New Supergrind on Bowsers Castle @Kierio05

Kierio05 Video in the comparison is the 42.923 no glitch flap from Monster's channel (youtu.be/0gtfGke1io0).

As seen in the comparison, this currently saves 9 frames over the high trick entry. I suspect I can get it to at least double that number by the end of the night because of how quickly I threw this together (15-20 minutes).
Edit: Managed to get it to be 21 frames faster than high trick entry.

People have tried numerous ways to get a Bowser's Castle supergrind to work, using these stairs, using the offroad next to the stairs, and using any of the geysers in many directions. I can say with experience that none of these worked, as I have tested all of them at least two or three times each over the past 3 years.

What makes this different to the ideas we had before is the use of the shooting star, an outside drifting bike (ODB). ODBs have a different "range" of facing angles they can start supergrinds at relative to the direction you want to supergrind in. Inside drifting bikes have their own range, which has its limits, preventing something like Ancient's recent GV supergrind or this one. This range includes starting it facing forwards and moving sideways already, using the sideways momentum from a drift or hop (which is also exclusive to ODBs).

This is the first original Nintendo track supergrind strat I have found in well over a year and a half (618 days or 20 months) so I am very pleased that this worked as well as it did. Fingers crossed that I can improve it to save enough time to make the shooting star and the supergrind viable for glitch flap!

While writing this description I thought of another supergrind idea :o

updated 2 years ago

New Supergrind on Bowsers Castle

Kierio05 2022-03-22 | Video in the comparison is the 42.923 no glitch flap from Monster's channel (youtu.be/0gtfGke1io0).

As seen in the comparison, this currently saves 9 frames over the high trick entry. I suspect I can get it to at least double that number by the end of the night because of how quickly I threw this together (15-20 minutes).
Edit: Managed to get it to be 21 frames faster than high trick entry.

People have tried numerous ways to get a Bowser's Castle supergrind to work, using these stairs, using the offroad next to the stairs, and using any of the geysers in many directions. I can say with experience that none of these worked, as I have tested all of them at least two or three times each over the past 3 years.

What makes this different to the ideas we had before is the use of the shooting star, an outside drifting bike (ODB). ODBs have a different "range" of facing angles they can start supergrinds at relative to the direction you want to supergrind in. Inside drifting bikes have their own range, which has its limits, preventing something like Ancient's recent GV supergrind or this one. This range includes starting it facing forwards and moving sideways already, using the sideways momentum from a drift or hop (which is also exclusive to ODBs).

This is the first original Nintendo track supergrind strat I have found in well over a year and a half (618 days or 20 months) so I am very pleased that this worked as well as it did. Fingers crossed that I can improve it to save enough time to make the shooting star and the supergrind viable for glitch flap!

While writing this description I thought of another supergrind idea :o

Full Uncut MGTC MKWii TAS Interactive Tutorial

Kierio05 2024-02-28 | Recorded by and featuring commentary from @Jellopuff_TAS.

Made for the Multi-Game TAS Competition, and hosted in the discord: discord.gg/PbY433kFQN

I plan to make an abridged version of this with cuts and speedups, but have got many requests to upload the uncut version as soon as possible!

New IDB Supergrind on Bowsers Castle

Kierio05 2023-09-02 | *Short story:* This saves 1 frame over the double trick entry.

*Long story:*

Last night, out of nowhere, I thought "hey, the reason the ODB supergrind worked was because of the optimal counter-rotation, not the use of outside drift, so shouldn't IDB be able to get it with optimal counter-rotation too?"

When I found the time to test this today, I tried the exact same setup as the ODB supergrind (youtu.be/Xo9ejZvKXhw), but realised quickly that the stairs were not feasible to build and carry EV past the finish line.

Next, I tried a setup off the geysers (youtu.be/maxMPSGztwY), was able to get the supergrind (and carry it to the first ramp), but it was about 0.6 slower due to the fact that I didn't have enough EV to RFH the ramp (so was forced to eject early in order to hit the boost panel).

Finally, I tried a setup in the offroad (youtu.be/WFcCjWEEL-U), which let me carry much more EV onto the ramp; enough to let me RFH up it, giving me a low trick.

This saves 1 frame over the double trick entry (youtu.be/3anjf7Go2hE). However, this is just a prelim demo so could definitely save much more.

