Martini!
Underground Fun 1:03.38
updated
- Hello by OMFG
- In the Halls of the Usurper from Shovel Knight
- Main Theme from Shovel Knight
Shouldn't have taken over 10 months, but that's still about two years less time than Pied has taken for the 15.0x no WPR. (In all seriousness 15.0x no WPR is the hardest shit I've ever done, so I wish him the best of luck.)
Anyways this run was very solid, with single frame losses scattered throughout. Can see a low 15.7x being done with the roll jump start which the TAS uses.
Also attached 2-1 and 5-6 VWRs. 2-1 was a little sloppy, I wanted 11.43 or better but alas. The 5-6 VWR is over a year and a half old at this point and is similar in quality to my 22.46, with most of the timeloss occuring at the final lever. Nonetheless these invasions don't even remotely compare to the absolute cinema that is 2-2 virgin.
4-4 in 17.71
5th WW. This was my first sub 12 last room entry, which I finished with my fastest ever last room! Awesome run all round with the exception of second room which was a little unoptimal.
4-1 in 26.28
Was trying to copy Spartacus' middle rocket route but couldn't get it quite right, so I lost some time there. Still a very good run but I definitely want to optimise it some more. Sub 26 is still out of reach though without new tech.
3-2 in 22.43
Tied 5th WW (counting Hydra's and Frozen's no video proof times). Just a frame better over my previous best. I was grinding for WR back when it was 22.31 by Darth, and while I had some fantastic paces entering the 2nd zetu boost I was just not able to finish any of them off :(
Original film is A Clockwork Orange by Stanley Kubrick
Seed - AB 25 8F 80
It was recently found that Yuzu training room allows for manual seed setting through cheat engine, meaning that we can make TAS runs of past layouts! This was an obvious layout to try out as it is the fastest layout ever due to the extremely rare occurrence of two double bumper bricks (the double bumper brick saving two frames by pushing you above the speed limit for a brief period). This is the only layout of its kind to be found (or at least have its seed recorded).
The first run is done solo, the second is done with a zip. The first bumper GP proved annoying to optimise for both runs (hence the thumbnail). Neither myself, Enigma or Makione could make this bumper GP seamless (like in the second bumper GP), therefore a frame is unfortunately wasted on the GP. There may be an alternative setup that would allow for this frame to be saved. Also, I found in Kruchy's 5.13 pit speed with zip, Globox was able to turn left at 0.01 as opposed to 0.03. I was not able to replicate this.
- 009 Sound System Dreamscape by Trance
- Penance from The Binding of Isaac
I've wanted to make a video for a while now of playing through invasions with a handcam (especially with keyboard). I chose to do the two big swimming invasions first, although 5-6 is definitely the main event here considering just how insanely effective d-pad movement is on this invasion (and me being the only one to really take advantage of it). Both runs are solid but improvable, although to my credit I had a tripod covering my vision! If y'all like this stuff I'll definitely do some other invasions.
Virgin in 15.63
Some very slight strat variations compared to my 15.86 and Pied's 15.83 (I will beat it eventually).
D-pad only in 14.73
The main star of the show! I made a 14.99 TAS a few months ago using strategies far more similar to my 15.08. A PR boost at the start and the up kick ybys at the end (even when not optimised as well due to not using stick) save around 0.1 each. Maybe this shows that sub 15 can be done RTA...
Standard rules in 13.93
Some very slight strat variations compared to Galaxiee's 13.99, but this run is really still all him hahahaha
- City of Tears from Hollow Knight
- Luma Pools from Ori and the Will of the Wisps
0:03: 3-2 Virgin WR in 23.59
Solid run - around high 22.4x/low 22.5x quality execution. Pretty sure it's faster to go through the tunnel even without brakeboosting, but I couldn't be bothered.
