(World Record) Unlock Upstairs 1xA with 16 Stars in 35:59.78 yellowdice 2026-07-14 | I did it! 16 star Unlock Upstairs record, even with slope grind disadvantage! I thought this run was so dead so many times, to the point that I had mentally settled on having to do more attempts, but somehow, I'm done! Honestly, it's extra satisfying to see the time end with a 59 too. Wasn't aiming to break a minute barrier, but it's really cool that I did.To elaborate on what I mean by "slope grind disadvantage" or why the Dark World split is named "guaranteed timeloss :((((((", this is because the previous record in this category of 36:14 by Marbler was accomplished on controller (and on N64 too I believe). With access to more than 8 angles, this allows for the slope grind in Dark World to take around 2 minutes, whereas on keyboard, which is the only input device I have access to, I'm looking at a minimum of 4 and a half minutes for slope grind. While playing on keyboard advantages me in several other locations, namely the start of CCM, where I have a much easier time framewalking, playing on controller definitely still nets a significant time advantage overall.As you can tell from my splits, my previous PB in this category was a 37:08 (or maybe it was a 37:07, I'm not really sure since I'm splitting manually and I never framecounted my previous PB), which I achieved a little earlier today. In a way, this run was better played than Marbler's 36:14 due to slope grind disadvantage, but it's obviously still much slower. That run lost at least 2 minutes in SSL due to messing up the Klepto star, so I knew WR was definitely within reach, despite slope grind disadvantage. This run is fairly similar to my 37:07/37:08 up to the end of Dark World. I gained time on Cannonless, lost it back on CCM 100+Reds, gained time on Wall VSC, lost it back on Pipe Misa.I got some of the timesave I was looking for in SSL, but as you can see in the run, I still messed up Klepto 1 time, so there is still more time left on the table. However, once I died on Boil the Big Bully, I considered the run to basically be dead. I knew going into DDD that I needed a first try Sub split for a chance to eke out WR, and somehow that's what I got, for the 2nd attempt in a row. The small bit of Fire Sea practice I did earlier today clutched up, as apparently I had more timesave there than I had previously recognized. I was honestly really surprised seeing myself over a minute ahead of my last PB out of BitFS, and it was only at that point that I knew that I had gotten record lol.Anyways, this means that I have both Unlock Upstairs world records now! Getting the 16 star record was something I'd thought about in the past, but I only actually committed to after getting my 29:48 in the 9 star category, where you do 0xA Crackslide instead of all the basement stars. And, fun fact to end off the description, both the 29:48 and this run were on my 27th ever attempt at optimizing the category, which I think is pretty cool :)
(World Record) Unlock Upstairs 1xA in 29:48.23 yellowdice 2026-07-10 | After the discovery of Crackslide on the 22nd of November in 2025, it was immediately speculated that the trick could be implemented 0xA, thus allowing for a 0xA 30 star door skip which didn't require MIPS. Zyon would prove this possible RTA just 2 days later on the 24th, using a setup he and sapling had found.Soon after, I set out to become the first person to complete an Unlock Upstairs 1xA run with only 9 stars, where I would get the first 8 stars normally, beat Dark World 0x, then perform 0xA Crackslide and finish off the run with DDD Sub 0x and Fire Sea 1x. This accomplishment would happen on the 27th of November, 2025, and the first run of this category would be just under 2 and a half hours long. As of right now, the Unlock Upstairs 1xA record for 16 stars is 36:14, which was set by Marbler in July of 2024. Obviously, this was indicative of just how much a time of 2 hours 27 minutes could be cut down.When attempting All Signs 4xA a little over a month ago, I had inadvertently beaten the Unlock Upstairs 1xA record with a time of a little over 46 minutes. This 46 minute time included multiple deaths as well as a lot of time lost from reading signs. At this point, I decided I would eventually commit myself to taking this time down as much as I reasonably could.My grind would occur in mid June, when I was able to get a 39:00, 39:07, then finally, a 33:10, which officially made 9 star the faster category (as you would naturally expect it to be). This was despite the 2-3 minute time deficit I automatically incur during the Dark World slope grind from playing on keyboard in lieu of controller, but 9 star was fast enough of a category that I was able to end up with a faster time anyway.I would sit on the 33:10 for a bit before I came back to this category last night, since I knew there was still a lot of time that could be shaved off. The longshot goal was sub 30, but I was mostly just looking for a cleaner run. It took a few attempts to finally make it out of Dark World again, but when I did, I guess I managed to make it count.I'm pretty glad that the final time ended up starting with a 2; it would've been really frustrating if I had ended up somewhere in the low 30s. However, there is still a lot to be desired. Specifically, barely missing my first Bowser 2 throw led me to losing almost an entire minute, so this run should very well have been a 28. Nonetheless, this is a tough target to top and I did achieve what I initially set out to, so I will consider myself done with this category, at least for now.There's probably at least 2-3 more minutes to be taken off this category from playing better, and another 2-3 minutes can also be shaved off with controller slope grind. There were a couple sloppy moments in the run - Pre-Dark World wasn't super clean, I spent a long time on Dark World VSC, Crackslide could've been done faster, that DDD star grab was kinda embarrassing, and obviously that Fire Sea split was just horrendous. Of course, I also have to mention the actually clean parts of this run. Specifically, I somehow hit both pipe misa and DDD glgs first try, the latter of which I barely managed to squeak out.Anyways, if you've read this deep into the description, hi :)
(World Record) Cruiser Crossing the Rainbow 1xA in 2:00.96 yellowdice 2026-06-20 | Wow, I have a Rainbow Ride IL record now. That's pretty cool. This was my first attempt that made it onto the final carpet ride, and I managed to clutch the final edge DR out.You could probably take this down another second to break the 2 minute barrier with a better flyguy twirl at the start which lets you land outside of the wooden cage, instead of having to build speed using an unreferenced wall like I did during this run. Regardless, I'm still pretty happy with this time.Previous World Record of 2:03 by Ex320kyt: youtube.com/watch?v=wU2Ld76jKF8