Alternative Encode for RR TAF

Kierio05 2023-07-27 | Wanted this specifically for 0:44 and 1:40. Watch the original encode here: youtu.be/ugyglcyKJC0

The Longest Strat in MKWii TAS History (Reverse Ultra on Christmas Court)

Kierio05 2023-05-14 | As seen on tracks like Abandoned Boardwalk and Aquania, it's possible to complete a lap backwards without hitting all key checkpoints. While the traditional definition for an ultra is that it saves time, or even (for CTGP Backwards) that it makes some sort of mushroom/respawn improvement, this does neither.

The current no ultra route takes less than a minute and 1 mushroom, whereas this takes over 11 minutes and 2 mushrooms. Why demonstrate this at all? Well, just like how ultra longcuts show the more general challenge of finding a way to complete a lap forwards without hitting all key checkpoints, this reverse ultra in particular shows the more general challenge of finding a way to complete a lap backwards without hitting all key checkpoints. So, any track which has some sort of method of performing a reverse ultra, whether that be an improvement in respawns/mushrooms/time or not, will be documented separately, as making them happen is not trivial.

What is actually done in the video? Well, I perform the first part of the regular ultra to get onto and around the wall, then cross back into lap 0, drop down to the track below, and drive forward. I then get myself on the outside of the track and start wall hugging.

I am currently locked in CP26 (the last CP), so if I were to drive into CP0 the lap would increase back to lap 1 due to the reverse last CP rule, so I need to be in CP25. CP25 ever so slightly cuts into the region I can wall hug, so slightly that I have to lean to the side to reach it (which is incredibly precise to hit, let alone avoid reentering CP26 when returning to neutral).

Once I've escaped CP26 and locked myself in CP25, I have no choice but to wall hug forwards around the entire track to get back up (continuing backwards would unavoidably make me enter CP26). This takes about 9 minutes, during which the input limit is reached, so the rest needed to be TASed by manually entering the inputs into a csv file.

Once I get far enough, I'm able to use 1 mushroom to clip back in bounds, then use the newly found way to skip the 1st block (youtu.be/NEgukdx0Qlo), which was actually found while working on this strategy on a different savestate. Getting it with Spear is ridiculously hard compared to bullet bike as the heavyweight class has much worse trick height. From there, I used the shroomless clip for the 2nd block skip and completed the reverse ultra.

This reverse ultra was achieved more than a year ago, but as stated already went well over the input limit, so couldn't record it. I was not looking forward to redo it for documentation, but once we finally reached Christmas Court in the series, I had no choice but to try. I was able to do everything up to the input limit using normal TASing techniques, but was forced to do the last couple minutes of wall hugging, the clip back in-bounds, the 1st block skip, the 2nd block skip, and all driving in-between manually in a spreadsheet.

I reached a brick wall at the 1st block. I spent several days trying to get it without success, so sent it out to Thomas to try it, as he had experience with TASing past the input limit on DS Peach Gardens. He is not a particularly big fan of this series so didn't prioritise it, and after 2 months wasn't able to get any success. Another 2 months later, I sent it to Police, who was able to get marginally closer, and later that night, I redid the skip from scratch with new motivation and got it.

All in all, the reverse ultra cannot be done with less than 2 mushrooms, and using 2 or more mushrooms gives a 0 respawn solution. These statistics are separate from the no ultra route.

Optimal Supergrind Counter-Rotation

Kierio05 2022-12-20 | DISCLAIMER: This is not a brand new discovery, just one from several months ago which has never been formally documented.

Fast ghost with XZ speedometer: youtu.be/OXxBFRxwXD8
Slow ghost with XZ speedometer: youtu.be/_k8Z6ORU6nI

See, when working on making the Grumble Volcano ODB supergrind possible on the right side of the track (youtu.be/e4ikQuJnnMI), I found it was possible to maintain a supergrind while turning the opposite direction you're moving without slowing down significantly (I do it briefly to avoid offroad). This was a big development, and called for more testing.

So, I moved to the first thing I thought of - Bowser's Castle. I had poured hours of attempts into making a supergrind possible on that track, and the thing that always killed any chance was the fact that you couldn't move fast enough with a good angle. So why not just give myself a better angle by counter-rotating? That's exactly what I did (youtu.be/Xo9ejZvKXhw).

Shortly after that, Marth and Thomas found a way to get the slower IDB supergrind on DS Peach Gardens to work on ODB (Marth found the supergrind, and Thomas managed to do it facing the other way and carry it around the next turn), as well as finding a bonus supergrind and a way to make the flap-only supergrind viable in a 3lap run (streamable.com/sn1wtf, streamable.com/fmdsro, streamable.com/sxblko). The issue with the way they navigated the hedge maze was they were counter-rotating the usual way, one which gives speed loss and therefore significant time loss. So, when I made my flap LOTAD, I performed this hedge maze supergrind using my newly found optimal counter-rotation (youtu.be/JJHfzxDR0G4).

