Jellopuff | [MKW TAS] Mushroom Gorge (Ultra) - 18.613 (+3.432 Flap) [OUTDATED] @Jellopuff_TAS | Uploaded 5 years ago | Updated 2 hours ago
This is a tool-assisted speedrun. This time will not count on any real-time rankings. Done on Dolphin 5.0 with savestates, slowdown, frame advance, and RAM watch.
TAS BKT Cut: -0.847
Previous BKT: youtube.com/watch?v=TSC7rIfKEgg (by myself)
TAS BKT Flap Cut: -0.412
Previous BKT: youtube.com/watch?v=Bvo-lyKkIuo (by Estaloy)
This run has improvements on all 3 laps. Lap 1 is about .3 faster, lap 2 is about .1 faster, and lap 3 is about .4 faster. I still think that this time can be lowered, so don't think I'm done with the track just yet. With new findings, the sub 18 might be possible, but I'm going to have to work with Estaloy for that. I guess we'll see how far we can push this track in due time.
This is a tool-assisted speedrun. This time will not count on any real-time rankings. Done on Dolphin 5.0 with savestates, slowdown, frame advance, and RAM watch.
TAS BKT Cut: -0.847
Previous BKT: youtube.com/watch?v=TSC7rIfKEgg (by myself)
TAS BKT Flap Cut: -0.412
Previous BKT: youtube.com/watch?v=Bvo-lyKkIuo (by Estaloy)
This run has improvements on all 3 laps. Lap 1 is about .3 faster, lap 2 is about .1 faster, and lap 3 is about .4 faster. I still think that this time can be lowered, so don't think I'm done with the track just yet. With new findings, the sub 18 might be possible, but I'm going to have to work with Estaloy for that. I guess we'll see how far we can push this track in due time.
![[MKW TAS] Mushroom Gorge (Ultra) - 18.613 (+3.432 Flap) [OUTDATED]](https://i.ytimg.com/vi/B0lbGJP2jUc/hqdefault.jpg "[MKW TAS] Mushroom Gorge (Ultra) - 18.613 (+3.432 Flap) [OUTDATED]
This is a tool-assisted speedrun. This time will not count on any real-time rankings. Done on Dolphin 5.0 with savestates, slowdown, frame advance, and RAM watch.
TAS BKT Cut: -0.847
Previous BKT: https://www.youtube.com/watch?v=TSC7rIfKEgg (by myself)
TAS BKT Flap Cut: -0.412
Previous BKT: https://www.youtube.com/watch?v=Bvo-lyKkIuo (by Estaloy)
This run has improvements on all 3 laps. Lap 1 is about .3 faster, lap 2 is about .1 faster, and lap 3 is about .4 faster. I still think that this time can be lowered, so dont think Im done with the track just yet. With new findings, the sub 18 might be possible, but Im going to have to work with Estaloy for that. I guess well see how far we can push this track in due time.")
![[MKW CT] Subspace Factory (Shortcut) - 2:14.764 - Jellopuff (3rd Worldwide)](https://i.ytimg.com/vi/BC-e-5X6oD4/hqdefault.jpg "[MKW CT] Subspace Factory (Shortcut) - 2:14.764 - Jellopuff (3rd Worldwide)
Would you look at that? An MKW video that isnt recorded with an iPad. Nice! (For real though, credit goes to Jordan for capturing this for me)
Date Set - 20 March 2018
Player - Jellopuff
Country - United States
Character - Funky Kong on Flame Runner
Controller - Wiimote / Nunchuk
Total Time - 2:14.764
Lap 1 | 45.077
Lap 2 | 44.749
Lap 3 | 44.938
Ill be honest, I have no clue where this came from. Over the past 2 days (as of the upload date), I shaved a little over a second and a half off my personal best. This wont be my last go at this track, but its really been all Ive been tting for the past week, and Ive reached my personal goal of a 2:14.8xx (which actually makes this run the fastest time in North America), so Im probably going to be tting a different track for a while (most likely Sunset Ridge Glitch).
