zallard1 | Splatterhouse (Arcade) - *Former* High Score World Record [742,600 points] (TG/Aurcade ruleset) @zallard1 | Uploaded March 2019 | Updated October 2024, 1 hour ago.
EDIT: I beat this by 68,800 points! Check out the new run here: youtube.com/watch?v=t8PPDkUaWsQ
This was played on an original Splatterhouse arcade PCB on the H.A.S. supergun. This run was also done on default difficulty settings shown at the end of the video, and was recorded through an S-Video connection because my setup cannot support RGB signals at the moment. Here is a picture of my boardset w/ its original romset & supergun setup: pbs.twimg.com/media/Dy6ucfvUwAA4AZC.jpg
It's worth noting that this run follows the Twin Galaxies & Aurcade ruleset for score in this game, which bans the use of the double KO tactic on the final boss, which is otherwise the most lucrative option to cash out your extra lives. If this were allowed, then you would have to set up guaranteed double KO's with the blue orb timer, which is something on the order of 14+ minutes per double KO. You could also die to rocks at the end, but if you're on a good pace, the guaranteed route is the best option for consistency. Without the restrictions in this category, scores in the 1.2 - 1.3 million point range are possible.
I also need to make a note that showing your physical DIP switch settings for this game is worthless. Switch 1 enables the settings menu, while switches 2-8 do absolutely nothing, so the only thing that showing the physical dip switches does is tell you whether or not the game is in test mode... which should be already obvious if that's the case or not by looking at the screen. Once the run was over, I flipped DIP switch #1 to show off the actual settings the board was set to during that run.
Because I am doing these attempts on a SuperGun, whenever I reset the board, I am pressing the "Test" button on the SuperGun. This actually resets the board without me having to power off the board and cause potential damage to it over time.
I'm also posting an archive of the entire attempt session, which includes the boot sequence of the board from powering it on initially, up to the successful attempt:
Youtube - youtube.com/watch?v=U9JfMckWfzc
Twitch - twitch.tv/videos/393247354
And for convenience, here's the Twitch highlight of the run itself, complete with live commentary similar to the last video I uploaded of this game: twitch.tv/videos/393259083
---
Ok, now that we got all the fine details out of the way, let's talk about the run itself. My pace was pretty good heading into Stage 6 this time, but I still had some stage 6 loops that were cut short. Instead of them being embarrassingly short, they all at least broke 25k, as opposed to some from the last run that couldn't even break 20k. I did also have less than 60k for more of my loops, but my average stage 6 was still better overall, despite seeing less blue orbs than the previous record.
Something I need to elaborate on; the blue orb is a counter-measure so that players do not stay in one spot for an infinite amount of time. This is so there's something to kick idle players off of the game so any new players who walk up to the cabinet don't get any free games, as well as just make sure that players can't stall in one spot to prevent others from playing. This orb is great for score, since it puts a finite limit on how long you can point press any given spot.
I gotta give huge shoutouts to both Anthony Paparo and Caitlin Oliver for driving up this score over the years! Their efforts to optimize this score ultimately inspired me to do attempts of this once I finally got access to arcade hardware.
EDIT: I beat this by 68,800 points! Check out the new run here: youtube.com/watch?v=t8PPDkUaWsQ
This was played on an original Splatterhouse arcade PCB on the H.A.S. supergun. This run was also done on default difficulty settings shown at the end of the video, and was recorded through an S-Video connection because my setup cannot support RGB signals at the moment. Here is a picture of my boardset w/ its original romset & supergun setup: pbs.twimg.com/media/Dy6ucfvUwAA4AZC.jpg
It's worth noting that this run follows the Twin Galaxies & Aurcade ruleset for score in this game, which bans the use of the double KO tactic on the final boss, which is otherwise the most lucrative option to cash out your extra lives. If this were allowed, then you would have to set up guaranteed double KO's with the blue orb timer, which is something on the order of 14+ minutes per double KO. You could also die to rocks at the end, but if you're on a good pace, the guaranteed route is the best option for consistency. Without the restrictions in this category, scores in the 1.2 - 1.3 million point range are possible.
I also need to make a note that showing your physical DIP switch settings for this game is worthless. Switch 1 enables the settings menu, while switches 2-8 do absolutely nothing, so the only thing that showing the physical dip switches does is tell you whether or not the game is in test mode... which should be already obvious if that's the case or not by looking at the screen. Once the run was over, I flipped DIP switch #1 to show off the actual settings the board was set to during that run.