Along with the new rPG supergrinds came new rDS supergrinds. You may know I am a fan of DS Delfino Square, so my interest and motivation were piqued. I made a 3lap LOTAD which once again utilised the optimal counter-rotation throughout the TAS (youtu.be/njC_z7c4K6c), but 2 days later had the idea of doing it to the extreme, where I wouldn't just correct my angle, but actually take a turn backwards. That's exactly what I did (youtu.be/w8KG8TRRV6I). Explain the last 10 seconds of lap 3 to me from 2020 and I wouldn't believe that it was real.

Anyway, what even is optimal supergrind counter-rotation? Well, supergrinds are made up of two repeated frames, the B press and the B release. In terms of hop mechanics, the first is the angle hop frame (you can choose any direction and it won't affect the hop/drift following), and the second is the committing frame (whatever you choose on this frame is the direction you rotate). You always input a direction on the angle hop frame, and always input neutral on the committing frame.

The angle hop frame is responsible for the momentum increase during a supergrind, and the second is responsible for the lack of rotation you receive relative to simply holding a hard direction every frame. In late 2019, I found that on shallow curves or short straights, the best way to supergrind is to use the least amount of input on the angle hop frame (which is a magnitude of 2 since magnitude 1 behaves like a neutral input - this was eventually named "softgrinding").

When working on the second rMR supergrind in early 2020 (youtu.be/5ugKk-B-jW0), I realised you could also use hard inputs (magnitude 7) on the committing frame to turn yourself the opposite direction, but it slowed you down massively. It wasn't until working on the GV SG early this year when I realised the hard input was the thing slowing me down, and found the least amount of input on the committing frame (which is a magnitude of 1) would not slow you down significantly, as that magnitude behaves like a neutral input.

This tech is now central to how supergrinding is performed, and has been crucial in making some strategies even viable (it is used throughout the entirety of the new MMM ODB SG in order to maintain high speed without turning right: youtu.be/31cScT_JZls).

Aquania Hidden Boost Panels Exploration & Seam Riding

Kierio05 2022-06-07 | In this video, I show an area with boost panel collision on Aquania (v1) that allows for some exploration all the way around the building on the first turn of the raised open road. I also explain some extremely basic mechanics to do with velocity while driving on OOB seams. Nothing much more to this lol

Various Superhop Strategies on N64 DKs Jungle Parkway

Kierio05 2022-05-12 | I meant to upload the first one up to the cannon 3 days ago but forgot to publish it and decided to test the rest of the lap instead yesterday. I then forgot to record that lap so here it is finally. Although the lap time is slower than even the RTA no glitch (no spiral skip) lap 1, there are 2 places where I am clearly slowing down where I shouldn't need to (grass hit after cannon, wall hit on bridge), and 2 places where I am slowing down more than I need to (cannon exit and wall hit at shroomspot), so I do think this might have some potential, we just might need to use a different vehicle (hinting at cannon exit since that's the most annoying timeloss).

Note that even though a supergrind already existed on the long turn leading up to the bridge, superhops accelerate much faster and have no turning limit, so pretty much invalidate any boostless supergrind on a curve (sad).

Downhill Supergrind using a Kart (DSG)

Kierio05 2022-05-11 | The speedometer in the bottom left is showing my "XYZ speed", or the absolute amount of units I travel in one frame.

The reason I'm using a kart is just to demonstrate the fact that the speed gained in this video has nothing to do with the traditional method of supergrinding (repeated angle hops, which karts are unable to do). All of this speed is being gained thanks to the hill I'm moving down. Why this happens, and our movement speed isn't capped at 120, we still don't know.

Special thanks go to _tZ (youtube.com/channel/UCFHoTC2NXG-i-rHEhDrS1RQ), who demonstrated this to be possible in April 2021 (cdn.discordapp.com/attachments/356185849337085953/830630105435996181/CDpOuz1vK7.mp4), and reached a peak of about 100 XYZ speed, which was ultimately my inspiration for trying this again due to all the new superhopping developments.