(World First) All Signs 4xA in 2:02:36 yellowdice 2026-05-31 | The first ever completion of an RTABC All Signs run, done in the optimal amount of A presses! This run also beats the previous record of Unlock Upstairs 1xA with 9 stars by over an hour and a half, even though I lost a bunch of time there to reading signs. I couldn't really have asked for more with this run.Just to clarify, by "the optimal amount of A presses", I mean the lowest possible amount that can be achieved in an RTA setting as of now, ignoring WiiVC Platform Raise in Fire Sea. Out of the 4 A presses in this run (1 to get past the pole in Fire Sea, 1 to enter THI, 1 to reach a sign in JRB, and 1 to reach a sign in HMC), only 1 of these A presses is strictly required in a TAS setting at the moment (the one in JRB), as the other 3 can be skipped with WiiVC Platform Raise (or perhaps Bully Battery in the future), mini goomba manipulation, and swooper manipulation respectively.For anyone interested in the backstory, I noticed yesterday morning that the only category on the full-game tab on the RTABC spreadsheet without an existing entry was All Signs. Thus, I decided I would set out to be the first to complete an RTABC All Signs run, and I didn't really see a point in doing so if I wasn't going to aim for the optimal score. The main issue that this entailed was Cavern of the Metal Cap entry, which required either a really precise Hyperspeed Walking setup which notably includes a 20 frame left input with no clear visual cue, something that took me hours to do even once in the past, or a frustratingly difficult rollout off of Dorrie's head, something that I hadn't even ever attempted before.I had heard that the latter option was so horrendous to deal with that it was basically RNG. There is no single component of it that is random, but getting everything to work really isn't easy. So, over the past two days, I threw myself at this trick for a fair bit before I garnered at least some semblance of consistency with it. I also had to derust every other trick in the run - Wall VSC, Seesaw Framewalk, Pipe Misa, 0xA Crackslide, DDD GLGs - though those didn't take nearly as long, since I already had experience with them.All in all, this run went past so much quicker than I expected. I came into this with a whole free afternoon and evening, I was ready for it to take like 7 hours. Somehow though, the pieces didn't end up taking too long to align - Dark World had some significant timeloss from Seesaw Framewalk, even though my Wall VSC game was on point at the start, but Crackslide and DDD were both really clean.The real scary part of the run was THI, since a premature death automatically incurred an extra A press, and I had to be holding A for like the whole time. The next stressful section was Island Hop, where getting to hold A here made it so much easier than what I had to go through in Max Stars, and I'm really glad I hit it first try and didn't have to go through the agony of holding A any longer.Somehow, Mips was the real bossfight in this run. I did a tiny bit of savestated Mips practice right before the run, but I guess it wasn't enough, cuz getting into HMC ended up taking longer than both Dark World and Dorrie DR. I'm really glad resetting the game doesn't save the fact that I had grabbed the star, otherwise we would've been in a bit of trouble.Finally, there's Dorrie DR. After hitting that final sideflip in HMC, all the nerves were gone, and I knew it was just a game of patience and endurance, whether the rollout would work before I mentally collapsed. And it turns out, the race wasn't close at all - I don't know if it was a fluke or whether what I was doing is actually just semi-consistent, but by some miracle, it only took 15 minutes to make it into CotMC and finish the run.Anyways, this has been a pretty lengthy description, if you're reading this I presume you've probably just scrolled to the bottom. I doubt anyone's going to actually read through everything I've wrote, but this video is taking a moment to upload and I figured it wouldn't hurt to document all my thoughts somewhere.
(World Record) Behind Chain Chomps Gate 0xA in 38.06 yellowdice 2026-05-08 | what just happened lmaothe first rtabc IL I ever tied the A press score on, and now I have record! I thought the previous wr of 41 seconds by trekq was already insane enough, I don't really know how I saved 3 and a half seconds over it tbh lol. 3 frames short of the 37 tho, pretty unfortunate that I'm just at the cusp of another second barrierPrevious World Record of 0:41 by TreKq: youtube.com/watch?v=7IcuY8Zjtfg
(World Record) Secret of the Haunted Books 0xA in 1:41.63 yellowdice 2026-04-22 | I'm really happy with my execution overall throughout this run. I had probably at least a dozen and a half attempts into the final room before I managed to grab the star first try, and I'm really glad that the attempt I eventually succeeded on just so happened to be my fastest pace ever.There really isn't too much timesave over the course of the run, in the sense that I didn't necessarily make a glaring mistake during the attempt. Any future records from here on out will probably just be cleaner or tighter execution throughout or maybe some new strategy that may be found some time down the line.Honestly, my goal for this star was just to beat the previous record of 1:50.16 in general, since it no longer has a video attached to it. I figured sub 1:45 would be nice, but 1:41 was definitely way beyond my expectations lol. I don't deny this record can be beaten, but I imagine it'll stay on top for at least awhile.Previous World Record of 1:50 by Ivan178: youtube.com/watch?v=wRHgoQWxsyM
Getting on the Starting Cork Boxes in Bob-omb Battlefield without Pressing A in 6 Different Ways yellowdice 2026-04-07 | Let me know what y'all think of this kind of video. I just randomly had the idea to make this today and figured it wouldn't take too long to do, so I really didn't have much to lose. I doubt this video will garner much traction but I still think it's pretty cool.In case anyone watching this isn't too well versed in the dark arts of SM64, here's a more detailed explanation on each of the 6 methods:1. MisalignmentThis is a 1x1 unit of floor triangle at the cork box's corner which extends 78 units further down than usual, thus allowing a rollout to ascend 78 more units than usually possible. A misalignment can be found in the corner furthest away from the map's origin and is a result of walls and floors having different methods of truncation.2. Vertical Speed ConservationThe sign next to the cork box interrupts Mario's rollout while he is moving upwards. Thus, his vertical speed at that moment is positive. By refraining from touching the joystick and moving forward with only punches, this positive vertical speed is stored. After punching off the sign, Mario briefly falls upwards, thus allowing him to be airborne for long enough to ledgegrab the floor of the cork box.3. Goomba BounceWhen you rollout onto a goomba, you bounce. That's kinda it.4. Edgegrind RolloutAfter walking off the sign, I immediately dive once Mario's airborne. By then holding left to strain back onto the sign, I instantly land, thus allowing me to do a rollout on the sign and reach the cork box.5. Goomba CloneI set the HOLP nearby the cork box, then pick up two wing caps in quick succession to achieve Hat in Hand, where I'm able to pick up objects without overwriting the HOLP. I then grab a bob-omb just after it explodes to hold a vacant object slot, load a goomba into said object slot, then release it at the HOLP for me to bounce on.6. Star Dance ClipI repeat a similar process to the goomba clone, but this time, with a coin. Then, after collecting 99 other coins throughout the course, I sidle against the cork box's wall to collect my 100th coin, thus spawning the 100 coin star next to the cork box. By carefully rolling out into the star next to the cork box, I enter the star fall action, which, unlike the rollout action, allows me to ledgegrab. Thus, Mario is able to ledgegrab the floor of the cork box and snap up to it following the star dance.
