle.com/bbs/?tid=41010.Holy bingle. I've never seen a 1 death run and I never will. I figured I was going to get this at some point, I've been playing well, but I had no idea it would be this quick. Huge huge shoutouts to Gonengazit for the determination setup it's actually so cracked and the old setup was indescribably awful. Even with that, this must be the hardest thing I've done in any video game by far. This took a huge amount of practice, strat finding, and actually doing runs. I have a strat for pretty much every possible backup as well as cues for how to do every part of the run. At some point I might make a writeup in case somebody else wants to go for this; there's a lot of small stuff that helps a bit.I actually like completely zoned out this entire run. I realized between the final spring bounce in dt and the updash that it was the bounce and I needed to go, and I feel like I never really focused on Caroline very well (which isn't weird, it's how I practice, I just expected adrenaline to kick in a bit more). I kind of got to contemplation like "hey that's a new deathless pb" (which it was btw which is insane) and then started focusing a bit more, but it still doesn't feel real. Couldn't believe it when I got the first dash in determination and lost it when I got the last one. This is such a good run and I'm so happy to have gotten it.As far as I know, nobody actually knew until now that the end screen for Trueskip deathless is green (except obviously taco and potentially other people who helped token minimize the mod). Not something I was expecting at all and the green heart really surprised me. This is completely inconsequential and nobody cares but it's interesting to me :DProbably forgetting something else, might update the description later. So so so happy with this run. Shoutouts to Taco360 for the mods, now she has to release Foreverred 2 to keep up with me smh
Foreverred Trueskip 100% in 2 deaths Lord SNEK 2023-01-28 | 1 death in aspiration due to brain fart, 1 death in exasperation due to the level being hard :(. Do not ask me what happened in final climb. I do not know and I do not want to. Deathless soon hopefully but idk 2 deaths is still significantly more than 0. We'll see.
Foreverred trueskip 100% 4 deaths Lord SNEK 2023-01-19 | New minimum deaths wr for trueskip (both any% and hundo)! 3 deaths in contemplation and a death exit in determination. This run is very promising, and it may be the first time I've been deathless from determination, although I"m not entirely sure about that. Hopefully less deaths soon? We'll see.
Celeste Classic 2 any% in 1:21.967 [TAS] Lord SNEK 2022-08-11 | The current tas tool is not really a good model of the game, it rounds numbers differently and there may be other syncing errors since ikuyo could not get the inputs to line up in the actual game. Still, this is the most accurate tas tool we have, and for the most part it's probably fairly accurate (most of this stuff works rta to my knowledge). This is the first ever sub 1:22 tas.00:11.57 (346)00:17.83 (187)00:28.83 (329)00:52.67 (714)01:02.50 (294)01:12.07 (286)01:21.97 (296)lol snek what a bad mod when are you going to play a REAL game????
Hallowneste 38 Lord SNEK 2022-07-06 | The end of the run continues to be my weakness. Some nice highlights though, and if you ignore the awful menuing at the shop/forgetting to do a dash boost at the end, it's a pretty good run. 37 is my end goal.
Hallowneste 41 gem% Lord SNEK 2022-07-05 | I like running this mod, runs go by fast and the movement feels pretty nice. Drim's 42 is pretty good, especially considering the routing and stuff that went into it. You can save quite a bit over this by having better ending movement, I largely didn't practice that and kind of just went with whatever felt tolerable, using timesave from earlier as a buffer. I think that and slightly better movement in places should make a sub 40 possible gem%. Also, what if we killed some aspids going to the left? It would make the run harder, but especially for the first aspid in the second room (that I killed with an upright dash), it seems very likely to be faster with good movement, since my current route loses a lot of time going to the right, but it could be fairly quick going to the left.I just learned you can't put angled brackets in yt descriptions (less than/greater than signs, I can't put one as an example), what's up with that lol
Fault 45 Lord SNEK 2022-07-02 | Definitely a lot more time to save, I'm pretty sure sub 40 is possible. My movement isn't great, this was just me messing around for a bit since I saw wr was pretty doable. Shoutouts to gonengazit for helping me upload this to youtube lol
Celeste Classic 100m 0 dashes Lord SNEK 2022-05-30 | Timestamps:0:00-1:15 Setting up the wait on different windows1:15-17:41 Waiting for the dash effect time17:41-18:20 Practicing the walljump to get muscle memory18:20-19:10 Gaming!This is a video of me completing 100m in real time without dashing. The setup/trick itself are kind of weird, so I'll do my best to explain how my setup works. When you dash in celeste classic, a variable called dash_effect_time gets set to 10. Every frame, it decreases. If it's above 0, the game thinks you're dashing, and certain interactions will work like you're in the middle of a dash. However, this variable always decrements by 1 every frame, even if it's below 0, and in pico 8 variables over/underflow if they go beyond +-32767. This means that, if you wait for around 18 minutes and 12 seconds without dashing, dying, or entering a new level, you can activate some of the effects of dashing without dashing. This doesn't change your speed or anything commonly useful like