Because I am doing these attempts on a SuperGun, whenever I reset the board, I am pressing the "Test" button on the SuperGun. This actually resets the board without me having to power off the board and cause potential damage to it over time.
I'm also posting an archive of the entire attempt session, which includes the boot sequence of the board from powering it on initially, up to the successful attempt:
Youtube - youtube.com/watch?v=U9JfMckWfzc
Twitch - twitch.tv/videos/393247354
And for convenience, here's the Twitch highlight of the run itself, complete with live commentary similar to the last video I uploaded of this game: twitch.tv/videos/393259083
---
Ok, now that we got all the fine details out of the way, let's talk about the run itself. My pace was pretty good heading into Stage 6 this time, but I still had some stage 6 loops that were cut short. Instead of them being embarrassingly short, they all at least broke 25k, as opposed to some from the last run that couldn't even break 20k. I did also have less than 60k for more of my loops, but my average stage 6 was still better overall, despite seeing less blue orbs than the previous record.
Something I need to elaborate on; the blue orb is a counter-measure so that players do not stay in one spot for an infinite amount of time. This is so there's something to kick idle players off of the game so any new players who walk up to the cabinet don't get any free games, as well as just make sure that players can't stall in one spot to prevent others from playing. This orb is great for score, since it puts a finite limit on how long you can point press any given spot.
I gotta give huge shoutouts to both Anthony Paparo and Caitlin Oliver for driving up this score over the years! Their efforts to optimize this score ultimately inspired me to do attempts of this once I finally got access to arcade hardware.
![Splatterhouse (Arcade) - *Former* High Score World Record [742,600 points] (TG/Aurcade ruleset)](https://i.ytimg.com/vi/HuO4CC5Gjbk/hqdefault.jpg "Splatterhouse (Arcade) - *Former* High Score World Record [742,600 points] (TG/Aurcade ruleset)
EDIT: I beat this by 68,800 points! Check out the new run here: https://www.youtube.com/watch?v=t8PPDkUaWsQ
This was played on an original Splatterhouse arcade PCB on the H.A.S. supergun. This run was also done on default difficulty settings shown at the end of the video, and was recorded through an S-Video connection because my setup cannot support RGB signals at the moment. Here is a picture of my boardset w/ its original romset & supergun setup: https://pbs.twimg.com/media/Dy6ucfvUwAA4AZC.jpg
Its worth noting that this run follows the Twin Galaxies & Aurcade ruleset for score in this game, which bans the use of the double KO tactic on the final boss, which is otherwise the most lucrative option to cash out your extra lives. If this were allowed, then you would have to set up guaranteed double KOs with the blue orb timer, which is something on the order of 14+ minutes per double KO. You could also die to rocks at the end, but if youre on a good pace, the guaranteed route is the best option for consistency. Without the restrictions in this category, scores in the 1.2 - 1.3 million point range are possible.
I also need to make a note that showing your physical DIP switch settings for this game is worthless. Switch 1 enables the settings menu, while switches 2-8 do absolutely nothing, so the only thing that showing the physical dip switches does is tell you whether or not the game is in test mode... which should be already obvious if thats the case or not by looking at the screen. Once the run was over, I flipped DIP switch #1 to show off the actual settings the board was set to during that run.
Because I am doing these attempts on a SuperGun, whenever I reset the board, I am pressing the Test button on the SuperGun. This actually resets the board without me having to power off the board and cause potential damage to it over time.
Im also posting an archive of the entire attempt session, which includes the boot sequence of the board from powering it on initially, up to the successful attempt:
Youtube - https://www.youtube.com/watch?v=U9JfMckWfzc
Twitch - https://www.twitch.tv/videos/393247354
And for convenience, heres the Twitch highlight of the run itself, complete with live commentary similar to the last video I uploaded of this game: https://www.twitch.tv/videos/393259083
Ok, now that we got all the fine details out of the way, lets talk about the run itself. My pace was pretty good heading into Stage 6 this time, but I still had some stage 6 loops that were cut short. Instead of them being embarrassingly short, they all at least broke 25k, as opposed to some from the last run that couldnt even break 20k. I did also have less than 60k for more of my loops, but my average stage 6 was still better overall, despite seeing less blue orbs than the previous record.
Something I need to elaborate on; the blue orb is a counter-measure so that players do not stay in one spot for an infinite amount of time. This is so theres something to kick idle players off of the game so any new players who walk up to the cabinet dont get any free games, as well as just make sure that players cant stall in one spot to prevent others from playing. This orb is great for score, since it puts a finite limit on how long you can point press any given spot.