0:30: 3-2 Virgin without mini boosting in 26.46
I personally believe that any form of mini boosting, such as double attack boosting or zetu boosting, should be banned for virgin runs. This is more a demonstration as to what a run without these boosts would look like. Interestingly it is faster to skip the first trumpet and play the whole invasion as big Rayman!
1:01: 3-3 in 24.73
Really solid run without the opening bonk, which I tried for but unfortunately could never finish a run with :(
1:30: 1-4 in 23.86
Got this for KOTH. Very similar to my 23.99 from way back, with a slight strat change at the end.
Interestingly this run is slower than my 11.36 FWR, at least when comparing the shadow placements of the two characters at the end. Although for some reason rolljumping to the end flag activates it earlier, and it was enough of a difference here to save a frame. So either your shadow shifts slightly back when in a rolljump animation (and I'm an idiot for doing the comparison wrong...) or Ubisoft are dogshit at coding their games.
First WR I've set in over 1000 days that wasn't 2-1 or 5-6, kinda surreal ngl.
Anyways this run didn't feel that great (edit: this was maybe the fastest pace entering the skidboost ever LMAO), especially near the end where I hit the slanted ceiling. I'll try and improve to a 22.8x. Will be interesting to see who fucks this invasion over first with the WPR (my bet is on Unaware).
No Javi, no Shisui, Virgin in 12.01:
Couldn't find an additional frame to squeeze out unfortunately. If there is one, it's most likely at the start.
No Javi, no Shisui in 11.83:
WPR (OMG MARTIN DID A WPR!!!!!!) is normally slower, but here it was faster as I did not need to hover to any shields to set it up.
Frozen strat in 11.43:
Could most likely save a frame at the spinjump.
Virgin in 11.39:
Nothing to comment here.
Standard Shisui strat in 11.33:
Same run as my 'human theory' run from a few months ago. Couldn't find any new optimisations.
Galaxiee strat in 11.21:
Teensy was used here, as for some reason with other characters the fireball patterns are different and block your path. I've gotten a faster Javi start than this once, but even when copying the inputs I couldn't get it to work :(
Lag Shisui strat in 11.11:
Another reupload from a few months ago. Again, couldn't find any new optimisations.
It has been a long time since I've made a huge compilation archiving my whole career (the last one I made like this was at the turn of the decade). While I rarely upload anymore, I still have lots to show from the last five years that warrants a new, and in my opinion my best, compilation! Hope you all enjoy :)
Kitchen Pipe in 30.65 youtube.com/watch?v=kU2eGOJNm7s
Finally starting to get this nonsense optimised. Can still be improved a bit more though!
Leaky Pipes in 49.33 youtube.com/watch?v=JglZh4PBU6Q
Amazing start, although I've had much better speed conservation in the second half. Sub 49 is most likely possible.
1-5 in 26.71
Good run using d-pad friendly strats. Will aim for a sub 26.5 in the future.
2-4 in 18.56
New PB by over a second using the wind boost! Pretty solid run, will leave it.
4-4 in 17.93 (splits 5.43/9.53/12.11/17.93)
Very solid run, 7th WW. Might return to this in the future to grind for a 17.6x/3rd WW, although it would require a more optimised beginning.
4-6 in 26.58
Still very improvable. I'll come back to get the sub 26.5/low 26 in the near future, but I'm a bit sick of this invasion at the moment.
I reacted the exact same way as in my 11.39 :sob:
I doubt anyone will beat this using these strats. Comparing this to my 11.33 human theory TAS, a frame can be saved at Shisui strat (a frame that has never been saved in any top runs), or by going for the YBY tunnel vault into water jump. An 11.34 would be the god run and most likely unbeatable for years.
Music - Relaxing Within the Campground Lodge from Psychonauts
2-2 virgin in 15.86:
2-2 is definitely the best invasion to play virgin as even without advanced movement it's still highly technical and challenging. Really solid run overall with a slight execution fumble at the first bumper GP - will come back to this if Pied beats me again...
