Kierio04 | Quest to Beat the MKWii TAS Competition - Task 4 2020 @Kierio04 | Uploaded 1 month ago | Updated 4 hours ago
Join the discord to help us on the next TAS: discord.gg/CJURvN4f3f
Check out the original result video here: youtu.be/aHZKyzL6UDo
TAS by @TASLirWm, @TASmarvus, @MKWiiTomas, @Alego11-TAS, @Kierio04 (myself), @thepolice3822
Original winning run by @Jellopuff_TAS, @EjayB
Music: Mario Kart Wii - Grumble Volcano
After almost another year, it returns again! Like last time, work has (very) slowly progressed since the release of the previous video!
In short, the following improvements were made:
- All-round better optimisation, including use of wheel inputs
- Using Spear instead of Flame Runner for higher speed
- Entering the lap as the rock is shaking to utilise bounces for the reo drifts
- Getting a slightly later lap entry (this can be seen in the ghost comparison)
- Delaying the second shroom to end as the RFH reaches the wall, in order to get a significantly bigger wall bounce SSMT
- Changing the shroomspot from after the SSMT to after the triple trick
- As a result of the new shroom, getting a much faster triple trick
In October 2023, immediately after the Task 3 2020 video was released, I posted a setup ghost using Spear to present a possible change combo for this task. A few days later, Police made a decent attempt at entering the lap with the "low DLC" strat. The tediousness of the lap entry caused a massive hiatus for the next 7 months.
In June 2024, LirWm picked up this project and started optimising the lap entry. Thomas then had the idea to change the lap entry timing to take advantage of the shaking rock for the reo drifts. He performed this and LirWm optimised it and later parts of the run over the next 2 months to finally reach a time of 7.747.
In August 2024, marvus helped improve the bounce and tricks, and worked with LirWm (and later Alego) to optimise the ending to bring the time all the way to 6.690.
This task made us realise that complexity was not the thing stopping us from making progress, but rather a lack of interested TASers was. Please consider contributing to this series by first joining the discord (link is at the top of the description)!
Join the discord to help us on the next TAS: discord.gg/CJURvN4f3f
Check out the original result video here: youtu.be/aHZKyzL6UDo
TAS by @TASLirWm, @TASmarvus, @MKWiiTomas, @Alego11-TAS, @Kierio04 (myself), @thepolice3822
Original winning run by @Jellopuff_TAS, @EjayB
Music: Mario Kart Wii - Grumble Volcano
After almost another year, it returns again! Like last time, work has (very) slowly progressed since the release of the previous video!
In short, the following improvements were made:
- All-round better optimisation, including use of wheel inputs
- Using Spear instead of Flame Runner for higher speed
- Entering the lap as the rock is shaking to utilise bounces for the reo drifts
- Getting a slightly later lap entry (this can be seen in the ghost comparison)
- Delaying the second shroom to end as the RFH reaches the wall, in order to get a significantly bigger wall bounce SSMT
- Changing the shroomspot from after the SSMT to after the triple trick
- As a result of the new shroom, getting a much faster triple trick
In October 2023, immediately after the Task 3 2020 video was released, I posted a setup ghost using Spear to present a possible change combo for this task. A few days later, Police made a decent attempt at entering the lap with the "low DLC" strat. The tediousness of the lap entry caused a massive hiatus for the next 7 months.
In June 2024, LirWm picked up this project and started optimising the lap entry. Thomas then had the idea to change the lap entry timing to take advantage of the shaking rock for the reo drifts. He performed this and LirWm optimised it and later parts of the run over the next 2 months to finally reach a time of 7.747.
In August 2024, marvus helped improve the bounce and tricks, and worked with LirWm (and later Alego) to optimise the ending to bring the time all the way to 6.690.
This task made us realise that complexity was not the thing stopping us from making progress, but rather a lack of interested TASers was. Please consider contributing to this series by first joining the discord (link is at the top of the description)!
![CTGP Tracks Done Backwards - Aquania 0x Respawns (& New Reverse Ultra Shortcut) [OUTDATED]](https://i.ytimg.com/vi/g6X8Os_7d4U/hqdefault.jpg "CTGP Tracks Done Backwards - Aquania 0x Respawns (& New Reverse Ultra Shortcut) [OUTDATED]
THIS VIDEO IS OUTDATED - PLEASE SEE A BETTER VERSION OF THIS VIDEO HERE: https://youtu.be/GUSsD9ofndE
In this video, I drive a full lap of the custom track Aquania backwards with the 3 mushrooms available in time trials. However, because it doesnt require all of them, I dont use all of them. In order for a track to count as being completed backwards, you must start in lap 1 and travel through all of lap 0 backwards, finishing by reaching lap -1. We will be going through the tracks of the CTGP 1.03-1112 update (29 Sep 2021) alphabetically, and returning to any past tracks if breakthroughs are made that change the statistics of the progression of the challenge (more on this later in the description).