I gotta give huge shoutouts to both Anthony Paparo and Caitlin Oliver for driving up this score over the years! Their efforts to optimize this score ultimately inspired me to do attempts of this once I finally got access to arcade hardware.")
![Punch Out!! (Wii) - Bear Hugger [0:26.70] (WR)](https://i.ytimg.com/vi/J7yBhPnA2m8/hqdefault.jpg "Punch Out!! (Wii) - Bear Hugger [0:26.70] (WR)
Now these ILs are starting to heat up in difficulty. This fight requires 9 frame perfect inputs to match the WR/TAS time of 26.70, and to set the first attack of Bears to come out as soon as possible, I did not skip the cutscene in any of my attempts. I threw tons of attempts at this fight with numerous really close calls. I had gotten six 26.76s, and five 26.82s before finally getting this perfect time. I also had a whopping 62 attempts that got out of phase 1 with a 2:54 (Ill explain the significance of this knockdown time in a little bit).
Lets break down this fight a bit. The name of the game for this fight is to land star punches during the first frame of startup from Bears next attack. Every single attack and taunt Bear has is critically vulnerable to a star punch, even if it doesnt seem like it should hit!
After collecting the first star from Bears taunt, he starts reeling to the side. If you throw a delayed hook, you can cancel his reel animation so he can start his next attack sooner than normal, which is absolutely mandatory to master if you have any hope of getting sub 27 seconds on Bear. After cancelling his reel, throwing a delayed star punch with the proper timing will meet his hook perfectly to send him to the mat at 2:54, which is a possible first knockdown time ONLY if youve hit the first 3 inputs frame perfectly. One frame slow on any of the inputs in phase 1 will get you a 2:53 first knockdown.
When he gets up, countering his hook will net you a star, then hell bust out his signature hug attack, which you want to hit with a star as early as possible. If you hit it perfectly, you should see Bears head tilt up slightly right as the punch is about to connect, too early and the star will deal normal damage, and too late will obviously have you take a ton of damage.
The final phase of the fight is effectively the same as phase 2, except his hug comes out much quicker than before, so as long as youre ready for it, it should be no problem.
I cannot begin to describe how satisfying it was to crush my decade long PB on this fight almost instantly when I started attempts on him. This fight was *so* hard for me to optimize when I was just starting out with speedrunning. Its just such a tangible reminder in how far Ive come in running Punch Out games, and speedrunning as a whole over the years.
For the next fight, Ill be working on Piston Hondo, since I want to put off Disco Kid as much as logically can, lol. Hondo is another 9 frame perfect input record, held only by Pottoww right now.
The Hondo strategy is pretty new, but the fight should prove to be quite a bit more difficult than this one for me, on the basis of this fight having similar inputs repeated, whereas that fight has mostly unique inputs that have to be mastered and strung together one after the other. Not to mention that fight also has mandatory frame perfect move cancels, and a similar situation where you can get a phase 1 knockdown thats immediately identifiable as frame perfect, since 2:47 is possible to get on the very last frame. Expect me to upload my progress on that fight every now and then (i.e. it almost definitely wont be immediately a record tie like these other ones have been).
Anyways, thanks for watching!")
![Breakfast Brawl - Speed Run [1:14 in-game time]](https://i.ytimg.com/vi/JGLkew-tx2c/hqdefault.jpg "Breakfast Brawl - Speed Run [1:14 in-game time]
This was a flash game put out by Dannon as an advertisement for their Frusion smoothies quite a few years back. Needless to say, this is one of the greatest advertisements ever and I wish more ads were in the form of Punch Out games lol
I routed this run back in 2011, and had a 1:16 run submitted to Speed Demos Archive in early 2012. At the time, I was content with this run because I had hit every single PB time for every boxer in a row, and figured thatd be the end of it.
But since Flash wont be supported in 2021, I figured I should try and see if any improvement could be found to beat that run before Flash becomes completely unusable. I decided to go back and look at the SDA verification thread to refresh my memory on a couple of things, and remembered that verifier #3 (who was named ExplodingCabbage on the forums) had some alternate strategies that supposedly got the same times that I got in my run.
At the time, I just assumed that the strategies were going to end up the same times as mine because thats what ExplodingCabbage reported, so I figured double checking that myself wasnt necessary.
However, its 9 years later, and Im a lot better at Punch Out games than I was in 2011, so it might be worth pushing all of these strategies as hard as possible to see if any of them end up faster if the execution is tightened *everywhere* in every fight. Turns out, his strategies for both Dave Jenkins AND Papi C. Debagel BOTH have potential to gain a second when doing the strategies really fast. Theres a small catch for the Jenkins fight that Ill get into later, but Papis is just straight up faster overall.