[TAS] Rainbow Ride 100+Reds 14xABZ yellowdice 2026-04-01 | Skip to 8:43 for pedro spot exitIn this TAS, I collect the Rainbow Ride 100 coin and red coins star in 14 total A, B, and Z presses. With current knowledge, this is the lowest possible amount of ABZ presses RR 100+Reds can be completed in. This TAS has been finished for like half a month now, but I just never bothered recording and uploading it until now.Also, for anyone wondering, the frame in the thumbnail is from a previous iteration of this TAS that collected the two stars 15x, but since the creation of that TAS, an optimization has been found to lower the score to 14x, which also altered where I collect the 100 coin star in the route.
(World Record) Shoot to the Island in the Sky 0xA in 2:31.63 yellowdice 2026-03-20 | -1:01.50 over the last record, hell yeah :Dbrief rundown of the route, for anyone not bothered skimming through the vid:- I set the holp beforehand then instant release a bobomb- I grab another bobomb bloat 3, regrab it hands free using an open cannon, then groundpound to stall and land in a warp- the bomb is now in limbo and constantly pushes me backwards, and I make my way up to the summit- I slidekick off the edge while the bomb continues to push me back and groundpound to stall in the air, making it onto the island- I get rid of the handsfree bobomb and use an out of bounds failsafe mechanic to reactivate the instant released bomb on the island- I grab it fake off the island's edge and clone a goomba- I release the goomba using bca and bounce on it to grab the starshoutouts to ivan, who idk if he's even around anymore, for inventing ivan hop and holding the previous record, crisk for his 2 year old 1:58 theory tas that they've probably forgotten about, and auto-scout for unintentionally informing me that you can take the warp up the mountain which they probably don't know they didalso, if you've read this far down in the description, leave a comment suggesting what star I should do next :)Previous World Record of 3:33 by Ivan178 (vid no longer exists): youtube.com/watch?v=WlpmaiA5ix8
(Former World Record) To the Top of the Fortress 0xA in 3:17.16 yellowdice 2026-03-01 | Current World Record of 2:57 by TreKq: youtube.com/watch?v=3wmOIOBllWkThe previous record for this IL of 3:38 was featured in the pannenkoek tower ABC video, which I watched months before I even began touching the challenge at all, so it's pretty damn cool to be the one holding the record now.This took somewhere between 5-6 hours to accomplish. This included 14 total attempts that made it to the platform release, 9 of which died to the platform tilting the wrong way, 3 of which died to not making it onto the tower, and 1 of which I just choked.This record still has a fair bit of improvement left. Just by my rough estimates, sub 3 is probably reasonable. With tighter movement, you can save around 13 seconds by catching an earlier cycle on the wooden rotating platform. Additionally, while climbing the tower at the end, you can perform 2 ledgegrabs per cycle, which is something I didn't go for here because I didn't feel like choking another attempt in the endgame.Previous World Record of 3:38 by TreKq: youtube.com/watch?v=PvlDhwBAVCg
Kaizo Mario 64 0 Star in 7:29 yellowdice 2026-01-10 | Kaizo Mario 64 is a 2009 romhack of Super Mario 64 that basically makes the game much harder. I first played this romhack in late September of 2025 where I spent around an hour and a half getting to the top of the endless staircase with 1 star before finding out that the entrance to Bowser in the Sky was just not there. Turns out, the intended way to beat this romhack is to collect all 120 available stars and unlock the cannon in the castle grounds, where there is a pipe that sends you to Bowser in the Sky. In short, I never got the satisfaction of beating the romhack, and that didn't sit right with me.I decided to revisit this romhack around 2 days ago and took a look at Hyperspeed Walking in the castle grounds. After a bit of playing around and testing, I eventually got Mario to start building up speed, which I could then preserve with a frame perfect crouchslide out of a swimming state to clip below the cannon's grate and therefore into BitS.I'm pretty daft when it comes to the technical aspects of SM64 so the "setup" I used here is literally just to HSW for somewhere between 15-16 seconds, yolo the angle for the reactivation, pause buffer the frame perfect Z press, hold up right, and dive once he's in the cannon hole. This wasn't actually the most inconsistent thing in the world, and combined with the fact that I get a shot at this every 3 or so minutes, clipping into BitS really wasn't the worst thing I've had to go through.Sky itself is a really tough stage, especially since I get exactly one shot at it, in addition to the fact that I'm not allowed to take literally any form of damage whatsoever. It also doesn't help that there are extra firespitters and amps everywhere, a chain chomp right before the pipe, and of course, most the stage is invisible as well. Somehow, despite a bunch of practice, I managed to die in BitS a total of 11 times. On my 12th attempt into Sky, I finally made it into the pipe, where Bowser was nice enough to give me a 1/8, letting me cruise to the finish. Overall, I played this run pretty safely and therefore pretty slowly. It's nowhere near the fastest run out there, as there's apparently a 4:08 on the leaderboards, which sounds like a time you'd have to play unnecessarily risky for. Nevertheless, I can finally be done with this project and move onto other things.