About the run itself, its pretty good, but definitely improvable. Aside from not tricking into the cannon on lap 3, I didnt really make any huge mistakes. Of all my pbs, this run is probably the best in terms of the shortcut entry across all 3 laps.
[UPDATE: 3/27/18] My pb has been further decreased to 2:14.549 (2nd WW)")
![[MKW TAS] Moo Moo Meadows Flap (Ultra) - 9.207](https://i.ytimg.com/vi/BQ8knpxt_Vc/hqdefault.jpg "[MKW TAS] Moo Moo Meadows Flap (Ultra) - 9.207
This is a tool-assisted speedrun. This time will not count on any real-time rankings. Done on Dolphin 5.0 with savestates, slowdown, frame advance, TAS code, and Info Display (created by SwareJonge). Input Display Overlay created by WhatIsLoaf.
Disclaimer: This run takes advantage of the analog inputs of the Wii Wheel. In order for this run to be performed in real-time in a playback, pause buffering would potentially need to be used.
TAS Flap BKT Cut: -8.521
Previous BKT: https://www.youtube.com/watch?v=A_38TZSWDDk (by Ejay)
A little over a week ago, Ejay proved the newest and most elusive ultra shortcut in the game, the Moo Moo Meadows ultra, to be possible. If you havent already, go give him some love for his discovery. He really deserves it :). But anyway, before the ultra shortcut was formally released, I had been given some insider info that got me theorizing as to how the time could be brought down from the low 17 Ejay got in his demonstration. This run is the result of that theorization.
Before we begin, lets make something clear: yes, this run is an official best known time. For whatever reason a lot of people seem to forget that Flaps are considered official categories, and so whenever an ultra with a long setup shows up, they immediately think it doesnt save time. This track is by far the most extreme example that we currently have in the game (though it may not last too long). The idea for this lap is to perform the shroomless out of bounds entry the same way Ejay did, but after getting out of bounds, we drive an entire lap without going back in bounds. This is where 95% of the time for this strat execution comes from since we cant use any boosts while on the heavy offroad. However, because we dont care about the lap 1 time, how long we spend isnt really important.
After driving out of bounds for about 5 minutes, the actual flap starts. The main difference from the original discovery of the ultra, as you can probably tell, is that we perform the outside drift clip (or ODC) before the flap begins. This section is a lot harder than it looks since the wall becomes double sided too early for us to clip using the same method Ejay used. If we used that method, wed never enter the finish line and the lap wouldnt begin. To circumvent this, I had to figure out some weird shroomless momentum nonsense that would allow me to initiate the shroom while already having momentum directed towards the wall. The space we have to get the clip without bonking early, missing the wall, or being respawned is quite small, so I was quite surprised that this ended up working at all. Upon clipping, we steer the vehicle midair to start the lap as late as possible then perform an OOB bounce to land on the OOB seam. Notably, we land on the seam far enough forward that we dont need to back up to get enough room to perform the final part of the ultra. We can simply turn around and shroom as soon as we turn far enough to the left. The final bit utilizes a technique that weve coined the seam swing. By landing on the OOB seam as low as we can while performing a right drift, we can land back on the seam, start a drift to the right (which sends us to the left due to outside drift bike physics), and then immediately fall off the seam using precise control stick movements. The end result is a redirection of our speed while underneath the floor. If done properly, we are able to barely reach the final key checkpoint without having to delay our acceleration with a hop prior to the seam swing. We respawn right before the finish line and do a small wiggle to shave one final millisecond off the run, nearly cutting the flap time for this ultra in half.
Special thanks to Ejay for telling me about the ultra being confirmed early as well as for helping me understand what I was doing wrong when attempting the shroomless clip! Thanks for all the help youve been to me, and congrats again on discovering this behemoth of an ultra!