3-2 in 22.44:
Got this during the KOTH event. Fantastic run although I unfortunately was playing against Bugs who got 22.38 :-( It would have been tied WR if I had set it just a few days before! Might come back to this although I haven't gotten close to this pace since...
Shoutouts to Galaxiee whose 11.15 script I worked off.
A faster Javi strat and reverting back to the standard ending saved two frames, which is what both Galaxiee and myself had theorised. I paused the emulator at 3.33 for a few seconds to create 'lag' that prevents the imps below from shooting. This is technically possible on PC but isn't allowed in 'official' world records. The current best TAS time without this exploit is 11.28.
- Luma Pools from Ori and the Will of the Wisps
- Sync from Fez
'Human theory' refers to a run that is theoretically possible for a human to perform. Of course it's debatable what is and isn't 'humanly possible'. Galaxiee has produced an 11.28 and 11.29 TAS with a more advanced middle section that is 'technically' possible to perform, but I personally don't think will ever be implemented in a complete run at WR pace so therefore I don't deem it 'human theory'. This TAS was made mainly to see what the fastest time was with the strategies currently used in the WR (although the ending is different, but still entirely possible). I didn't experiment with hover skip after Shisui strat (basically using an attack or attack + hover to clip onto the platform) mainly due to laziness, so I might return to this TAS in the future to see if it's viable in any way, but I doubt it is due to having no success with it in the past without Globox's chunky attack hitbox.
I also attached three other 2-1 runs (RTA this time) that I got when grinding for my 11.38. The first is 'Frozen only' (Frozen strat is a variant of Shisui strat involving a small jump entry), 'Shisui only', and 'Javi + Frozen'. All three runs are records and have small room for improvement.
I also attached the 3-1 VWR I got during the last KOTH competition, which I never made public. This run is great and I have no plans to improve it.
Sorry Tenollie :'(
Finally incoporated the ending water jump! The water jump boosts you as high as 23m/s (although your speed gets reduced straight back down to 20m/s). The spindash exiting the tunnel, first used by Unaware, puts you at 20m/s, compared to the 19.73m/s from rolling off. All this allowed for a frame or two of timesave. Timeloss in this run came from the hover after Shisui strat, and not framespinning on the hill.
4-1 in 26.39
Huge improvement from the 26.99 I uploaded last week. Still wanna try improving this a bit more!
1-4 in 25.56 (VWR)
Good run although I botched the end.
3-1 in 14.51 (tied VWR)
Tied Bugs. Really solid run all round!
Pit distance in 27.32km
I'm (very slowly) working towards 10km all distance/20km average. Got this a few weeks ago and is easily the best distance score I've done! Although not anywhere near as impressive as my speed records...
4-6 (achieved around August 2020)
Definitely the easiest of the four invasions to get under 27 seconds on. Well executed run with basic strategies.
4-2 (achieved May 2022)
I was considering grinding for the sub 27 back in 2019, but I lost interest when System beat me to it. I didn’t go for the yby up before the chain (it’s way harder without stick angle), so it was important I got all the small optimisations right. Really solid run overall.
1-5 (achieved on the 22nd)
The biggest issue I faced here was the top route. For the longest time I attacked to clear the gap, losing 0.4 seconds to the angled brakeboost. Galaxiee recommended I try a PR boost on the top path, which allowed me to skip the attack and make up for that timeloss. I played as Globox, opting for the faster and more consistent PR boost, but using the slower but more consistent groundpound brakeboost ending. Kinda sloppy execution overall with some missed RCBAs.
4-1 (achieved on the 26th)
This was the most daunting invasion of the four, and I heavily considered playing with controller. I’m glad I continued with keyboard, because it turns out there was far more timesave in my previous PB of 27.54 than I thought. The day before I achieved 27.03. When I woke up, MikoX posted a 26.93 with a more optimal rocket path. I achieved the sub easily following this route. Solid run up until rocket skip, where after that was weak. I’ve already improved to 26.86, and want to improve it further! So stay tuned for that…
Kitchen Pipe - youtube.com/watch?v=DuHWmTpBPtU
Got this run ages ago but never made it public. Still convinced I can keep more speed after the fourth straight, but this level is inconsistent so I'm content with this time.