Max Stars 0xA - 41 Stars yellowdice 2025-12-31 | I've never come anywhere close to doing anything for 11 hours straight, so it's pretty safe to say that this was one of the most mentally taxing experiences I've ever had. This was definitely a run I underpracticed for, but nonetheless, I somehow managed to pull through and get the job done. This was a pretty epic way to end off the year, and start it too, because this is what I was doing as 2025 turned into 2026 lol.As of right now, this run puts me in 2nd place on the Max Stars 0xA leaderboard, way behind Marbler's 49 and just ahead of Zyon's 40. I have no plans on improving this anytime soon, but if I ever do, the next thing to add is definitely Metal Cap Entry, which unlocks Metal Cap Reds, Whomp King, and Whomp Tower, or maybe even JRB Jet Stream if I ever learn the framecount for that.I doubt anyone's going to be reading through this, but regardless, here's every star in the run and my thoughts and comments on each:1. Chain ChompI think this is the first full game run to have Chain Chomp as the first star of the run. Bomb glide is such a sick way to start off the run, and I'm pretty happy to have gotten it done in just 10 minutes. With how consistent it can be, I could definitely have gotten this first or second try though.2. Bomb King and 3. Koopa the QuickNot much to say here.4. ShipI've done this star literally once before, and it took 26 minutes, so I was pretty glad to have finished here in under 10.5. EelI totally expected this to take at least a couple tries, each of which would be a pain as I'd have to swim to the back of the kingdom each time, but fortunately, I lucked out and managed to get it first try.6. JRB RedsI did not expect this one to be a major hurdle. That jet stream really screwed me over, huh.7. CannonlessSince I'm only allowed to have 1 joystick magnitude, I can't reliably do texture setup, so I instead just did an iffy framecount, which eventually worked.8. Whomp's 100 and 9. Whomp's RedsEmbarrassing how many times I fell down from the floating isle lol.10. Wild Blue and 11. PSSFreebies :)12. Wing Cap RedsI didn't know you couldn't rollout into the star lol. Also that switch press took much longer than it should've.13. Dark World RedsThe whole name of the game here just ended up being refiguring out how to do seesaw framewalk. Also, big shoutouts to the 4.5min slopegrind, which made hitting the misa much less stressful.14. BOB Reds and 15. Island in the SkyThese were the stars I feared the most. Island in the Sky was the only one I would've been unwilling to retry had I failed the first time round, and perhaps the hardest out of all 41. It's genuinely a miracle that this all worked out, and these 2 stars are the primary reason that I have no further plans for Max Stars runs.16. BOB 100Got to use a cloning "setup" that I found on like day 1 of me doing RTABC ILs, which is pretty sick.17. SSL 100Even with a late-ish death, I'm still somewhat satisfied with my performance on this star, at least compared to when I was trying this as an IL. Let's just say that it could've been much worse.18. Talon and 19. PyramidMore freebies :)20. Hot-Foot-ItI had an attempt that should've had it literally 7mins in, but I choked a pause on the final lavaboost, which was pretty unfortunate. I had another lavafall choke a bit later on before I finally closed out an attempt on my 3rd dustless rollout that managed to trigger the heart.21. LLL 100, 22. Bullies, 23. Bully, 24. Log Rolling, and 25. LLL RedsAll pretty easy stars, yet I still somehow managed to fail LLL 100 for 15mins.26. Vanish Cap RedsAnd this is where it all fell apart. On my very first attempt, I hit the switch, then fell off trying to get the Vanish Cap, and it took me another hour and a half to get back to the ending and close it out. I literally got so bored halfway through I had to put a YouTube video on in the background.27. Mips, 28. Toad, and 29. Swimming BeastI'm happy with how quickly I got the SDC into HMC here, I definitely did better here compared to practice attempts.30. Haunted Books and 31. Ghost HuntThis was the part that sucked the most. Haunted Books literally took me only 15mins to get the first time I tried, yet somehow, everything kept falling apart here. I basically had to relearn staircase VSC, which was being really uncooperative for some reason.32. BBH 100Somehow first tried the bookend bounce. We take those.33. Bowser's SubI expected this to be a major hurdle, but somehow, I managed to get this one done in under 10 minutes. Finishing this so quickly was an incredibly morale booster that carried me through to the end of the run.34. MantaDidn't expect this to take too long, and for once, it didn't. We love to see that.35. Wall KicksI was mostly just winging this one, so I'm glad it didn't take too long to work.36. Slide, 37. Baby Penguin, 38. Penguin Race, 39. CCM 100, 40. CCM Reds, and 41. SnowmanMore freebies to end off the run :)
(World Record) Great Tiger in 29060 points yellowdice 2025-12-18 | Run begins at 2:15It's been awhile since I've last really played this game, but D-man decided to remind me of the 300k SOB goal the other day, and I thought this would be a fight where I could pretty easily tack on some points. So I went ahead and looked for a new strat with a higher ceiling, and eventually came up with this one. The odds are the exact same as last time (around 1/10), and the execution really hasn't gotten harder either, so I knew this wouldn't be the toughest fight to finish.With that said, I do believe that, for now, 29060 is as high as it gets without implementing a select fill (which doesn't even actually benefit that much) or frame perfect 19 damage star punches (which is the main difference maker between this fight and the TAS score). I'm pretty satisfied with this run overall, this grind did take me a bit longer than the last one but at some point I finally got the hang of the execution and stopped making mistakes. I blame rust on that one.Also, this is the last 1000 barrier on this fight, probably ever, so that's a pretty cool milestone. Don't expect more MTPO activity from me any time soon though, I'm still largely focused on SM64 right now.
[TAS] Nightmare Punch-Out!! in 51:40.58 (Obsoleted) yellowdice 2025-11-02 | This TAS is (finally) obsoleted! A 41.97 is possible on Glass Joe with rapid punch storage.Preliminary TAS of an MTPO romhack known as Nightmare Punch-Out. The clock gets a bit janky at times, and the fighters are in a strange order. Don't expect all of these fights to be fully optimal.This TAS is optimized in terms of IGT, and sort of optimized in terms of RTA.Here are the times achieved in this TAS for each of the 14 fights:Glass Joe: 0:42.00Von Kaiser: 1:45.61Super Macho Man 1: 1:02.00Don Flamenco 1: 0:28.25Piston Honda: 6:19.97King Hippo 1: 6:51.48 (The time glitches out on the bluescreen, but by looking at the relevant ram address, I can decipher that the final time is 6:50.139, which I've converted to 6:51.39, which rounds to a time of 6:51.48.)Mr. Sandman 1: 8:06.00Don Flamenco 2: 1:14.25Mr. Sandman 2: 2:15.25King Hippo 2: 6:51.48 (Same situation as King Hippo 1.)Great Tiger: 1:49.99Super Macho Man 2: 1:35.82Mr. Sandman 3: 8:06.00Mike Tyson: 4:32.48Total Time: 51:40.58Please let me know if you find an improvement to any of these times.
[TAS] Mike Tysons Punch Out!! Low Score in 35940 Points yellowdice 2025-09-19 | This TAS is optimized in neither RTA nor IGT. Clockstop Glitch is banned. Here's a list of the score on each fight for anyone who is too lazy to keep track:Glass Joe - 1010Von Kaiser - 3340Piston Honda 1 - 3030Don Flamenco 1 - 2300King Hippo - 1200Great Tiger - 3030Bald Bull 1 - 3170Piston Honda 2 - 1010Soda Popinski - 3500Bald Bull 2 - 1010Don Flamenco 2 - 3250Mr. Sandman - 3510Super Macho Man - 3380Mike Tyson - 3200I'm pretty confident there are no more improvements to be made to this category, so I figured I'd upload a full TAS of it to the internet. If you have any questions, or have somehow managed to find an improvement to one or more of these scores, feel free to leave a comment below.
