Uploaded June 2024 | Updated September 2026, 2 weeks ago
This video contains rapid pausing resulting in mild flashes, which may be uncomfortable for some viewers.
This is the abridged version of the Any% No A Button run. It has many cuts and skips to bring it down to a reasonable amount of time to view. This is a large part of why this took me so long to upload it all, so much apologies for the wait. There isn't anything super special or flashy about this run, it just cuts out a lot of the repetition, particularly for the failed attempts for some stars. I didn't include any of the failed attempts of the frame counting setups because if you don't know how each one works then it's not particularly clear how or why it's a failed attempt, but for everything else I didn't cut out any failed attempts, just the repetition part (i.e. getting back there) was cut out.
Unabridged version: youtu.be/tzMbBh9dTWA
And now for the description no one reads :)
Shortly after finishing my 2xA run last October, I claimed there were still a few things I wanted to look at, one of which being Chuckya Drop, which was the solution to one of the only two A presses in the run, the other being the obvious BitFS rising platforms on Wii VC, which I didn't use because doing so for anything other than a perfectly optimized run seemed pointless. The run including a 1xA BitS opener was already fairly impressive, but with how consistent I had gotten with the double misalignment opener, surely something could be done for chuckya too, considering its movement is entirely deterministic.
One thing I had asked pannen to do when he was looking for a double misa setup for me was to find a set of backup movement for me after I got the first misa, in case I failed the second misa in a way that kept me on the platform. Although he did find one, I decided not to use it for the 2xA run because I felt confident enough with the double misa, but the fact that he found one was good news for Chuckya Drop, as it requires Mario to drop down to turn chuckya. The one he found wasn't compatible with the turning, though, so I asked him to find one that was and he delivered.
Meanwhile, I would still need a way to get up onto the blue block, so I analyzed that and found that the blue block doesn't quite come all the way to the edge of the gray platform underneath. The perfect set of conditions for a glitchy ledge grab. Unfortunately it wasn't as simple as just punching off, but a short low magnitude side strain to get you off the wall was the only extra thing you needed. That takes away the requirement of hitting a misalignment on a moving platform, which was a HUGE relief. From there it was simply a matter of finding a set of inputs that got the chuckya to a good position.
After a little bit of digging, I found a sweet spot for the first phase and tasked pannen to find a set of inputs for the last phase, which once again he delivered. Over the course of the next two months I kept finding improvement after improvement that made the whole thing more consistent for me to do. One of those improvements was instead of planning to break out of the chuckya's grasp with a measly ~30% to succeed even if I did break free, I could get the chuckya to the edge and abuse the same mechanic that makes the throws a 50% chance in wet dry world. Suddenly it was seeming more and more feasible.
The biggest change, was allowing B to be pressed at the end of a set of inputs instead of at the start. While a seemingly small change, this actually meant I could use the same hand motion for both unpausing and repausing, which meant functionally it was no different than just a standard pause buffer, so long as my button inputs were synchronized. This was a game changer for consistency, and before I knew it my consistency with *the rest of Chuckya Drop after the 1xA opener* came to about 50%. Tack on a 56% for a successful chuckya throw, and you've got yourself an rta viable chuckya drop.
From there, all that was left was to remove the 0.5xA presses from the 2xA run, and to improve my execution of it to make it more watchable. That run pause buffered EVERYWHERE, needed to retry in so many places, it was honestly embarrassing. I simply wasn't gonna let this run be a repeat of that, so the additional 5 months between that run and this run was spent endlessly practicing, making things more consistent and removing pause buffers wherever I could do so, so long as it didn't take a significant hit to consistency, and this was the result. It's been a long time coming, but SM64 has finally been beat without the A button.
Now time to start practicing for 120abc...
This video contains rapid pausing resulting in mild flashes, which may be uncomfortable for some viewers.