I decided to go through the menus pretty slowly, since not a lot of people know about this game and might get a kick out of being able to read the sections in between fights. This run is measured by the sum of each whole second that has elapsed on the fight timer per fight. Ill likely do a follow up video on this that prioritizes real time over game time since this game is so short, but I figure uploading this type of run first would be more watchable overall.
As far as mechanics go, all opponents have guards similar to Mike Tysons Punch Out and Super Punch Out, however, their guard alternates high and low on their own without being possible to manipulate. This guard shifting will also carry over to subsequent phases, which is important to always keep track of so you can hit opponents as fast as possible. You also get Frusionators (aka star punches) from landing 3 consecutive punches in a short period of time.
Its also worth noting that unlike any other Punch Out games, buffering jabs or hooks after youve hit an opponent once is VERY BAD because as an opponent is reeling from a previous punch, they have some frames where theyre intangible, so buffering a subsequent punch will miss outright, therefore every punch aside the first one in each phase must be timed carefully.
Lets get into the fights...
Ray Zinbran (10 seconds) - This fight shows off the standard 10 second knockdown that will be a key part of every fight from this point forward. If possible, its faster to use a Frusionator as one of your 3 required punches in order to get a second Frusionator.
Moe Finntop (20 seconds) - The first phase is the exact same strategy as Ray Zinbran,, but if you are very close to frame perfect, Moes guard will be in the wrong spot to start the next phase, so keep that in mind. Start phase 2 with 2 low punches with a high punch to get a Frusionator, then a jab with 2 low punches to get a 2nd Frusionator, the let a Frusionator rip. Now for the phase 3 Frusionator, you cant buffer this, you actually have to time it exactly as Moe is fully centered to finish him off.
Dave A sausage, egg & cheese sandwich Jenkins (24 seconds) - This is the hardest fight in the game, and by far has the most strict execution to get his goal time. Starting with the standard 10 second knockdown will reveal that Dave has way more HP than anyone else, so instead of throwing the 2nd Frusionator, throw 2 hooks into a jab to collect another Frusionator, then send him to the mat with one. If you are too fast on the first 6 punches, your 7th will miss with this strategy, so be very mindful of that. For phase 2, throw 2 low punches into 2 supers and you should take him down a 2nd time. Phase 3 is just 2 jabs into 1 hook, and a Frusionator to hopefully finish him at 1:36 left on the clock.
Papi C. Debagel - Papis guard is way faster, so the 10 second knockdown needs to be adjusted for it, but itll still work since his HP is low like the first 2 fights. Start with 2 jabs into low to get the first Frusionator, then do the EXACT same sequence to get the 2nd Frusionator, then throw one for the 10 second knockdown. Similar to Moe, you want 2 Frusionators immediately, then throw one to send him to the mat. After timing the phase 3 Frusionator, hell be down for the TKO at 1:40.
Anyways hope you enjoyed this, thanks for watching!")
![Punch-Out!! (NES) - Mr. Dream [2:06.00]](https://i.ytimg.com/vi/JSYXtCoYwqA/hqdefault.jpg "Punch-Out!! (NES) - Mr. Dream [2:06.00]
This run was done on an original cartridge of Punch-Out!! on an NES that was modded with the Tim Worthington NESRGB kit. For some more in depth information about my RGB setup, you can check out this pastebin here: https://pastebin.com/m7tzR5Cq
Ive also decided to include a non-stretched version of this fight in an unlisted video. If youd like to learn more about why Im including this too, you can check out the description in the second video here: https://www.youtube.com/watch?v=2jxZsQx2_88
I uploaded a 2:06.61 on Mr. Dream a while back, since I figured it would be sort of fun, albeit kind of silly, to kind of track what my best Dream fight is alongside my Tyson fight. My Tyson PB is a 2:05.82, so this is now my current Mr. Dream PB even though the fight itself is considered completely equivalent to the Tyson fight, and is ranked as such on leaderboards. I also feel like it would be kind of fun to get my Dream/Tyson PBs to match each other if possible, and with this, theyre now pretty close. The exact time for this fight is a 2:06 on the 3rd frame of the .00 increment, according to the Tyson Finder spreadsheet created by Omnigamer.