that, but, importantly for this level, it does allow you to break berry blocks you're touching.Because it takes 18 minutes to break the berry block, it's very tedious to do this rta. Fortunately, there is a slightly better way to do this than just waiting 18 minutes for every attempt. It's possible to have multiple windows open all running apps, so you can run multiple versions of pico 8 at the same time. I chose to do 3, because I worried any more would be too much for my computer, and any lag at all is bad because it desyncs the game with real time. This accounts for some of the weird alt tabbing that you can see I did at the beginning between celeste1 celeste2 and celeste3 (untilted document leak :o:o:o). Theoretically, with a much better computer, you could absolutely have a better setup; in fact, with a really really powerful computer you could probably just continuously attempt this trick with like 70 windows open. I don't have that luxury, and so I had to resort to playing 3D World until the timer reached 17 minutes.This strat was originally floated around by FlyingPenguin223 as an idea back on August 23rd 2020, but myself and others deemed it unviable because it didn't seem like you could do the second half of the level. On June 1st 2021, Sparky99 said it was probably possible, and when I tried to prove it wasn't possible I discovered the second half of the level was doable. A few days later, I figured out that you can get from the berry block to the middle bump, by landing on the ground by the block with VERY specific rem. Two days later, when tasing the entire level, I realized you could make the berry block bounce with less specific rem if you had a better x position when hitting the berry block. This final tas was then verified on a tas cart that was hacked to emulate underflow, since the tas tool doesn't normally underflow like pico 8 does. At the time, it seemed pretty unviable for rta runs, and the tas did an extremely long setup because of how precise I thought the rem had to be. Over time, I discovered a great setup for the ending walljumps and thought of the multitab strat, and became convinced 100m dashless could be done rta with a better setup. 4 days ago, I made a few different setups that made this strat definitely rta viable, but extremely hard. FlyingPenguin223 started to practice the strat, and had some success, but then gonengazit labbed out the strat shown in this video. Compared to the previous strat, which had 7 frame perfect inputs including a jump that you have to do solely off of your timer, this strategy was much easier, and it has only 2 frame perfect inputs. This run was gotten a few hours later using the new strategy. It's hard to emphasize enough how helpful gonen's setup is; with the best strategies I had come up with, this level would have been a painful long grind even for people with good enough computers to run 30 tabs at once.Shoutouts to gonengazit for the setup and FlyingPenguin for motivating me to focus on mindashes tricks rta, I definitely wouldn't have been able to do this without them.
Farland Normal Route any% 1:44.333 Summit timing [TAS] Lord SNEK 2022-04-18 | "Summit timing" is not an official timing method, but rather a placeholder; Farland isn't entirely supported by the tas tool, which means we can't see the actual final time (when Madeline touches the flag) to the decimal place, which makes it difficult to compare tases. To compensate for this, I just used the final time I could find, which was the time Madeline enters Summit. Hopefully these errors are fixed at some point so we can use the final in game time with decimal point precision. The title says "normal route", and not just any%, because there is a secret path that allows you to beat the game faster, which would be real "any%" for this game.This is the first tas of Farland uploaded to youtube, but a tas of this game was created for a tas comp almost a year ago, and that tas was recorded even though it never got officially uploaded anywhere. This video is an updated version of that tas, with 51 frames, or 1.700 seconds, saved since the original tas comp. While most of the frames saved were just a result of better optimization, it's worth noting that many of these frames are a result of a physics modification discovered by bacon_good that makes your terminal velocity faster if you are holding down. This improves almost every section where Madeline is falling, and makes up a significant number of the 51 frames saved. A full breakdown of splits and time saved:00:05.73333333333333 (142) (-2f)00:09.36666666666667 (81)00:12.8 (76)00:15.46666666666667 (53)00:17.3 (23)00:20.36666666666667 (61) (-2f)00:25.23333333333333 (120)00:27.56666666666667 (42)00:29.93333333333333 (46) (-1f)00:32.63333333333333 (57)00:35.066666666666667 (46)00:38.83333333333333 (81) (-4f)00:44.16666666666667 (134) (-8f)00:47.53333333333333 (69) (-2f)00:51.33333333333333 (89) (-2f)00:55.36666666666667 (89) (-5f)00:59.4 (89)01:04.46666666666667 (124) (-2f)01:07.66666666666667 (72)01:11.13333333333333 (72) (-6f)01:16.46666666666667 (128)01:20.9 (107) (-3f, although 1f was saved right after the original tas comp, not in the recent tas update)01:23.23333333333333 (44)01:28.36666666666667 (122) (-9f)01:30.7 (46)01:34.46666666666667 (85)01:38.36666666666667 (86) (-2f)01:44.33333333333333 (147) (-3f)The weird audio and color palette is due to tas tool issues. The audio for one area never stops playing once it starts. This is a problem in most games, but is especially apparent in this mod because the music changes so many times. The weird visuals are because the default palette is annoying to actually tas the game with, because it is hard to see the input display.