This is the abridged version of the Any% No A Button run. It has many cuts and skips to bring it down to a reasonable amount of time to view. This is a large part of why this took me so long to upload it all, so much apologies for the wait. There isn't anything super special or flashy about this run, it just cuts out a lot of the repetition, particularly for the failed attempts for some stars. I didn't include any of the failed attempts of the frame counting setups because if you don't know how each one works then it's not particularly clear how or why it's a failed attempt, but for everything else I didn't cut out any failed attempts, just the repetition part (i.e. getting back there) was cut out.
Unabridged version: youtu.be/tzMbBh9dTWA
And now for the description no one reads :)
Shortly after finishing my 2xA run last October, I claimed there were still a few things I wanted to look at, one of which being Chuckya Drop, which was the solution to one of the only two A presses in the run, the other being the obvious BitFS rising platforms on Wii VC, which I didn't use because doing so for anything other than a perfectly optimized run seemed pointless. The run including a 1xA BitS opener was already fairly impressive, but with how consistent I had gotten with the double misalignment opener, surely something could be done for chuckya too, considering its movement is entirely deterministic.
One thing I had asked pannen to do when he was looking for a double misa setup for me was to find a set of backup movement for me after I got the first misa, in case I failed the second misa in a way that kept me on the platform. Although he did find one, I decided not to use it for the 2xA run because I felt confident enough with the double misa, but the fact that he found one was good news for Chuckya Drop, as it requires Mario to drop down to turn chuckya. The one he found wasn't compatible with the turning, though, so I asked him to find one that was and he delivered.
Meanwhile, I would still need a way to get up onto the blue block, so I analyzed that and found that the blue block doesn't quite come all the way to the edge of the gray platform underneath. The perfect set of conditions for a glitchy ledge grab. Unfortunately it wasn't as simple as just punching off, but a short low magnitude side strain to get you off the wall was the only extra thing you needed. That takes away the requirement of hitting a misalignment on a moving platform, which was a HUGE relief. From there it was simply a matter of finding a set of inputs that got the chuckya to a good position.
After a little bit of digging, I found a sweet spot for the first phase and tasked pannen to find a set of inputs for the last phase, which once again he delivered. Over the course of the next two months I kept finding improvement after improvement that made the whole thing more consistent for me to do. One of those improvements was instead of planning to break out of the chuckya's grasp with a measly ~30% to succeed even if I did break free, I could get the chuckya to the edge and abuse the same mechanic that makes the throws a 50% chance in wet dry world. Suddenly it was seeming more and more feasible.
The biggest change, was allowing B to be pressed at the end of a set of inputs instead of at the start. While a seemingly small change, this actually meant I could use the same hand motion for both unpausing and repausing, which meant functionally it was no different than just a standard pause buffer, so long as my button inputs were synchronized. This was a game changer for consistency, and before I knew it my consistency with *the rest of Chuckya Drop after the 1xA opener* came to about 50%. Tack on a 56% for a successful chuckya throw, and you've got yourself an rta viable chuckya drop.
From there, all that was left was to remove the 0.5xA presses from the 2xA run, and to improve my execution of it to make it more watchable. That run pause buffered EVERYWHERE, needed to retry in so many places, it was honestly embarrassing. I simply wasn't gonna let this run be a repeat of that, so the additional 5 months between that run and this run was spent endlessly practicing, making things more consistent and removing pause buffers wherever I could do so, so long as it didn't take a significant hit to consistency, and this was the result. It's been a long time coming, but SM64 has finally been beat without the A button.
Now time to start practicing for 120abc...


![SSL: 100 coins HOLP setup [RTA]
Setup found by @pannenkoek2012 after making and running a brute forcer to look for setups compatible with RTA inputs.
To see a proper execution of this setup, check out this video here: https://www.youtube.com/watch?v=oDBE-BDNFRc SSL: 100 coins HOLP setup [RTA]](https://i.ytimg.com/vi/rpjTm8i42vo/mqdefault.jpg)
![WDW: Express Elevator - Hurry Up!! 0xA 4:26.88 [RTA]
Apologies for the double audio, there was an issue with OBS that I have since fixed.
This is neither an A press improvement, nor a community time improvement. Its actually not even a PB, my PB for this star is 4:03. However, what this is is a consistent way of doing the transport CWT required for the star. See, the way it works is...