100% Consistent Blindfolded Tyson Buffer yellowdice 2025-08-02 | No matter what RNG Tyson gives, this strategy will guarantee a TKO at 5:33 every single time, as long as you don't make a mistake. Phase 2 is routed in such a way that regardless of refill and pattern, Tyson will always be knocked down at the same time. I did not perform this fight blindfolded as I don't know how to buffer punches without looking.
(World Record) Bald Bull 0 in 43.97 yellowdice 2025-04-26 | it is 4:21am as I am typing this so I do not have the mental capacity to type out a thorough explanation in the description. hopefully I'll be motivated and bothered enough to come back and edit this in the future so that informative information will actually be present herebut basically I just did a modified version of the tas strat that requires a 2 frame up press and a 1/4 refill and a gut punch which I really don't know the odds of. and this is perhaps the last record ever for this fight, unless someone is crazy enough to perform a frame perfect guard raise into a special reverse max damage star punch which sounds like a term I made up, probably because it is sorta a term I made upthe last record didn't last long I guess
[TAS] Bald Bull 0 in 43.61 yellowdice 2025-04-23 | I said I was gonna upload this yesterday and didn't despite the fact that I had the vid and thumbnail ready just because I was too lazy to type up a description and I'm realizing that I'll probably be too lazy to do so for the rest of today as well so I'll just edit this description to have actual info and details and stuff whenever I feel like it lol
(Former World Record) Bald Bull 0 in 46.61 yellowdice 2025-04-21 | For the uninitiated, Bald Bull 0, or Demo Bull, is the fight that plays on the title screen if you wait for too long, and can only be accessed with the assistance of Game Genie codes. The previous world record was held by @yourlocalpizzaguy with a time of 48.61, meaning this fight is a whole 2 second improvement!This fight utilizes a new strategy I came up with earlier today, and has total odds of 1/16, consisting of a 1/4 gut punch at the start of Phase 3 and the 1/4 Low Refill. The TAS of this strategy also gets a 46.61, meaning this fight ties the strategy's limit. The only point where time can be lost is the second facejab in Phase 1, which must be delayed a minimum of 1 frame. Everything else in this strategy is buffered.I'm pretty satisfied with both the strategy I came up with and the fight that I managed to execute. I don't think it's too hard to tie, the 1/16 luck is probably the biggest barrier, but beating it would require a whole new reroute, and there is no currently known strategy that is both faster and humanly viable.
[TAS] Mike Tysons Punch-Out!! 2015 Remix in 13:27.83 (Obsoleted) yellowdice 2025-03-22 | This TAS is now obsolete :)Note that the run is timed in IGT (In-Game Time) and not RTA. This TAS is not optimized for RTA timing. A lua script, which provides a bunch of additional useful information, was enabled during the playback of this TAS. This does not directly affect the TAS in any way.Three days ago I made my first TAS of the Mike Tyson's Punch-Out romhack 2015 Remix, which had a final time of 13:48.78. Since then, 7 out of the 14 fights have been improved, lowering the final time by over 20 seconds.Let me know if you have any other ideas for things that I can try TASing, as I'm still kind of burnt out from the highscore TAS at the moment and am finding new things to look into. Also, feel free to ask any questions you have about this TAS in the comment section, there's probably a lot to unpack.Here are the splits for each fight:Glass Joe - 0:42.00Von Kaiser - 0:37.61Piston Honda 1 - 0:43.25 (-1.23)Don Flamenco 1 - 0:11.25King Hippo - 0:35.25 (-2.36)Great Tiger - 0:34.97 (-1.03)Bald Bull 1 - 0:54.00 (-1.99)Piston Honda 2 - 0:47.61 (-1.21)Soda Popinski - 0:55.48Bald Bull 2 - 1:02.82Don Flamenco 2 - 0:50.25 (-2.00)Mr. Sandman - 2:14.61Super Macho Man - 0:43.25Mike Tyson - 2:35.48 (-11.13)Final Time - 13:27.83 (-20.95)The time written within brackets indicates the amount of improvement on that specific split. Any splits with no corresponding bracketed time are unchanged compared to my previous TAS. Courtesy to FartButt97 for the strats behind the Hippo, Bull 1, and Don 2 timesaves.also yeah I just reused the thumbnail and like half the description what do you want from me
[TAS] Mike Tysons Punch-Out!! 2015 Remix in 13:48.78 (Obsoleted) yellowdice 2025-03-18 | Multiple fights in this TAS have been obsoleted now. Here is the new, updated TAS: youtube.com/watch?v=G_W84l8n-MwNote that the run is timed in IGT (In-Game Time) and not RTA. This TAS is not optimized for RTA timing. A lua script, which provides a bunch of additional useful information, was enabled during the playback of this TAS. This does not directly affect the TAS in any way.This entire TAS was made in less than a day, so it is very likely that further optimizations can be made. Feel free to let me know in the comment section if you find an improvement on a fight.I decided to look into this romhack last night when I saw a TAS of it by a channel called "OOBS - Gaming and More!" with a time of 16:06.15 (with the decimal on Tyson disregarded). This current TAS is over 2 minutes faster, with improvements on every fight besides Glass Joe.Let me know if you have any other ideas for things that I can try TASing, as I'm still kind of burnt out from the highscore TAS at the moment and am finding new things to look into. Also, feel free to ask any questions you have about this TAS in the comment section, there's probably a lot to unpack.Splits for each fight:Glass Joe - 0:42.00Von Kaiser - 0:37.61Piston Honda 1 - 0:44.48Don Flamenco 1 - 0:11.25King Hippo - 0:37.61Great Tiger - 0:36.00Bald Bull 1 - 0:55.99Piston Honda 2 - 0:48.82Soda Popinski - 0:55.48Bald Bull 2 - 1:02.82Don Flamenco 2 - 0:52.25Mr. Sandman - 2:14.61Super Macho Man - 0:43.25Mike Tyson - 2:46.61Final Time - 13:48.78 (13:48.17 if the decimal on Tyson is disregarded)also yeah I'm horrible at thumbnails what do you want from me
(World Record, Matches TAS) Don Flamenco 1 Starless in 32.00 yellowdice 2025-03-16 | I'm too lazy to explain anything in the description, if you're curious about anything regarding the fight, just ask in the comments.