Playlist of these videos: https://bit.ly/ctgpbackwards
Here, there are NO respawns that are mandatory to complete the track backwards, and there is ONE section requiring a single mushroom.
There is 1 block on this track:
1. Getting to the beginning of the cannon
This video is an improvement over the previous strat that completed the track backwards in 1 respawn shroomless (https://youtu.be/yJYl9zT6o3w). Since then, @EPIK95 found a way to wallclip from the last CP out to the start of the raised road section using a mushroom (https://youtu.be/iH00DQwxfAQ), which in combination with driving forward to hit the last KCP, allows for the only block to be avoided. While re-TASing this respawn skip to document it publicly, I found some hidden boost panels (https://youtu.be/Iw-BtMEfXfM), and while messing about with that I drove backwards and realised I accidentally performed a reverse ultra. This is shown before the no ultra strat so that it makes sense when I say no ultra as a caption.
Side note: This is the 3rd and final track affected by EPIK95s individual discoveries/improvements, the others being Candy Coaster and Abandoned Boardwalk.
In summary, Aquania has the following statistics to offer:
Minimum Respawns for Infinite Mushrooms: 0
Minimum Respawns for 3, 2, 1, 0 Mushrooms: 0, 0, 0, 1
Minimum Mushrooms for Minimum Respawns: 1
Minimum Mushrooms for Track Completion: 0
Special thanks to Melg (https://www.youtube.com/channel/UCi7BmIG91AG7lBt3sC_PkGw), for creating the debug panel in this video.")
![[MKWii TAS] DS Airship Fortress v1.3-ctgp - 1:42.592 (Ultra)](https://i.ytimg.com/vi/hAsF3AIkUYg/hqdefault.jpg "[MKWii TAS] DS Airship Fortress v1.3-ctgp - 1:42.592 (Ultra)
This is a tool-assisted speedrun (TAS). Played on SwareJonges Dolphin 5.0 LUA Core v3.5.2. Tools used: slow-down, savestates, frame advance, TAS input, TAS code, RAM watch (info display), CSV files, and LuigiMs TASToolkit. Download the ghost here: https://cdn.discordapp.com/attachments/930016627351756850/930949231345291364/BC_kiro_pls_1m42s592.rkg
There is no former full 3lap TAS to compare to.
Current RTA WR (1:47.085 by Jerko): https://youtu.be/O2hc3N33kls
Music: Mario Kart DS - Airship Fortress (https://youtu.be/azC_7dm41yE)
This track was created by Sniki, and updated by SpyKid. Download it here: http://wiki.tockdom.com/wiki/DS_Airship_Fortress_(Sniki)
Some notes about the TAS:
- All the strats done on lap 2 are faster than those done on lap 1
- The slide strat used here is optimal (provided by Citrinitas), but is stopped early due to not actually wanting to be too far left
- The up trick in the ship on lap 2 is faster than side tricks (RSTA slows you down too much and STA gives you too much air), but the air movement looks strange because of the restricted area for adjustment due to the size of the gap between the two crates
- The ultra can be done with 1 mushroom! Not a surprise since it was done with 2 but for the RTA strat this is actually a big deal since it means I was able to use a mushroom on lap 2 as well as lap 1
- The turnaround after the respawn actually changed the time of the run significantly. I had a 1:43.276 that just reversed straight, then a 1:42.884 that reversed diagonally to get the optimal line to the KCP, and now a 1:42.592 with SSMT turnaround & SSMT wall bounce (suggested by CampbellMop). If you are specifically interested in seeing these for any reason then DM me on discord (kierio04#9607)
- The shroomspot used here for the no glitch laps is much faster than the other 2 that I had tested (the one Marth used in his lap 1s and the one I showed in one of my strat tests using a bat clip off the rocky wrench)
- The turnaround right at the beginning of lap 3 was difficult to figure out what to do, but Ejay and Ancient both suggested not dropping my wheelie so that I could decelerate later and losing less time without having to worry about a down trick happening from being too close
- The spiral is handled very nontrivially. This is due to the fact that its a 600-700 degree turn, so just going tight wont cut it. Plus, outside drifts get wider and wider the longer you hold them (thank you to CampbellMop for telling me this), so its actually faster to do chain drifts (without a trick boost, which goes too wide too early) for the first third of the spiral, pseudo-chain drifts for the second third, and then terrain drifts for the final third
When Jerkos WR for the RTA glitch category was published on CTGP Records, my interest piqued and I wanted to TAS the track due to it using outside drift (something extremely rarely seen as an optimal combo). I begun working on lap 1 a few days ago, and posted small progress updates for most of this on my second channel (https://www.youtube.com/channel/UCGPuKK14yEQrqkkXV2NmYbA).