The Slide of your Life - youtube.com/watch?v=81iAxWJiAbU
Much better speed maintenance after the 30 second mark. At 47 seconds I tap back by accident (you can hear the input in the link above), but I don't think this made any difference.
- A Yearning for the Sky from Ori and the Will of the Wisps
- Desperate Romantics from LittleBigPlanet
These are all the runs I submitted to this season's KOTH tournament!
Mystical Munkeys ILs - I wanted to improve this by 1 - 2 seconds but this was still good enough for 5th! More than enough to qualify for the next round.
When Toads Fly Virgin - Very solid run using an inconsistent strategy. Came in 4th.
Snakes on a Cake - Got put against Hydra so I wasn't hopeful. Tied my PB with a pretty decent run using 2020 strats, but wasn't motivated to continue grinding.
Finally got the 11.3x! Still improvable though, maybe now it's time I implement the water jump...
This run was insane except for the last jump which I barely undershot, meaning I had to attack at the finish line, costing two frames. Disappointing, but overall I'm very happy that one of the few invasions I can still compete on is getting some much needed attention!
Much better first half and really good speed conservation in the second half, very happy with this run overall
Underground Fun - youtu.be/W_gMVx8c-qU
New strategy for the spiral is a little reckless but I think it's faster?????
The Slide of your Life - youtu.be/PtGrKPDfmj0
Ending was much better, although I still feel like I'm missing something...
Solid run with a great speed boost on the fourth straight, although execution overall can be improved
Leaky Pipes - youtu.be/4iSAZD8j_ng
Vey solid run although I'm not too sure if my movement in the first half is fully optimal
Underground Fun - youtu.be/SOKOBzM5JTk
Another very solid run although I most likely can maintain more speed entering the final section
The Slide of your Life - youtu.be/shgMK2P3tbo
Yet again a very solid run, but with room for improvement across the board
This run had an amazing start, although I bled time due to slow execution from book split and onwards. The only real mistake was at 11:12 where I wasted 5 seconds trying to defend myself from the shrimp...
Will return to this if this category receives more attention, or a new strategy is found that would make sub 16 a viable grind.
Will eventually go for sub 14:50, but I'm happy with this for now!
Decent run. Bit peeved at how close this was to being sub 15:10, but it's still good enough for 3rd place! My goal for this category is a sub 15 which still requires some work, mostly at salad.
I also turned off game music because I had a stream playing in the background and it would have been too distracting.
I spent about two hours total on these two runs. 4-5 is improvable, although I'm happy with the sub 25. 5-3 was really good, definitely better than I expected to get. While sub 22 is possible for me I'm still very happy with this time!
I also attached an IL run of O-1-3 that I achieved a few weeks ago. Compared to my previous O-1-3 run from two years ago, I used an up kick yby up into a ledge grab in the 1st room, and the water jump in the 2nd room.
First run was pretty good! Big timesave at the beginning by using a double WMB (something I should have been doing already but whatever). Execution isn't as good as in my 22.66. I'll hopefully have some time to improve this some more, but I'm still very happy to have gotten the 22.5x!
Second run was solid. I wanted to beat my legendary 22.96 without WMB, and this run did it :) Still improvable but I'm content with this.
22.66 - Great run. Will be very hard to improve, although a low 22.6x/high 22.5x should be possible!
22.99 - Good run, but still unoptimised. 22.8x is possible but I'm still happy with this :)
Run was sketchy but still pretty good. Will improve to a 22.6x in the near future!
Also attached two invasion PBs from a few weeks ago. Both good runs but still very much improvable!