[TAS] Don Flamenco 1 Low Score (2300 points) yellowdice 2025-03-15 | A bit over a year and a half ago, mPap made a video where he tried to attain the lowest possible score on each of the 14 fights in MTPO with the help of a TAS; the score he ultimately got was 35990. Pap obtained a score of 2310 on Don Flamenco 1 in his video, meaning that the fight shown here beats his fight by 10 points. The new lowest possible score for beating the whole game is now 35950, when you include an additional 10 point save that I found on Tyson, which can be found here: youtube.com/watch?v=9CnAvuVC7w4 as well as another 20 point improvement I found on Soda.Don Flamenco 1 is a fight where, after you counterpunch him enough times, Mac will get what's known as rapid punches. As the name implies, this allows Mac to counterpunch him faster, but it also slows the clock down each time. In order to beat Don 1 in 2 knockdowns, you'll need to knock him down twice in a minute, meaning that rapid punches should be used whenever possible. They're by far the most efficient way to deal damage to Don 1, besides star punches of course, which score too many points.Rapid punch storage is utilized excessively in this fight. After first gaining rapid punches, instead of continuing to counterpunch Don 1, a left gut punch is thrown to cause Don to return from stunned back to his idle animation, where a second left gut will cause him to retaliate with another uppercut. As Mac never actually finished counterpunching Don 1, the game doesn't recognize that he should no longer have rapid punches, which allows the following counterpunch to be a rapid punch.On Don Flamenco 1, the first counterhit after dodging an uppercut always deals 5 damage, while consequent hits deal 3 damage. Optimally, you want every punch to deal 5 damage, but this isn't possible under the strict 1 minute time limit, unless clockstop glitch is used. This glitch is strictly banned in normal play, but a fight demonstrating the theoretical minimum can be found here: youtube.com/watch?v=TI9e2K-77qAThis fight crams in as many 5 damage punches as possible, but towards the end, it becomes necessary to implement some 3 damage punches. My first knockdown score ends up being the exact same as Pap's, but the key difference is that it comes at 51 seconds, compared to Pap's 54 seconds. This 3 second advantage allows me to remove a punch in Phase 2, by delaying my final facejab to the final frame to deal 5 damage. As you can see in the video, this strategy barely works out, ending with a time of 59.97.I doubt this video is going to garner much attention, but just in case anyone was wondering, I've already informed and talked to Pap about this fight, so don't bother him about it if that was your plan. The reason Pap missed these 10 points while making his video is because he assumed that rapid punches last for a set amount of punches rather than being a flag that gets toggled on or off. To be honest, we're still not exactly certain why rapid punch storage works and I do not have the technical expertise to figure that out.If you're interested in Pap's now obsoleted fight or a commentated and clearer explanation of rapid punch storage, please refer to his original low score video: youtube.com/watch?v=psMBM3y52Hs&t=747sIf anyone is interested in further details or confused by the above explanation (which is probably quite likely, since I'm not the best at explaining things), feel free to ask in the comments or something, I don't know. I figured I'd just upload this video alongside a written explanation to ensure the most updated version of each low score fight is somewhere on the internet.
(Former World Record) Great Tiger in 28610 points yellowdice 2025-02-16 | New World Record: youtu.be/sNVkImBAQ3oRun begins at 6:06This beats the previous record of 28500 by mPap by 110 points, which, at the time, tied the TAS. 28500 was the best known possible score for the past 7 or so years, but when I decided to look into the fight 3 days ago, I managed to push the TAS score up to 29110, and so I thought I'd also quickly come up with a human viable strategy that can beat the record. This also means that the amount of TASes humans have tied dropped from 2 to 1, and as far as I'm aware, this is the first time this has happened in MTPO history.This strategy has odds of around 1/10 and doesn't even require pressing select to regain health. There are no frame perfects required in the run and none of the execution is really that difficult. This was my 3rd attempt of the day although I had put in around an hour and a half of attempts yesterday in addition to probably an hour of practice. To be frank, I did not expect to get it so soon and genuinely thought I was gonna somehow choke it.This record can pretty easily be beaten by playing Round 1 tighter, countering an extra right uppercut in phase 3 for an extra 110 points, and still landing the final star for 19 damage. This shouldn't be too difficult to do but it wasn't required for the record and so I didn't have to go for it. Rounds 2 and 3 can be played in the exact same way as in this current record as there is actually wasted HP in this current record when I get hit in R2 while Tiger has already recovered to 34HP (he can only recover up to 32-35HP).This is my 2nd ever highscore record and the 6th mPap highscore record to fall, with Hippo, Sandy, Don 2, Don 1, and Tyson being the other 5. As for tying the TAS on this fight, it's definitely still the next most likely candidate. I'm doubtful it will happen anytime soon though, as the TAS requires 1/40 odds (which really isn't that bad) and 6 frame perfects (that's the bad part). If highscore was a bit more active, I'd probably be more optimistic about the TAS tie happening, but as it stands, there really aren't that many of us that engage in this category.Update: It’s probably worth noting that the Tiger Highscore TAS now obtains a score of 29450 points.
(Former World Record, Matches MTPO TAS) Don Flamenco 1 Starless in 33.61 yellowdice 2025-01-31 | I found out around 2 hours ago that a frame perfect left quick dodge on a forced uppercut on Don Flamenco 1 can be buffered, meaning this fight boils down to 2 points of time loss: the quick dodges at the start of phase 1 and the start of phase 2. I had initially thought tying the TAS on this fight would be very difficult as I assumed each left quick dodge had to be manually timed, which is really stupid of me to think since I'm literally the one who made the TAS for this fight.Anyways, I figured it wouldn't take long to quickly grind out this TAS tie, and I got this attempt a little over an hour in. I can't count how many times I got 33.82, it's probably in the high 10s or low 20s, but I'm relieved to finally get the job done. When entering phase 2 on this attempt I was actually pretty sure that I had lost too many frames to be able to match the TAS, but as with many prior attempts in the session, I decided to play it out just to be sure, and received quite a pleasant surprise seeing the .61 on the blue screen.Not many people care about starless and I didn't think I'd really bother either but the idea of being the first to get an unbeatable record was pretty alluring. I believe the prior world record was a 33.82 by Nate09, who seems to be the only other person running this pretty obscure category. I don't think I'm gonna be grinding for any other ILs anytime soon but that may very easily change.As for the technical details about this fight, I use rapid punches to slow the clock down while simultaneously damaging Don 1. The most effective way to do so (aside from using Clockstop Glitch, obviously) utilizes rapid punch storage, where after obtaining rapid punches after countering Don's first upper, I intentionally throw a left gut punch that gets blocked to ensure that my first counterpunch following Don's next uppercut will be a rapid punch. This is repeated for phase 1, and phase 2 is identical to the speedrun.