Despite the difficulty of this fight, its actually pretty simple to explain. To deal maximum damage on Dream in the uppercut phase, you need to mirror what side he throws his punch on when you throw a jab to stun him. Each time you do this, it deals 5 damage to his 96HP health bar, as opposed to 1 damage if you stunned him with the opposite fist. The 2nd punch for each of these also needs to be delayed to the very last frame of the stun. If you land it, itll also deal 5 damage, otherwise youll be blocked or deal 1 damage. Phase 1 requires you to hit 9 out of 10 of these, while getting decent enough luck to get past phase 1 at 56 seconds or below (or below 1:00 if you just care about him throwing 6 uppercuts in phase 2).
Phase 2 is pretty much the same as phase 1, except you wont see the 8 second delay from Dream. You only have to hit 5 of 6 frame perfect punches here, due to his 56HP health refill, which he gets if you have full health when hes knocked down. For being in contention of the best non-phase-3-uppercut time, you need Dream down by 1:28 or earlier.
Phase 3 will start with Dream going into his hook phase at 1:30, and you need to hit 8 hooks frame perfectly in order to get a perfectly executed fight. His health refill is 40HP this time, and dodging/punching will get you 2 damage per punch, whereas if you wait until the very end of Dreams hook, and punch on the very last frame, youll land 1 punch that deals 5 damage. As long as you get to the start of this phase at an early enough time, your final time is completely dictated by what pattern you get from Dream, with the very best phase 3 pattern netting you a 2:05.25. This phase 3 pattern is literally the only difference between this fight, and my 2:05.82 on Tyson.
Another big reason I wanted to upload this is to use this as an opportunity to mention to people here that Ive been actively de-rusting and doing full game runs of Mike Tysons Punch-Out on my Twitch channel lately. Stop by if you want, and maybe you might catch a new PB live! http://www.twitch.tv/zallard1/
Anyways, thanks for watching!")
![Pokémon Snap - Mew [10,000 points] (WR)](https://i.ytimg.com/vi/JVWuJu7dyek/hqdefault.jpg "Pokémon Snap - Mew [10,000 points] (WR)
Oh boy how do I start this description off. Well first off before I talk about why I uploaded this, I should mention that at the start of May, I suffered a stroke. I got to the hospital quickly and my recovery has been good so far, but if you want to catch up on how Ive been doing (and what exactly happened), this thread has everything, including the incident itself described in the pastebin, my gradual recovery, and any struggles Ive had along the way:
https://twitter.com/zallard1/status/1653471339100119041
A side hobby Ive had over the past few years has been soldering, primarily for console modding and repair. As a way to gauge my precision with hand-eye coordination after the stroke, I started doing small soldering projects and have been gradually ramping them up as I see fit. Like recently I ended up fixing a bundle of N64 Expansion Paks that I got for cheap with very small resistor soldering, and then later I ended up installing an advanced RGB mod in my childhood N64 that can reverse some of the blurring on the graphics on a hardware level. You can check out both of these here in these two threads:
https://twitter.com/zallard1/status/1664461054188064768
https://twitter.com/zallard1/status/1665234475301650438
I decided I wanted to upload something super different from what I usually upload to kind of set the tone that I probably will be taking it easy on seriously speedrunning things like Punch-Out for a little while, and will likely upload things that are a bit more chill in nature and slowly work my way up to harder things as I continue recover.
So to start, I decided to upload something from probably one of the most relaxing games I played as a kid, Pokémon Snap. I never got the perfect Mew picture as a kid, which is (I believe) a frame perfect picture right before it disappears by going through you, while it does its pose before it vanishes, and is fully centered.
This is almost definitely the easiest score record to tie in the game, as you have an entire several minute autoscrolling stage to bully this one Pokémon and exhaust your entire roll of film on it. This took me maybe 2 hours collectively of trying off and on to get the picture, so it wasnt bad at all. For those same reasons, Im pretty sure its the most tied record in the game too, likely with the original perfect 10,000 score set by someone back in 1999 when the game was released. Still harder than the comically easy tied world record for MTPO Glass Joe though, haha. Its a pretty fun and relaxing record to get; Id highly recommend giving it a shot if you feel like youre in a mood to play N64 for some reason.
I remember back when this came out, I actually got to rent the game from Blockbuster, beat it, and print out stickers on their special kiosk that they had specifically for Pokémon Snap. I still have a VHS tape somewhere with my best photos from back then recorded on it, along with some of those stickers still on it! Not sure if anyone would want to see it, but if I ever find that tape again, would people be interested in seeing the footage on it, as well as what the stickers looked like? I dont think any of my scores were any good, but it might be fun to document it?