Theres a hidden mechanic for Chuckya. It might have been to try to prevent throwing Mario OOB, it might have been a failed attempt to prevent it from throwing Mario off of a ledge, Im not sure. I cant retrospectively read the developers minds. But if theres no floor within 50 units vertically of a point exactly 150 units in front of Chuckya, then it will not throw Mario. It will instead continue turning (unless it has already continued turning in this way 15 times). This means if there is a floor at *some* point in front of it, then there are 16 out of the 32 possible directions it can throw Mario which will successfully find that floor. With careful planning, we can abuse this to take what should be a 6.25% chance of a successful throw and turn it into a 50% chance of a successful throw. Theres a catch, though. The throw angle to have Mario land on the plank is so precise that the amount he turns every frame can actually skip over the angle range we want, meaning that theres effectively a human element of randomness to the throw as well. This is simply unacceptable.
So what I did is, with a bit of a change to where I initially grab the Chuckya, I managed to find a setup that gets Mario and the Chuckya BCA, which means that I will *always* [have a 50% chance to] get the correct angle on the throw. This not only increases the overall success rate of the transport CWT, but just in general makes it less frustrating to do.
The setup is, theres a seam where the floor triangles meet. Stand on the point of that line where it meets the water, when the water is at the bottom of its oscillation, then go c right and c left to reset the camera. Do a tiny movement straight forward to change Marios facing angle to be in line with the camera, then go into Mario cam and follow that up with fixed cam. Press c left once, wait for the camera to come to rest, then walk forward for 6 frames. Once Mario comes to a stop, walk backwards for 6 frames. Press c right 3 times (or spam it until the camera buzzes) then wait for the camera to come to rest. Press b and start on the same frame, then do as short of a pause buffer as you can, and finally unpause and wait for Chuckya to throw Mario. If done correctly, there will be a 50% chance of Chuckya throwing Mario in the correct direction. If you fall short of the plank, then you need Chuckya to be further to his left when you initially grab him. If you land on the plank but then fall off before you regain control of Mario, then you need Chuckya to be further to his right when you initially grab him. WDW: Express Elevator - Hurry Up!! 0xA 4:26.88 [RTA]](https://i.ytimg.com/vi/sHkB6m8X6Dw/mqdefault.jpg)
![HMC: Watch for Rolling Rocks 0.5xA Faster PU Route [TAS]
I collect the star Watch for Rolling Rocks in 0.5x A presses, in 1°491677. This is TAS. This is a timesave over the previous record of 5°262210 by @pannenkoek2012 which can be found here: https://www.youtube.com/watch?v=mgNMKnkZEKY
A few months back I had theorized that instead of Hyperspeed Walking up to the syncing speed of the island triangle and then doing the standard PU routing of waiting out the speed loss for the further syncing speeds of each triangle, you could instead just use jump kicks to ascend the hill between the lake and the elevator. The previous route used triangles that had descending syncing speeds, which allows you to ascend up to 312 units in a single frame, but here I simply use jump kicks, which always have a syncing speed of 262144 as its air movement. The catch is you can only ascend 98 units every 4 frames at best, although in practice youll only be able to ascend 83 units roughly every 20 frames. One major benefit of this change in strategy, however, is that you can get yourself QPU aligned with the landing frames, whereas with the other strategy you must keep yourself QPU aligned throughout the entire strategy, or it causes a console crash upon trying to move partial frames.