This also happens to be the fastest speed ever travelled during a supergrind. You will notice I peak at about 125.5 in this video. Note that I was able to reach 127.8, but doing so with this setup made me run into the wall. A history of the fastest speeds accomplished can be seen below:
120 - First supergrinds (2014-15, on Luigi Circuit)
121 - Thomas (2018, on Luigi Circuit)
122 - myself (2019, on Luigi Circuit)
123 - Jellopuff (2019, on N64 Mario Raceway)
124 - myself (2020, on N64 DK's Jungle Parkway)
125, 126, 127 - myself (2022, on GCN DK Mountain)

Fast Wheelie Glitch Timesave on Rainbow Road

Kierio05 2022-05-08 | Faster not wheelie glitch

Lap time was 46.177 (rs's 1 shroom laps are 44.882). shroomspots are probably going to be second turn, turn before cannon, and turn after cannon, but I did this to have an idea of how much slower I still was against RS. For the record, I'm 0.6 behind by the time I get to the cannon. With the knowledge I have on the category system, this means Rainbow Road now has 3 categories - Unrestricted (ultra), No Ultra (fast wheelie glitch at turnskip), and No Glitch (no fast wheelie glitch at turnskip).

This timesave was independently discovered by _tZ (youtube.com/channel/UCYgxToY9uwv_9TtoG-c42OA) in January 2021 (cdn.discordapp.com/attachments/735589843513638940/806286290030690375/3fKwroUi3H.mp4), however it wasn't until I was uploading this video that I realised he actually used fast wheelie glitch to take the turn tighter. Note: The video linked above shows _tZ racing an earlier attempt of his timesave, rather than RS's original lap 1.

rSGB glitch 3lap LOTAD (Shooting Star / Twinkle Star)

Kierio05 2022-04-28 | 59% of you were right!

See the Zip Zip / Rapide run here: youtu.be/trmk5u-ssyY

Thanks to all the people in voice chats with me over the past few days giving me motivation to get the ultra (especially the second time). Special thanks to @justin3055 for discovering the new ODB-exclusive 3lap viable ultra strategy (youtu.be/oGn2mE27czE)!

I love that I can use the fact that this vehicle bounces too much to my advantage.

On any shroomless lap, zip zip gains a significant amount of time over shooting star because it can take the water route and shooting star has to go all the way to the ramp path. Therefore, on both 3-0-0 (2 shroom ultra lap 1) and 1-2-0 (2 shroom ultra lap 2), shooting star is slower. For TAS, shooting star is only barely faster because of its speed and drift relative to zip zip's offroad and acceleration. Yoshi is not faster than Daisy here. Shooting star would probably be slower if it were not for the OP island section strats that RTA would absolutely not be doing, meaning even for 1-1-1 (1 shroom ultra lap 3), zip zip is still likely faster for RTA. However, RTA shooting star laps seem to be faster than zip zip (which may be due to zip zip not having been pushed to its limit) so shooting star seems to be the current fastest option for 1-1-1 (1 shroom ultra lap 3).

rSGB glitch 3lap LOTAD (Zip Zip / Rapide)

Kierio05 2022-04-28 | 41% of you were wrong!

See the Shooting Star / Twinkle Star run here: youtu.be/G_4trkj7ACo

Thanks to @EjayB and @campbellmop355 for helping me out with deciding on the best way to approach the first turn of lap 2, and for all the people in voice chats with me over the past few days giving me motivation to get the ultra lol. Special thanks to @justin3055 for discovering the new ODB-exclusive 3lap viable ultra strategy (youtu.be/oGn2mE27czE)!

Funnily enough, the ending strat saves 0.001 over a complete stick on the offroad.

On any shroomless lap, zip zip gains a significant amount of time over shooting star because it can take the water route and shooting star has to go all the way to the ramp path. Therefore, on both 3-0-0 (2 shroom ultra lap 1) and 1-2-0 (2 shroom ultra lap 2), shooting star is slower. For TAS, shooting star is only barely faster because of its speed and drift relative to zip zip's offroad and acceleration. Yoshi is not faster than Daisy here. Shooting star would probably be slower if it were not for the OP island section strats that RTA would absolutely not be doing, meaning even for 1-1-1 (1 shroom ultra lap 3), zip zip is still likely faster for RTA. However, RTA shooting star laps seem to be faster than zip zip (which may be due to zip zip not having been pushed to its limit) so shooting star seems to be the current fastest option for 1-1-1 (1 shroom ultra lap 3).

New Supergrind Strat on ODB GCN Peach Beach

Kierio05 2022-03-29 | Obviously, the previous video showed that Shooting Star isn't gonna be useful for 3lap, but this is a demonstration of a timesave for a hypothetical Shooting Star TAS. Special thanks to CampbellMop (youtube.com/channel/UCTlFF4haRu3QFzfAZlCbgIg) for discovering this supergrind (cdn.discordapp.com/attachments/735589843513638940/958191957006499860/rpbstraight.mp4), I just made this video to prove it saved time.

rPB NU 3lap LOTAD

Kierio05 2022-03-29 | Unfortunately, not every single supergrind leads to timesave. Thanks go again to Wade474 (youtube.com/c/Wade474) for discovering this supergrind spot (cdn.discordapp.com/attachments/864397137667555329/958109348662497310/superground.mp4). I tried a few other shroomstrats but this was the fastest one. Ramp should be faster on lap 3 but for some reason I wasn't able to get the boost panels to activate (?) and so couldn't get the offroad immunity and speed required to get across the water.