Im not sure if I want to try any other picture score tied records, but this was a nice way to test out recording this console with a Datapath capture card and upscaling it to 4K. I think it looks great for footage off of a real N64, and the deblur function combined with that probably makes this the best quality Mew score record off of an actual N64 uploaded to YouTube right now, lol.
I just started streaming again on Twitch yesterday too. Itll likely just be very very easy going stuff, like maybe a playthrough or two of a few RPGs, with some action games here and there depending on how I feel on any given day. For yesterday, I started a playthrough of Paper Mario on this exact same RGB modded N64, then I ended the stream with a round 1 Tyson just to see if I still got it or not, haha. If anyone would like to keep up with me and my streams, Ill be streaming again in a couple days at: http://www.twitch.tv/zallard1/
Anyways, thanks so much for checking this out and taking the time to read this. Hopefully Ill be fully back to speedrunning sooner rather than later.")
![Punch Out!! (Wii) - Soda Popinski [0:42.89]](https://i.ytimg.com/vi/K-P5PxNbLqs/hqdefault.jpg "Punch Out!! (Wii) - Soda Popinski [0:42.89]
So... I improved my Soda time by the smallest possible increment; an improvement of 0.01 seconds lol. Every punch in this fight was frame perfect aside from the final star punch, which I lost 2 frames on. Still closing the gap on Kefkamaydies record slowly but surely haha. This fight will probably be my main focus until the record has video proof, so expect more improvements inching closer to that time! Ill post the previous videos description below for posterity.
As of right now, this is the fastest Soda fight to have video proof, but the commonly accepted WR is currently held by Kefkamaydie at 42.84, set back in 2009.
This is one of two fights in this game that requires a technique called the TKO Recovery which, as the name suggests, is when Mac resists being TKOd and gets back on his feet instead of going down for the expected TKO. This happens completely randomly, estimated to be anywhere from 15-25% on average. This makes optimizing this fight far more difficult than it would otherwise be, considering 75%+ of your attempts wont make it past this point. Champions mode is used for this fight to skip his rage uppercuts whenever he gets up from a knockdown. It also has Soda throw a jab whenever you get up from a knockdown, which can be countered for a star!
Lets get into the fight. The first thing thats done is a delayed gut punch off of Sodas first jab, which if done correctly, yields a star. If you follow that up with a delayed jab, then throw a star as hes reeling, Soda will be instantly knocked down!
Phase 2 involves taking a knockdown to skip Sodas rage uppercuts, then getting up and countering Sodas jab for a star. Off of this counter, I buffer 2 ducks into alternating punches to hit Soda on the first possible frame as he finishes drinking his soda and walks back in, which weirdly enough, rewards a star. The instant knockdown done in phase 1 works in subsequent phases... but will require an additional star each time its done (this time requires 2). This particular attack also lets me do a delayed jab to get him into position for the instant knockdown too, since countering his bottle would just make Soda mad and rush in for a sequence of rage uppercuts.
Phase 3 has Mac take his 2nd knockdown, to which Mac gets up and counters Sodas scripted jab. Now, this delayed star punch needs to be timed in a very specific way such that Sodas next attacks gets skipped. If you throw the star punch too early, then Soda will do slow hooks after Macs next knockdown, which instantly ends the attempt. The visual cue that youve thrown it correctly is if Soda just barely gets completely centered as the star punch lands.
After the star punch, Soda reels to the back of the ring and runs back to Mac to throw more rage uppercuts, which Mac takes. This is where luck comes into play, because thats Macs 3rd knockdown, meaning we need the lucky TKO recovery. Thankfully in this run I get it!
Once Mac is back up on his feet, counter Sodas scripted jab, and counter Sodas bottle. This will make Soda mad and hell approach you to do rage uppercuts, however a delayed star punch will hit him if you time it just right, and a double star punch here is just enough to take him out for good.
This was actually a fight I streamed on my Twitch page, so you can hear my reactions there if youd like!
https://www.twitch.tv/videos/754385350
Ill be doing more ILs streams there too, so if you like these fights, check out my progress live! I wont be doing *all* my ILs on stream, but Ill certainly do some amount of them for sure! Ill definitely be trying to improve my Soda time on stream at the very least.
Anyways, thanks for watching!")
![Super Punch Out!! - Hoy Quarlow [00832] (PAL WR)](https://i.ytimg.com/vi/KdqReaGLUOo/hqdefault.jpg "Super Punch Out!! - Hoy Quarlow [00832] (PAL WR)
This run was done on an original Spanish Super Nintendo cartridge of Super Punch-Out, as well as an original PAL Super Nintendo 1chip console that was recapped and modded with the Borti RGB amp.