[TAS] Mike Tyson Low Score (3200 points) yellowdice 2025-01-31 | Almost a year and a half ago, mPap made a video where he tried to attain the lowest possible score on each of the 14 fights in MTPO with the help of a TAS; the score he ultimately got was 35990. I've watched his video countless times but I figured that I'd look for an improvement for once. I found this 10 point save 3 days ago and am now finally getting to uploading it.mPap obtained a score of 3210 on Tyson in his video, using a similar strategy as displayed here but instead doing the initial 5 damage counterpunches on 4 uppercuts then finishing him off with 2 damage counterpunches following his 5th upper. For a clearer explanation, refer to his video: youtube.com/watch?v=psMBM3y52Hs&t=3113sHis second knockdown had a time of 2:25, although you can knock Tyson down as late as 2:28 and he would still perform his Blinking Pattern, which is set to occur at 1:30 and 2:30 in Round 2 and is necessary to perform a one hit knockdown. Here, I save an extra 10 points by doing an extra 5 damage punch on a 5th uppercut, thus removing the need for 2 2 damage punches.You may notice that I deal the 2 damage punches on his 1st uppercut, although this isn't strictly necessary. Tyson's guard is able to be manipulated after 2:00 in Round 2, which allows for early hook cancels that each save a couple frames. This is what allowed me to get a 6th uppercut in and save the 10 points. Doing the 2 damage counterpunches earlier allowed me to perform 3 early hook cancels rather than 1, although a singular early hook cancel is technically enough for an early enough knockdown.If anyone is interested in further details or confused by the above explanation (which is probably quite likely, since I'm not the best at explaining things), feel free to ask in the comments or something, I don't know. I figured I'd just upload this video alongside a written explanation to ensure the most updated version of each low score fight is somewhere on the internet.Also, I doubt this is going to garner much attention, but just in case anyone was wondering, I've already informed Pap about this fight, so probably don't bother him about it if that was your plan.As for the current updated lowest score for beating the whole game, it's actually 35950. Last night (as of the time of this upload), I decided to go through each fight and look for further optimizations while making a full game low score TAS. After some thorough analysis, I was able to squeeze out 10 more points on Don Flamenco 1 that Pap missed. A video of this fight can be found here: youtube.com/watch?v=3WvsE8oEJog Additionally, 20 more points have been taken off of Soda Popinski, meaning that there are 40 points worth of improvement over Pap's video altogether.
(World Record) Don Flamenco 2 in 16670 points yellowdice 2024-12-24 | The video quality is pretty choppy, because my computer has a tendency to lag a lot. For OBS to not have an encoding error, it forces me to render the video in a lower resolution, or something like that.I put countless hours into this fight back in early November, choking the record countless times before giving up. I thought I'd return to the grind yesterday and choke thrice more before Don 2 gives up the luck one final time and I finally execute. I had a whole document of exactly what to do written down but funnily enough, the one time I actually finished a fight, I just went off of instinct rather than following through with the plan.This record was previously 16560 held by D-manXD (which, my initial grind for this record was actually the reason he decided to attempt for WR as well). This is an 110 point improvement, primarily thanks to the 2nd unnecessary random star I got in round 2. Feel free to credit it as much as you will but I'm glad that unexpected RNG isn't the only reason this is the WR as I would've beaten it by 10 points anyways.If I've done my calculations right, this luck could've yielded a score as high as 16730 points (with my current strategy). I got the best refill in every possible spot, got 2 1/16 random stars, as well as a good select refill in round 3. The run didn't have any mistakes up until the end of round 2, when I was supposed to get by both his jab and his hook to allow him to recover 4 extra hp, which can result in me gaining an extra 20 points. Unfortunately, I dodged the jab instinctively, forgetting the strategy in the moment.The other mistakes are all in round 3. I definitely could've lowered his health further in phases 1 and 2 before landing the final star punch and I should've been able to get even more points had I been getting star punches with gut punch intercepts rather than buffered frame perfect face punches. I knew I had a lot of room to work with, considering that I got the extra 100 points from the random star, and not wanting to have to risk dealing with my probably 20th choke, I figured that I should just play it safe.Thanks to Pap for coming up with most of this strategy (I altered phase 1 to allow for 20 extra points which he missed due to an oversight, as well as adjusting round 2 to accommodate for double large fill) and D-man for repeatedly listening to me rant about this fight during my grinds.
(Former World Record) Bald Bull 0 in 48.97 yellowdice 2024-11-06 | The strat for this fight was by your local pizza guy, who also developed the strategy for the TAS of this fight. Phase 1 of this human viable strategy is identical to the TAS, and phase 2 is a slight variation. Phase 3 is nearly 4 seconds slower than the TAS, but I couldn't find further optimization because Bull 0 starts dodging stars. I haven't TASed this strategy yet but I doubt it can go lower as playing phases 1 and 2 too fast will allow Bull to dodge the first star of phase 3 for a reason I have yet to figure out, meaning the fight boils down to 2 frame perfect inputs - the entirety of phases 1 and 2, and the first guard manipulated gutter in phase 3 (technically that’s 2 frame perfects - the up release and b press, but who’s counting).The previous record was a 48.99, meaning this improves it by 1 decimal. I'm unsure exactly how many frames this record is ahead by as I used the increment version of the Tyson romhack rather than the decimal. It'd be really funny if I was exactly 3 frames ahead, although I highly doubt I was able to shave that much time off of the previous record.