Fortunately, this should still be faster for flap. Hopefully sub 21 is possible...

Another rDS SG 3lap LOTAD (new strat lap 3)

Kierio05 2022-03-27 | Skip to 1:55 for the new strat.

As with the previous TAS from 2 days ago, special thanks to dm248 (http://www.youtube.com/channel/UCSM97wYzq7HkeK0agLATZxw) for coming up with the stair supergrind, and Wade474 (youtube.com/c/Wade474) for coming up with the [lap 1/2] last turn supergrind and proving both to be possible and finding fast ways to perform them (youtu.be/t61I1UEd8qM).

However, the difference between the last run and this one is that I do a supergrind leading into the last turn on lap 3 as well, starting it from the bridge. This is, of course, much faster to do than the slope at the very end, since you spend more time above wheelie speed and enter the last turn at well over 100 instead of about 80.

The only downside to this one is you can't hold it as long (I think, I haven't done too much testing, but I always ejected early if I tried to hold it through the next uphill when I tested it on lap 2). This also is not applicable to flame runner since you can't start it with sideways momentum and are stuck at 60-70 speed for about 10 seconds before reaching the finish line.

I came up with this while I was falling asleep last night as well (same as with the LC shroomless jellogrind), so this seems to have been a very productive brainstorming period.

As this is a new BKT, previous BKT - 1:57.855 by myself: youtu.be/njC_z7c4K6c

See the video linked above for comments on where more time could be saved throughout the rest of the run.

Shroomless Jellogrind on Luigi Circuit

Kierio05 2022-03-26 | Yes, "Jellogrind" is and has always been the official name for the back straight supergrind (he did discover it after all and we needed some way to differentiate from the front straight SG, or the "Waltergrind").

I thought of this idea while I was falling asleep last night so wrote it down in a notebook and tested it this morning. Seems a bit slow so far (20.393 is the current fastest 1 shroom lap with a supergrind carried in, and this is 1.2 seconds slower), but I think a ton of time could be saved elsewhere (less time spent in the grass for setup, better SG movement in both SGs, better ejection in both SGs, a 120 speed trick off the ramp, a realign clip into the grass shroom) and time should be spent to see if this has potential.

rPG NU flap LOTAD

Kierio05 2022-03-26 | Special thanks to Marth (youtube.com/c/MarthTAS) for proving the hedge maze SG to be possible/faster on shooting star, as well as the same for the mole area SG (streamable.com/sn1wtf). Thanks also to Thomas (youtube.com/user/MKWiiTomas) for proving the hedge maze SG to be possible/faster going the other direction, making it possible to carry around the turn (streamable.com/sxblko).

As this is a new BKT, previous BKT - 36.709 by Akari, Marth, Luke, Rocky, Ejay, and myself: youtu.be/GKXXJE2b0oA

As that run is unlisted, and also not on mkwtas.com, previous BKT before that - 36.935 by Akari, Marth, Luke, Rocky, and myself: youtu.be/WUYo3Ug6qbM

NU means No Ultra / No Ultra SC / No Ultra Shortcut, which disallows skipping key checkpoints but allows everything else (in this case, supergrinds and highclips)

Some notes about the TAS:
- Probably possible to enter lap at slightly higher speed
- First turn SG could be taken tighter
- Second turn (1st shroom) is good, but exit is a bit off
- Third turn (before 2nd shroom) can maybe be taken tighter but idk if that'd actually be useful because I need to approach the next turn wide in order to be able to get momentum into the maze SG
- Maze SG is good, but uphill and approach to flowerbed is probably improvable
- Post-flowerbed ejection is good, but probably best to exit further right in order to be facing further left when starting mole area SG
- Mole area SG can probably be improved because I wasn't sure how I was gonna approach tree clip when I made this
- Tree clip setup (3rd shroom), airtime, recovery, and last turn can be improved

rDS SG 3lap LOTAD

Kierio05 2022-03-25 | Special thanks to dm248 (http://www.youtube.com/channel/UCSM97wYzq7HkeK0agLATZxw) for coming up with the stair supergrind, and Wade474 (youtube.com/c/Wade474) for coming up with the last turn supergrind and proving both to be possible and finding fast ways to perform them (youtu.be/t61I1UEd8qM).