Music: Mt. Chill - Thegur90
MKW TASing Discord Server: https://discord.gg/EPD9yCu
MKW TAS BKT Website: https://mkwtas.com/")
![[MKW TAS] Grumble Volcano Flap (Ultra) - 1.441 by Jellopuff and Ejay](https://i.ytimg.com/vi/BRyQF7_C5x0/hqdefault.jpg "[MKW TAS] Grumble Volcano Flap (Ultra) - 1.441 by Jellopuff and Ejay
This is a tool-assisted speedrun. This time will not count on any real-time rankings. Done on Dolphin 5.0 with savestates, slowdown, frame advance, TAS code, Info Display (created by SwareJonge), and TASToolkit Lua Scripts (created by LuigiM). Input Display overlay created by WhatIsLoaf.
TAS Flap BKT Cut: -0.025
Previous BKT: https://www.youtube.com/watch?v=hH9cdEMP9cw (by Ejay and myself)
Ejays channel: @EjayB
Soooooo this took longer to upload than I expected. Sorry about that. This is the final of four flaps I completed at the end of 2023, and its one that I had never really expected to return to either. Grumble Volcano has, for the past four years, been the home of the fastest single lap in all of Mario Kart Wii, and after all that time, the Grumble Volcano Unrestricted Flap is still the only category in the game to have been pushed under two seconds. The result of this unparalleled shortness was a flap that went from a mere proof of concept to a fully fleshed out run that utilized every improvement we could have possibly thought of, all in less than a month. Much of the community felt that this TAS was effectively unimprovable, and as such, we moved on from it. However, with some motivation from the MKW TAS Competition and four more years of experience under my belt, I finally returned to the category to improve the seemingly unimprovable.
First, lets lay the groundwork for this flap to explain why we felt it was at its limit. At its core, the GV UR Flap is incredibly simple: backwards hop into the lap, then instantly accelerate forward upon landing. Thats it. There are literally zero frames where we arent moving as fast as the game can physically allow us to move in those circumstances. Further, the former BKT delays the flap as long as possible to minimize time wasted in the air, and it also hits the final KCP for only a single frame upon accelerating towards the finish, so how can this possibly be improved? The answer...a new vehicle, and an incredible amount of precision.
Lets start with the easy one, the vehicle. Throughout the entire flap, we are moving at shroom boost speed (and were in the air for half of that time anyway). Thus, the only place where our vehicles speed stat matters in this flap is in the drift after landing from the backwards hop, and it only matters marginally. As such, it seemed clear that using a vehicle with better drift and handling than the Spear would be preferable, and the Mach Bike solidly fills both of those roles while still having a great speed stat (just to minimize any timeloss from the drift). So, by using Mach Bike, we are able to more quickly straighten our line towards the finish line, and as such, we also gain the ability to throw in a wheelie before going airborne to gain some more speed.
Now for the complicated part. Remember when I said that the former BKT minimizes the time wasted in the air? Well, that was actually a lie. This new flap lands a frame earlier than the former BKT does. But how? The former BKT only hit the final KCP by one frame, so surely landing a frame earlier would cause you to miss the KCP outright. Normally, this would be correct, but theres a crucial detail that was missed when the former BKT was made. We land on the rock outside range of the final KCP, and then upon accelerating, the facing angle of the vehicle causes us to drive just barely into the KCP from the side. Therefore, if we land too far to the left, our vehicle will need an extra frame enter the KCP from the side, and to compensate for the distance we travel in that frame, we also need to land a frame later so as to not miss the KCP. This ended up happening in the former BKT without us realizing. To fix this, we needed a midair trajectory during the backwards hop that keeps us as close to the checkpoints as possible. But remember, to enter the lap optimally, we need to hit the finish line as late as possible which, due to how the checkpoints are laid out, means we must have a trajectory that passes through CP 37 (the final regular checkpoint on the track) but with frame alignment that allows us to avoid CP 37 entirely. As a result of both of these optimizations being put in the same run, our vehicle comes within a FOURTEENTH of a unit from hitting CP 37. In the real world, this would be like avoiding a randomly placed obstacle by roughly half a millimeter while driving at 110 mph (178 kmh) blindfolded with your arms tied behind your back entirely on purpose. This insanity allows us to land far enough to the right to land a frame earlier than the former BKT while still hitting the final KCP. So yeah, thats the run. Enjoy TAS!