Terra Australis Any% 1:28.133 {TAS} Lord SNEK 2022-03-19 | Recently I improved most of the levels in this game (every level except for 200, 900, and 1400 was tased in the last two months). Since the mod has some sick strats, and some of the recent timesaves were huge (1300m alone was 3.667 seconds faster than the previous tas!), I decided to upload the tas. This recording saves 9.567 seconds, or 287 frames, over the fastest recorded tas on the pico arcade channel.I don't usually upload tas videos, since the celeste classic community normally uploads all tases through pico arcade (youtube.com/channel/UCa4880BXY1H2Jka3sD2iV7g). However, the channel is pretty backed up, with farland and vanilla 100% waiting for over 3 months for an upload, despite having a video recorded. Because of this, I decided it would be best to try recording a tas myself, to see how the quality comes out and to get the video out quicker. I may unlist this later if pico arcade is able to upload it on their channel. If not, I might record other tases soon, such as farland (but an updated version since something big was discovered), or celeste downhill (which has over a second saved since the previous tas, despite being a relatively simple mod to optimize).My recording here was literally just a screen record, and the audio didn't come out amazingly, but it was still a lot better than I expected. In future, I might try a different recording method, or recording video and audio separately and editing them together.00:07.6 (200)00:11.73333333333333 (92)00:07.6 (200)00:11.73333333333333 (92)00:17.96666666666667 (159)00:22.86666666666667 (103)00:27.7 (109)00:32.13333333333333 (107)00:36.3 (97)00:41.46666666666667 (128)00:44.066666666666667 (54)00:53.43333333333333 (235)01:00.16666666666667 (170)01:11.46666666666667 (293)01:19.066666666666667 (200)01:22.46666666666667 (72)01:28.13333333333333 (142)
Foreverred v2 trueskip 100% 4:18 Lord SNEK 2022-03-17 | 11 left to go. Maybe some day I will have coordination.
Everred gemskip any% 1:22 (World Record) Lord SNEK 2022-02-27 | Obviously there was a death in 600. Beginnings was not the best (I can clip the platform consistently if I get the updash through the gem at a good time, which I didn't this run), as well as a missed cornerjump in 100 and an intentionally slow but safe strat in 1500. If you cleaned all that up, you could get a 1:19, or maybe a 1:18 if you're good. Still, I'm pretty happy with this for the time being; it's hard to get a clean run with all these strats and I already got a deathless run with slower strats just to get the golden reddish.
Everred normal route any% 1:17 (World Record) Lord SNEK 2022-02-22 | Pretty decent run. 100 I missed a cornerjump, which loses like half a second I think, and at summit I missed the heart and lost a few frames. 1:16 is possible and will someday happen, but 1:17 felt like just getting a good run with hard strats; 1:16 would require either optimizing movement a large amount or adding strats I haven't looked into a lot.
Foreverred 2 demo 37 Lord SNEK 2022-02-20 | Pretty good run, looks like a pretty high decimal too. You can get 5s leaving the first room relatively consistently, but I was slow at mashing here. You can also get a cornerjump at the top of the second room, which is pretty fast, and I could've cut the last two rooms a lot closer if I wasn't scared/hadn't practiced them since they're after the hardest level. I think altogether that plus actually routing climbjumps can get you a 36, but you'd need like everything as well as probably pretty good movement, and I'd rather quit while I'm ahead before the grind gets too grindy.
31.27 midgame Lord SNEK 2021-12-22 | 2000 was really really bad and I'm still not doing 1300 because no. I could probably 30 but at least for now I don't feel that motivated to care.
Simplere - 1:48 Lord SNEK 2021-12-21 | Sub 1:50 with simple strats. I would probably recommend going for some additional strats such as 1900 clip, going for 900 when you get good rng (I didn't here to show you don't need to rely on good rng), left side 400, etc, but this is supposed to provide an easy blueprint which you can build off of. Also, it's supposed to make people realize you totally don't need fast 1000 or 1600 strats to get sub 1:50, since I see that misapprehension a lot.