As this is a new BKT, previous BKT (which doesn't use any SGs, so is now the no SG BKT) - 1:58.764 by Luke: youtu.be/u_RX0ok2wXY

This TAS can probably be improved by one or two seconds, so I expect a 1:55 to be possible with better driving, better supergrinds, and proper bridge strats. Some other comments regarding the drawbridge:
- Lap 1 bridge strat was only possible with a drift into it in order to get sideways momentum. Without it, the RFH wouldn't hold up to the top and the wheelie trigger wouldn't happen
- Lap 2 bridge strat would be improved with a double trick, like in Luke's flame runner no SG TAS
- Lap 3 bridge strat might be improved with a double trick, but idk if it'd still be too low to get a double trick off of (even with the major improvements leading up to this point). If that's the case then I'm not actually sure how much time would be saved since I optimised the second half hop strat quite a lot.

For the record, flame runner 3lap with SG was projected to get around a 1:57.9/1:58.0 and flame runner 3lap without SG is projected to get around a 1:58.5/1:58.6

Faster Supergrind Strat for ODB Grumble Volcano

Kierio05 2022-03-12 | This pretty much confirms shooting star being the optimal vehicle for no ultra 3lap. Please note the speedometer is showing XZ speed.

Original supergrind strat on left side by Ancient. Below are links to the progress that has been made and the frame counter values when we reach the second ramp.

Release video: youtu.be/Cwjs7buORPI - frame 759
Improvement: youtu.be/etf66b5_oTA - frame 746
Lap 1 progress: youtu.be/-lZ_uNOa3ec - frame 746
Prelim full lap 1: youtu.be/se5s2K4Kz9Q - frame 746
Ghost in this video (by Ancient and ThatsSlick) - frame 737
Player in this video (by myself) - frame 735
Potential - frame 72x

Special thanks to DwainiumB for requesting I test this.

rPB 200cc no glitch: 0-1-2 racing 0-0-3

Kierio05 2022-03-02 | ive been making a lot of lotads recently but for this track there wasn't a way to compare two routes that were this close without properly optimising, so here is a comparison from the slower ghost's perspective (0-1-2). im going to be uploading the full 0-0-3 TAS to my main channel in a few days so look forward to that :)

BC 200cc glitch LOTAD

Kierio05 2022-02-13 | -5.170 from ancient's tas: youtu.be/Vvqw9dhzGuU

MH 200cc LOTAD

Kierio05 2022-02-11 | strat

DDR 200cc no glitch LOTAD

Kierio05 2022-02-11 | :(

ddr is hard

current rta bkt is 1:15.632 (~7.5s slower)

902 racing 913

Kierio05 2022-01-23 | ghost is ejay+justin 6.913 flap

dry bowser might be faster but we're optimising both characters' flaps since they are too close to tell just from stats

im starting to feel like ive done this before

Kierio05 2022-01-18 | maybe it's because i have

rBC further SG improvement

Kierio05 2022-01-17 | ...

rBC start comparison

Kierio05 2022-01-16 | start slide by ejay
sg setup strat by marth
all inputs by me
ghost is current BKT

turnaround

Kierio05 2022-01-12 | ...

DS Airship Fortress - Lap 1+2 Progress Update

Kierio05 2022-01-12 | found a few slightly faster things lap 2 (tighter first turn, higher trick off 2nd ramp, no RSTA off 3rd ramp, double MT at end of spiral) and am going to test if using the 2nd shroom for turnaround accel lap 3 is better than where I've used it on lap 2.

The First Shroomless DS Airship Fortress Ultra Shortcut

Kierio05 2022-01-10 | long story short i am sad

DS Airship Fortress Lap 1 Progress Update

Kierio05 2022-01-10 | watch in 2x speed to be less dissatisfied. recording went slow for some reason ¯\_(ツ)_/¯

if you think spiral looks weird it's because it does. your drift gradually gets wider the longer you hold it so redrifting before the miniturbo is charged allows the drift near the end of the spiral to go much tighter than if you were to continue chaining miniturbos.

Strat Test for Lap 1 of DS Airship Fortress Glitch 3lap (LOTAD)

Kierio05 2022-01-09 | Outside drift bikes are very fun. I tested out using mushroom for mole clip (bat clip but retextured) instead of the spiral exit, like in Marth's inside drift lap (youtu.be/s3aVaRjf9Us). I've optimised the start slide (thanks to Citrinitas for providing the inputs for a non-sliding bike) and first trick, and the rest is just preliminary. This is a low-optimisation tool-assisted run made simply for the demonstration of strats I have come up with (LOTAD).