Music: Snowball Park - One of the composers for the Super Mario 3D World soundtrack
MKW TASing Discord Server: https://discord.gg/EPD9yCu
MKW TAS BKT Website: https://mkwtas.com/")
![[MKW TAS] Coconut Mall (No Ultra) New Strat](https://i.ytimg.com/vi/CX3f6SXzAHs/hqdefault.jpg "[MKW TAS] Coconut Mall (No Ultra) New Strat
Props go to LuigiM and BennyMKW for theorizing parts of this strat. Do note that this video was simply showing that this strat is possible and faster. The lap itself is nowhere near optimal. Also, sorry about having no audio. I wanted to upload this fast, and capturing frames and audio at the same time takes forever.")
![[MKW TAS] N64 DKs Jungle Parkway (Glitch) - 20.535](https://i.ytimg.com/vi/CfB8vACRDYs/hqdefault.jpg "[MKW TAS] N64 DKs Jungle Parkway (Glitch) - 20.535
This is a tool-assisted speedrun. This time will be counted on every real-time leaderboard lol.
So I figured out that if you start the race in 2nd gear, MKW turns into MK64. Pretty cool glitch if you ask me.
In all seriousness though, April fools!
Huge props go to my man LuigiM for editing the collision of the track by removing the two wall triangles (to allow me to perform the MK64 version of the glitch) and also making the offroad driftable (for further MK64 accuracy). None of this could have been done without him. I would like to point out that none of the key checkpoints were edited to make this version of the glitch possible. The only thing that makes this version of the glitch impossible is the wall itself, not the position of the checkpoints.
Of course, this run wont actually count on any real-time rankings (or any tool-assisted rankings). This TAS was done on Dolphin 5.0 with savestates, slowdown, frame advance, TAS code, and RAM watch (created by Swarejonge).")
![[MKW TAS] GBA Shy Guy Beach (Ultra) Flap - 17.186 by Jellopuff, cf, and Ejay B [OUTDATED]](https://i.ytimg.com/vi/D2Ax0Sy-Lzc/hqdefault.jpg "[MKW TAS] GBA Shy Guy Beach (Ultra) Flap - 17.186 by Jellopuff, cf, and Ejay B [OUTDATED]
This is a tool-assisted speedrun. This time will not count on any real-time rankings. Done on Dolphin 5.0 with savestates, slowdown, frame advance, TAS code, Lua scripts (input writer/reader), and RAM watch (created by Swarejonge).
TAS Flap BKT Cut: -0.543
Previous BKT: https://www.youtube.com/watch?v=S_SMQQiTWRo (by Aus Tom)
cfs channel: https://www.youtube.com/channel/UCPfb0RuttHWFvPd1oyPx6tg
Ejays channel: https://www.youtube.com/channel/UCL6IvK60jAvowbyeDnu0Ybw
The creation of this run was mainly separated into two parts, specifically before and after the respawn. By using Ejays basic alignment, cf managed to improve the bomb clip by a significant amount. At the same time, I was improving the 2nd half of the flap by using the Lap Completion tool in RAM Watch and an improved landing from the respawn. By using Lua scripts, we managed to put the two runs together to create the final run. I would like to give a special shoutout to Derek/TASPlasma for spending the time to combine the inputs of the two runs for us. Thank you for helping us out, man. :D
This run is still improvable by a fair bit. Improvements could be made in a few ways. First, Luke showed that it is actually faster to use the Manual transmission on this flap simply due to better lines and a miniturbo near the end of the run. Another possible improvement may come from the use of the Magikruiser, which actually benefits from its high acceleration stat allowing it to not have to charge an SSMT after hitting the last KCP. While were still not certain if the glitch is possible with Magi, cf has come close to making it with it, so keep your fingers crossed. And the final possible improvement simply comes from waiting for the 3:24 bomb to drop, since its slightly closer to the position we need to fall in for the lap to count. With further optimizations, the sub 17 second flap is pretty much inevitable.")