[TAS] Bald Bull 0 in 44.82 (Obsoleted) yellowdice 2024-11-05 | This used to be the fastest known way to beat Bald Bull 0, otherwise known as Demo Bull. This is the same fight as the one that plays when you do nothing on the title screen for too long, and can only be accessed with the help of Game Genie codes. If you couldn't tell, the Select press at the start of the fight to lower my health to 48hp isn't necessary.The previous best TAS of this fight was by @yourlocalpizzaguy and is 3 frames slower, clocking in at 44.97. This 44.82 is the exact same as the 44.97 except for a single punch in Phase 2. The previous TAS had a face punch that required delaying 3 frames to seep through Bull's guard, although I accidentally found out that a buffered misdirected gut punch can achieve the same purpose. When getting a star, a left gut punch and face punch take up the exact same amount of frames on Bald Bull 0, hence why this TAS saves exactly 3 frames. Almost all the contributions to this TAS was made by your local pizza guy, all I did was accidentally implement a misdirected gutter, so he definitely deserves more credit than me.The blocked face punch at the start of Phase 1 definitely seems suboptimal, but I couldn't find a way to improve it. It helps set up a 30 damage star punch as soon as possible, which allows for a 17 second first knockdown. This TAS is almost definitely humanly impossible as there is a pretty difficult star punch with multiple frame perfect inputs strung together in a short duration.
Can You Beat Mike Tysons Punch Out Without Pressing Up? (Obsoleted) yellowdice 2024-11-01 | I can't quite upload an actually optimized TAS, apparently. I just found an up press save on Mr. Sandman today (4th March 2025) and the updated amount of up presses required to beat Mike Tyson's Punch-Out!! is now 5 (1 from Hippo, 4 from Sandy).Hopefully this one does well, I put quite a bit of effort into it.Links:Opponent Behavior Guide - gamefaqs.gamespot.com/nes/525246-mike-tysons-punch-out/faqs/75406MTPO Wiki - mtpo-speedrunning.fandom.com/wiki/MTPO_Speedrunning_WikiPunch Counter Video - youtube.com/watch?v=tTXDgdyGGWMMusic Credits:00:00 Undertale - Spear of Justice00:39 Super Mario 3D World - Super Bell Hill01:37 Five Nights at Freddy’s Pizzeria Simulator - Smashing Windshields03:09 Super Mario Odyssey - Steam Gardens04:13 The Legend of Zelda: The Wind Waker - Dragon Roost Island05:42 Castlevania - Vampire Killer06:43 de Blob - Euphoric07:15 Super Mario World 2: Yoshi’s Island - Athletic08:11 C418 - Wait09:11 Tekken 5 - Moonlit Wilderness09:48 Sonic Unleashed - Windmill Isle Day11:04 Bitzel - GO GO GO!12:22 The Legend of Zelda: Ocarina of Time - Gerudo Valley14:26 Super Mario World - Castle17:06 Super Meat Boy - It Ends (Retro)17:56 Super Mario Galaxy - Gusty Garden Galaxy19:21 Super Smash Bros. Melee - Final Destination20:59 Sonic Mania - Mirage Saloon Act 2 / Rogue’s Gallery22:08 Super Mario World - Ending
[TAS] Don Flamenco 1 in 2240 Points (Theoretical Minimum) yellowdice 2024-08-15 | I literally started making this like 45 minutes ago but I thought it was cool enough to upload. I was going to do something similar for a Macho Infinite Spin Glitch with no RNG manip (no idea how that works) but I never got to it. Also that would technically beat Zallard's record for "longest Punch-Out fight ever", though that can also be done with Clockstop (the glitch abused in this video).mPap's low score video - youtube.com/watch?v=psMBM3y52HsI don't think this obsoletes anything since Clockstop is obviously a glitch and Pap probably knew about it and chose to avoid it intentionally when making that TAS. He did use Rapid Punch Storage, though, which is also a glitch, so I'm not quite sure.If you're curious what Clockstop Glitch is, it's essentially when you get Rapid Punches when countering Don's uppercuts, then proceed to force him to block, which is essentially Rapid Punch Storage. The glitch is finished when you punch Don again and get him to retaliate, where the clock will pause until you punch him again. A better explanation can be found in Pap's Don 1 WR Progression video.
[TAS] Can You Beat Mike Tysons Punch-Out!! With Only Gut Punches (Obsoleted) yellowdice 2024-07-24 | Important update! I'm not sure if anyone will see this but I'll leave this here just in case anyone does come along. Anyways, the TAS has been obsoleted and additional face/star punches can be taken out from the TAS. I've made an updated TAS file but frankly don't know what to do with it, since I don't think many people will be interested seeing as they've seen most the fights and understood the basic concept. Regardless, thanks to FartButt87 for correcting a couple basic and stupid oversights that I missed due to lack of research. Oh well.Obviously a twist (not sure if that's the right word) on an mPap challenge video, just much lower budget and less accommodating to anyone who can't see. If you don't know who he is (highly unlikely, considering you saw this and clicked it), please check him out: youtube.com/@UCIhITN99GcmbJ8fUcdxx1IwHopefully you didn't think this was an mPap video. If so, sorry to disappoint lolI mentioned in the video that I'd link the Opponent Behavior Guide that he made in the description, which is: gamefaqs.gamespot.com/nes/525246-mike-tysons-punch-out/faqs/75406Also, I thought I'd mention that this idea originated from a comment on Pap's minimal punches video from a user called "TheGrindingIce". I thought it was worth TASing and here we are.For anyone wondering what a TAS, or Tool Assisted Speedrun is, the brief paragraph explanation I flashed at the beginning of the video is just info you need to know for this video. Essentially, I can control Mac's input for every frame as well as manipulate randomness as a result. TASing in other games mostly involve re-recording though, and not just adding inputs in a TAS editor (or at least I think so, since I haven't gotten to trying). If you're interested, here's a Bismuth video on it: youtube.com/watch?v=Ietk1-Wb7oYIf you're interested in Mike Tyson's Punch-Out speedrunning, there's also a leaderboard at speedrun.com/mtpo as well as a discord that has links to a ton of resources and tutorials, which is discord.gg/QQkCecPd as well as a wiki, which is mtpo-speedrunning.fandom.com/wiki/MTPO_Speedrunning_WikiPresuming people see this and put in the effort to read to the bottom of the description, I also made a TAS using minimal gut and star punches, one collaging every individual level world record currently into a full game run, and my own spins on minimal punches and dodges (which really aren't that different). None of them are as interesting as this one, and it's not like this one's SUPER interesting either, but if someone wants to see them I could release them onto the internet.
![[TAS] TTM 100 Coins on 0x y-scale](https://i.ytimg.com/vi/JtFrs8zhtvA/hqdefault.jpg)