Probably not going to finish this TAS now that I've been told kart is probably faster.

Task 3 2020 - Slower Supergrind Strat

Kierio05 2021-12-29 | :(

MT 200cc glitch LOTAD

Kierio05 2021-12-26 | :)

MMM SG flap - 22.202

Kierio05 2021-12-24 | Ghost being raced is 22.253. Current BKT is 22.284. Mid to low .1 is probably happening. Timesaves galore

MMM SG flap - huge improvement to ejection

Kierio05 2021-12-24 | so as it turns out you can save time here too

MMM SG flap - no drop ramp

Kierio05 2021-12-24 | :o

MMM SG flap improvement

Kierio05 2021-12-24 | Because of the NG flap uploaded today, I wanted to take a look at the SG flap. This is what I have so far

MC 200cc glitch LOTAD

Kierio05 2021-12-20 | I'm not actually intending on finishing this. I just happened to be on this track and wanted to see what a time with a better shroomstrat would be like.

Backwards Longcut Route on TF 200cc

Kierio05 2021-12-20 | As of right now, there exist these kinds of routes (where you hit a KCP, respawn, end up past the next KCP, drive backwards to hit it, and complete the lap) on DK Summit, DS Yoshi Falls, DS Delfino Square, and DS Desert Hills. It's likely there are more but I enjoyed making this one in particular so decided to upload it just as proof.

Strat/Route Test for Skyline Avenue (LOTAD)

Kierio05 2021-12-13 | LOTAD = Low Optimisation Tool Assisted Demonstration
i.e. a prelim TAS made to demonstrate strats

Previous BKT, on v2 (49.596 by MilesAlley66): youtube.com/watch?v=Fu-_bas1yBY
Previous BKT, on v1 (48.991 by DrakeTheBeast): youtube.com/watch?v=iMRE2izsnKc

CM no ultra flap miniturbo strat

Kierio05 2021-12-05 | Ghost being raced is the Slick, Aaron, Ancient, and Ejay's 19.697. Still haven't beaten the 1f QM lock on the bonk exiting the OOB. I feel like this strat will save 3 frames when optimised fully (1 from current timesave, 1 from better shroom timing and decel stacking, 1 from better kcp2 turn and OOB exit). Idea by mkwLuke (youtube.com/user/MKWLuke)

Task 20 2021 - Proof Its Possible to Beat DashQC

Kierio05 2021-11-18 | In this video, I show a low-optimisation TAS made to demonstrate the time potential (LOTAD) for this task, that improves on both 2nd (37.117) and 1st place (36.267) with a time of 35.800. I first show this run in comparison against the 1st place run (by DashQC and AshMK), and then in comparison against the 2nd place run (by Jellopuff and myself).

This could be improved in 3 ways.
1. The ending is NOT optimised, however this doesn't particularly matter, as you'll see in the next 2 points.
2. The shroomstrat in Jellopuff and my submission was not suitable to allow this route to be maxed (we used 2 mushrooms before the dash cut spot and the dash cut requires 2 mushrooms to work). The mushroom at the beginning should be replaced with a mushroom to boost through the long mushroom section at 100 instead of 73 km/h
3. CampbellMop showed (in his DQ run) that it was possible to do the 1-pass starting boost panel strat with mach bike by getting stuck in the ground to do a large hop that minimises the boost time (therefore maximising the drift at lower speed, allowing the 1-pass strat to be possible). So, time could be saved from using a faster vehicle, performing t and also gaining time at the dash cut (due to Twinkle Star / Shooting star requiring awkward setup angle and getting bad exit angle)

Fun fact: this run was actually made before the task ended so if I had thought of this in the span of the task running, I could have got this time or better

Task 19 2021 - Strat Test Compilation ft. CampbellMop

Kierio05 2021-11-17 | Campbell's channel: youtube.com/channel/UCTlFF4haRu3QFzfAZlCbgIg

Task 18 2021 - My Submission and Bonus TAF

Kierio05 2021-11-14 | In this video, you see my submission to Task 18 of the 2021 MKWii TAS Competition in full, including a freerun afterwards. For more information about what you are about to watch or have just watched, please read the following:

0:00 Following the rules, the fastest combination is Daisy on the Standard Bike M, which has a base speed of 80.30. There were a couple people who didn't check the stats of medium vehicles and assumed Zip Zip was faster. Daisy on the Zip Zip has a base speed of 79.77. I was not one of these people, as I know my way around CampbellMop's stat sheet (docs.google.com/spreadsheets/d/1ZbrFItAoVqMiZP6n6vUEjjlJkK-uCdWkfY6U5g2kv-w/edit#gid=984220244).