Hoy Quarlow on the PAL version is pretty different compared to the NTSC strategy. If you would like to compare the NTSC fight to this one to see everything thats changed between the two strategies, you can check it out here: https://www.youtube.com/watch?v=QB_f64jZiA0
Just like the NTSC version, Hoy is very random and will ruin attempt after attempt, usually by doing random dodges when you try to throw unstunned punches, or by randomly changing his guard after manipulating it, which makes him shove you if he blocks your punch.
The fight begins with the usual guard manipulation by pressing nothing for the first frame, and Up + Y on the second frame, which will allow a left jab through his guard, assuming you get favorable luck (3/4 likely). From there, buffer another left jab (buffered unstunned jabs are 9/16 likely), then Hoy will go into a crane stance before attacking you. His sprite will twitch for a brief moment, and immediately following that twitch, you want to throw out a delayed right gut punch to intercept it.
This is where the NTSC and PAL strategies diverge. The NTSC strategy from this point forward would yield an 8”34, but Lixunis found a completely different approach based on a phase 1 strategy that Khananaphone found that can cut off a frame in the PAL version.
After intercepting Hoy’s crane stance, hell immediately attack low, which you can counter with a delayed right gut punch. Hoy can only be hit with alternating punches when he’s stunned, so youll need to end his stun with a buffered left jab. From here, you can buffer 2 left jabs, hope they land, and Hoy should go dizzy!
For NTSC, there’s an elaborate reason for real time optimization to knock Hoy down with a jab, but in the PAL version, we need the lower health refill from the super knockdown, so we do that instead by buffering a right gut punch into a low super. This phase 1 is interesting because it prevents Hoy’s move timer from advancing as far as the NTSC strategy, and Hoy ends up being setup for something pretty neat in phase 2.
For the start of phase 2, Hoy immediately throws 2 consecutive hooks, so you just have to buffer 2 hooks to intercept both attacks. You do want your hooks to be left and right sided, since you need the extra damage from your right hook for this strategy to actually knock Hoy down.
Once you land your 2nd hook, Hoy will be in a state where he’s receptive to having his guard manipulated, since he’s been immediately throwing attacks and hasn’t been idle for a single frame yet. Because of this, you must manipulate Hoy’s guard before you can throw your next punch and expect it to actually land.
After you land your 2nd hook, you need to delay for one frame without pressing up, then on the next frame, you need to press up & A to throw a guard manipulated uppercut. I timed this around when Hoy’s recoil animation was ending for the 2nd hook I landed. If all goes well and you get lucky (5/8 likely), then the uppercut will land and Hoy will be sent to the mat for the 2nd time.
For the NTSC version of the fight, Hoy’s move timer will be in a spot where you cannot land a guard manipulated super before Hoy’s next attack comes out, so you’re a bit limited in what your best time can be, since you need to land a super directly into that next attack.
However, with the PAL version, youre earlier in his pattern to where you *can* land a guard manipulated super early, so thats what well do. Start phase 3 by holding Up for 1 frame, then for the next frame, let go and press A for a low super. Assuming you get the 5/8 luck, the super will land and hell be down for the TKO!
For the PAL version, Hoy ends up arguably being the hardest fight to match the TAS time on that isn’t Narcis Prince. The execution for Narcis isn’t bad, but the luck is really rough, while Hoy requires you to hit 5 frame perfect inputs while dealing with the 2nd worst luck needed for a fight in the game.
Heres a breakdown all the luck needed to land this strategy:
Phase 1:
-Landing an unstunned jab (3/4)
-Landing a buffered unstunned jab (9/16)
-Landing a buffered unstunned jab (9/16)
-Landing a buffered unstunned jab (9/16)
-Total Phase 1 odds: (3/4)*(9/16)*(9/16)*(9/16) = 2,187 / 16,834 = 1 / 7.6973 (or ~12.99%)
Phase 2:
-Landing an unstunned uppercut (5/8)
Phase 3:
-Landing an unstunned uppercut (5/8)
Total fight odds: (2,187/16,834)*(5/8)*(5/8) = 54,675 / 1,048,576 = ~1 / 19.178344 (or ~5.21%)
I got extremely lucky when I did this fight, because on my very first TAS pace attempt out of phase 2, I hit the final punch in phase 3 perfectly while getting the best luck. It took me roughly 2 hours of effort to match this time because of that. I had 5 perfect phase 1s where 4 of which Hoy dodged in phase 2.
Anyways, thanks for watching!")