However, while working on the lap 2, I thought of an idea for a potentially faster way to perform the ultra by clipping up on the wall and then getting more distance off the wall due to it being much higher. I was unable to perform it due to the very small space required to accelerate. I then realised the edge of the wall was trickable and tried using a trick boost to accelerate slightly slower but also get more distance. Using 1 mushroom I was able to get the fast respawn by a mile (respawning about 20-30 frames after hitting CP41), so tried pushing the trick boost strat as much as possible to see if it was possible shroomless.
As you can see in this video (https://www.youtube.com/watch?v=WBmqZb7UD7A), I was successful. Even though that meant 3/3 ultra was possible, I was sad because my work on the 1/3 ultra TAS was obsolete. If youre reading this description, then you probably are confused as to why I uploaded this TAS at all. This is a TAS using optimal 1/3 ultra strats. Just because I made a TAS on this track with 1/3 ultra strats does NOT mean I want to make a TAS on this track using 3/3 ultra strats. I still may but please dont assume and start saying Im excited for the 3/3 TAS!.
Want to see more TASes and TAS projects? Visit https://mkwtas.com
or the MKWii TAS discord, at https://discord.gg/NCZghgy")
![[MKWii TAS] N64 Bowsers Castle Lap 1 - 51.213 (Shroomless - Bullet Bike) (PRELIM)](https://i.ytimg.com/vi/hmLjSs0oc08/hqdefault.jpg "[MKWii TAS] N64 Bowsers Castle Lap 1 - 51.213 (Shroomless - Bullet Bike) (PRELIM)
Ok so theres all 4 possible no glitch laps, now you can ponder these for a few days and figure out what is optimal on your own, or wait for me to explain it to you :)")
![(April Fools) [MKWii TAS] DK Summit Flap - 23.731 (Glitch)](https://i.ytimg.com/vi/iytXQkD4a54/hqdefault.jpg "(April Fools) [MKWii TAS] DK Summit Flap - 23.731 (Glitch)
TAS by @Kierio04 (myself) and @thepolice3822
-6.923 from TAS by Monster and Marth (https://youtu.be/mXmhqPMqSCE)
Music: Isana Remix - DK Summit Remix (https://youtu.be/L-4GJkBAcKU)
(Yes, I used this in the Wave 4 DLC Showcase, but I liked the music, and wanted to use it again)
When you enter a cannon, the game manually updates you position along a set path at a given speed determined by the cannon setting. In this case, your speed in the cannon turns out to be about 500. Whenever you exit a cannon, your speed must be returned to what it was before you entered it, so whatever action was occurring on the final frame before entering is applied to the frame you exit.
In the case of a low cannon, this happens because you graze a wall upon entering the cannon, and then experience the speed loss upon exiting it. Since wall hit speed losses are dependent on your angle relative to the wall, the more extreme the angle, the more extreme the speed loss.
So, what should be impossible under normal circumstances, if you hit a wall and lose almost all your speed upon entering a cannon, you will end up with negative speed exiting it. This should be impossible since if you drop to almost 0 speed and are facing towards a wall rather than the cannon trigger, you wont be able to enter the cannon.
Enter supergrinding. By supergrinding into a wall you can achieve the same angle you need for a 0 speed wall hit but have EV left over to bring you into the cannon on the following frame. This unfortunately comes with the caveat that the external velocity you use to get into the cannon is also considered in the speed loss on the final frame, and so you just end up with a regular low cannon.
By some stroke of luck, there is a way around this. For whatever reason, if you enter a cannon mid-air, your external velocity is NOT considered (presumably if you were to be accelerating in a fall, it wouldnt all of a sudden hurdle you out of the cannon on the other end with increased speed), so as long as you are in the air when performing a 0 speed wall hit with momentum directed towards the cannon, you will exit the cannon with negative speed.
But how much negative speed? Well, Nintendo wanted to ensure you wouldnt exit the cannon with more speed than usual, and wanted the higher speeds still retained a somewhat equal amount of speed exiting the cannon, so they used what looks like a negative inverse exponential with a linear asymptote to determine cannon exiting speed. As you might expect, this breaks down quickly when you traverse into negative speeds, and we end up with about -560 IV. Wait, IV?
Normally, velocity is capped at 120, however in the transition between cannon speed and exiting speed, there needs to be a grace period where this does not exist to make it look smoother (unlike when you enter the cannon). This grace period starts the instant the cannon is triggered, and ends when your airtime resets to 0, since there is no guarantee you will reach any point in the exact same amount of time.
So, we exit the cannon with this -560 IV and get shot immediately backwards, with enough speed to not only pass through the cannon exit wall in a single frame, but fly across to where KCP2 is located, skipping the entire post-cannon section.
As an extra note, you can only enter a new checkpoint if you exit the one youre currently in or adjacent to. This means we actually drive backwards over the cannon checkponts until we exit them, jump straight into CP23, and enter KCP2.
Want to see more TASes and TAS projects? Visit https://mkwtas.com
or the MKWii TAS discord, at https://discord.gg/NCZghgy")