0:02 Standard Bike M is the kind of vehicle that can slide but only while not holding A (some can't slide at all, some can but only when A isn't held, and some can always), so I perform the current best start slide known for this type of vehicle. Start slides just move you forward before the race begins. When the race begins, I do a readjustment which turns me faster than a straight wheelie.

0:10 Because I restrict myself to less time due to being a competition admin, I didn't spend a lot of time routing or strat testing this task, so I logically assumed Justin strat would be slower due to offroad glitch but didn't realise how much the shortcut saved. I also didn't test shroomspots and therefore I didn't test using mushrooms to go 100 over long driving sections, which proved useful in the top few placings.

0:15 Yes, the way I do these turns is faster than what you might think would be. If you watch 6th place in comparison to mine in the result video you will see that he loses a lot of time in that section. There's no simple objective explanation for why these are faster but through lots of rounds of optimisation, I'm confident that they are.

0:24 I performed these turns very badly. See 4th through 6th place for a good representation of an optimal turn here.

0:28 I made this look easy. It is not. Green mushrooms are still unimaginably bouncy, even though they don't have bouncy collision. Getting the RFH I wanted on the first sequence and the sticks on the second sequence took the majority of my time working on this TAS.

0:32 "Isn't the speed cap 73?" Say hello to mushroom abuse, where touching the edge of a mushroom in weird ways and performing side trick abuse can bring your speed up to more than the speed lock allows. With outside drift, it turns out that you don't even have to perform real mushroom abuse to get the speed benefits from it. You just have to be using the sideways momentum from the oustide drift and side trick abuse. 4th and 1st place showcase the usefulness of this very well and you'll see plenty more in the Task 20 results video.

0:34 see the comment two paragraphs above

0:37 similar to the old last turn strategy on GV when it was using outside drift, it's better to drift right here so that you can get sideways momentum left. Everyone who does this does it differently. What I do in my run is the best for my combination of approach position, timing, and angle. Yes you can trigger the lap from the wall. I'm not sure if it's possible to do this at the extreme left side of the wall but it obviously works where I do it.

And now for the TAF...

0:43 I convert the sideways momentum from the wall into outside drift sideways hopping momentum (comparable to side trick abuse) and accelerate.

0:45 I cancel my sideways momentum and get immediately decelerated below 10 speed, forcing me into a SSMT (comparable to wade trick abuse)

0:48 By wheelieing at the right time while accelerating from a boost, I give myself a large pocket of airtime, where I reo drift out of that pocket, which always looks cool :p

0:54 Slowing down here while making it look like I used instant finish code (sets your inputs to default 0,0 which is full down left - netural is 7,7), however it's really difficult to slow down on these without falling off.

1:02 Certified CampbellMop instant finish fakeout. Again I struggled to turn around without falling off.

1:08 I originally planned on either wallclipping up to get back on the track shroomless (as I had already exhausted all my mushrooms), but getting knockback from the wall to land on the edge of the mushroom, and barrel roll seemed way cooler.

1:13 I decided to use infinite mushrooms code (and eventually had to get Vabold to use that code on stream so that we could show my full TAF) so that I could escape from the green mushroom back on to the track. This is harder than it looks.

1:24 I use a mushroom off the boost pad to give myself sideways momentum at 100 speed so that I can use another mushroom to perform side trick abuse on a red mushroom and land on the wall, riding it along, barely not falling on to the ground, riding across to the other side, and then getting held in place due to the wall holding me up, even though I'm facing the other way. Coolest part imo

[TAS] Wii Mario Kart Wii 100% in 00:19.10 by Kierio04

Kierio05 2021-03-31 | For more information please visit http://tasvideos.org
Full explanation can be found here: http://tasvideos.org/7078S.html

Pretty cool glitch if you ask me.

the other cycle problem ft. kriller

Kierio05 2021-03-22 | ...

why did I have to save time in 1-1 and 4-4 this would have been amazing

Kierio05 2021-03-22 | ...

5-1 strat idea (slow setup - proof of concept) *terrible quality*

Kierio05 2021-03-19 | sorry sm64 people but this is just my secondary account so any other game goes here, not just that... :/

[TAS] AMA C2 - Shoot into the Wilder Blue - 803

Kierio05 2020-10-11 | Original video: youtube.com/watch?v=N-2Gp1U_iwM
(Originally uploaded on 25 September 2020)

Rerecords: 112 (yes this low)