I gained a newfound appreciation for this star when trying to do the final movement, though, that section is an absolute nightmare. Now remember that when Pannen did this all those years ago, he didnt have STROOP. STROOP wasnt made yet. He did this entirely with ramwatch and a calculator. Absolutely insane. HMC: Watch for Rolling Rocks 0.5xA Faster PU Route [TAS]](https://i.ytimg.com/vi/sKytFn1EgZk/mqdefault.jpg)
![SMSR: Bowsers Rainbow Rumble Clear 0xA Without Preset HOLP [TAS]
We collect the star in the overworld after clearing Bowsers Rainbow Rumble in 0 A presses. This is TAS. This is an improvement off of the previous record of 0 A presses, which was done by me and can be found here https://youtu.be/hOF0fDg47Kc
Note that the start of this TAS is the same as the 1xA variant done by lemon, and can be found here https://youtu.be/JZrxorc5BtI
Since then, there had been discussion about a 65 star run potentially being possible, as a Glitchy Ledge Grab to clip past the 80 star door to get to the warp was known to exist, and there was research being done on whether or not it could be done 0xA. Doing this, unfortunately, was assumed to need to use the crazy box to get up to where a misalignment exists. This turned out to not be necessary, and maybe Ill make a video on that soon too, but I decided to look into doing this star HOLPless regardless. This had already been looked into a bit by lemon here https://youtu.be/JZrxorc5BtI but neither lemon nor myself were able to figure out how to get past the last brown block. I had a couple ideas using CWT, but nothing panned out due to needing to not get thrown too hard, lest you fall off the block from the inside of the platform.
The main issue with it is that the ledge *after* the block is too high to dive recover onto, and theres no accessible misalignment, which is why you need to get on the inside of the tower. From there, its simple to do a dive recover and clip out once youre high enough to land on the floor. So when revisiting this, I picked up mentally where I had left off before around a year ago, and had the idea of using a type 2 CWT, otherwise known as a throw arc CWT, coupled with a chuckya escape. Chuckya escape solves the issue of getting thrown with too much speed, but it has the consequence of not getting as much height out of the throw, so in theory getting the extra height from the trajectory the chuckya takes from being thrown could cover that extra height. And as it turns out, it indeed does solve the issue, getting Mario to the inside of the tower on the brown block. From there, finishing out the level is trivial. SMSR: Bowsers Rainbow Rumble Clear 0xA Without Preset HOLP [TAS]](https://i.ytimg.com/vi/tmF9Nj4Ej7c/mqdefault.jpg)
![SM64: Unlock Upstairs 1xA in 36:14.09 [RTA][WR]
Been working on making this category a bit more polished. We have this as a category for a short one, since theres no way to beat the game with fewer than 70 stars, meaning this is our 16 star category. Timing ends on X cam for door touch.
Join the discord: https://discord.gg/73JKC6f SM64: Unlock Upstairs 1xA in 36:14.09 [RTA][WR]](https://i.ytimg.com/vi/uNT1SlW1br0/mqdefault.jpg)

![DDD: Board Bowsers Sub 0x [RTA]
I found a setup that goes as follows: Actively manipulate the camera so that its as close to directly behind Mario as possible. Sidle against the wall, the closer to the ledge the better since time is of the essence. Neutral joystick, then with the camera directly behind Mario, walk straight forwards and press Z. If the camera isnt directly behind Mario, youll have to adjust the joystick slightly. By walking straight forwards and pressing Z, youll go into crawling immediately. Turn to the right for 3 frames, then stop and start crawling to give the camera a moment to catch up. Crawl right up to the ledge, which will be when you can see half a fist sized bit of water between Mario and the cork box. Stop crawling, then do a standing punch. Neutral joystick for 2 frames, then full left strain for 2 frames. This gets Mario in the misalignment every time, and even with a little bit of wiggle room for if you didnt get the camera movement perfectly right.
Some people have decent luck with just yoloing it. I absolutely d0 not, and can not do this without the setup. DDD: Board Bowsers Sub 0x [RTA]](https://i.ytimg.com/vi/v4EHFPlAWbk/mqdefault.jpg)
![Pannens Ice Hack: Coins Amassed in a Maze + 100c [RTA]
@pannenkoek2012 made a romhack that replaced all of Marios movement with ice physics. Initially intended for a 16 star run, I decided to take it a step further and go for max stars, and this is easily one of the more frustrating stars. Pannens Ice Hack: Coins Amassed in a Maze + 100c [RTA]](https://i.ytimg.com/vi/vMOIuae_5D0/mqdefault.jpg)