![Splatterhouse (Arcade) - Speed Run [13:36]](https://i.ytimg.com/vi/KehaX8Mus6g/hqdefault.jpg "Splatterhouse (Arcade) - Speed Run [13:36]
Happy Halloween everyone!
This was played on an original Splatterhouse arcade PCB on the H.A.S. supergun. My Splatterhouse PCB also now has the EVERTEN stereo expansion board, so the audio is more clear now with separate stereo channels! This run was done on the hardest difficulty, which starts you at 2HP instead of the default 4HP.
This was also recorded through the Datapath E1 capture card via RGB and upscaled to 4k, so this quality is about as crisp as you can possibly get on real hardware. For comparison, you can check out my slower 13:37 run from a while back which was recorded via S-Video to see how much better the recording quality has gotten: https://www.youtube.com/watch?v=Uvgl8DadM6M
It is also worth noting that the game is running at a slightly higher framerate than the home release ports out there (JP Wii VC at 59.94 fps, and the Namco Museum version runs at ~60 fps). This board appears to run at about 60.60 frames a second, so this benefits from small time gains compared to those versions.
So stormcrow56k picked this game up for speedrunning recently, and found a couple of new strategies in a few spots. The main optimization he found was for the stage 2 boss, where he found an approach to kill the chair with one punch and one slidekick, however its extremely difficult and inconsistent, so he also found a backup strategy that minimizes time loss if you dont get it, which helps justify going for it.
Beyond the chair kill, theres a really hard strategy that requires good luck on knives such that they cross in the center of the room, as well as a really difficult jump-punch strategy to take out the 2 airborne knives, which requires pressing jump and punch on the same frame to get Rick to carry his punch state into the air and punch twice in the air. Normally when you do a jump, your first attack will always be a jump kick, and any subsequent attack will be a jump punch (unless you hold down for more jump kicks).
The other big find that stormcrow found was for the Jennifer boss fight in stage 5. Effectively, you can exchange one of the punches early in the the fight with a very late slide kick for the last attack, which saves you one attack and nets you a tiny bit of time.
As far as the overall run, I messed up the early game pretty bad, as stages 2/3/4 all have pretty significant mistakes. Granted, I did fix the problem I was having with the stage 3 boss, which helped, but I had issues picking up the shotgun after Rick accidentally picked up the rock, lol.
To make up for these shortcomings, I ended up getting pretty sizable gold splits on stages 5 and 6. Both were really good, but the stand out stage was definitely 6, with a very very good boss fight with favorable enough luck that allowed me to keep attacking it for big damage. The exact time of this run is a 13:36.961, which beats the old run by almost 0.4 seconds (old record was an 13:37.357). This ended up being a sub 13:37 by only *3* frames.
Still going to be aiming to beat this by a sizable amount, because I had way more promising attempts that kept getting shot down by either bad luck, or critical mistakes. Its kind of bittersweet that the record for this game is no longer a 13:37, but it had a good long stretch at staying leet for a few years, so it had solid staying power!
Anyways, thanks for watching!")
![Mike Tysons Punch-Out!! - Soda Popinski [0:46.00]](https://i.ytimg.com/vi/Kp1qkWNhxGo/hqdefault.jpg "Mike Tysons Punch-Out!! - Soda Popinski [0:46.00]
This run was done on an original cartridge of Mike Tysons Punch-Out!! on an NES that was modded with the Tim Worthington NESRGB kit. For some more in depth information about my RGB setup, you can check out this pastebin here: https://pastebin.com/m7tzR5Cq
Ive also decided to include a non-stretched version of this fight in an unlisted video. If youd like to learn more about why Im including this too, you can check out the description in the second video here: https://www.youtube.com/watch?v=lTeOdzExiEA
Ive also got my full attempt session from power-on to the successful attempt right here: https://www.youtube.com/watch?v=FLKKUUnPdfE
Heres a fight I havent improved since my 46.25 I got when I initially beat Matt Turks final untied record in Punch Out in January of 2015... almost 8 years ago. I had asked on one of my streams what my most likely IL PB would be for me to get in this game, and the suggestion that stood out was Soda. Since I got my 46.25, there had been a new development for the fight that allows for better luck, which in turn would make it more reasonable to push luck in other ways.
Soda is a very fast opponent with a couple of random elements, but has a massive weakness to exploit. The fight starts off with Soda attempting to throw 2 hooks at you, and they need to be cancelled immediately with 2 blocked hooks from Mac. Its more consistent to throw 2 right gut punches, but you can opt to instead buffer a left gut punch as your 2nd. It will save a couple of frames, since left gut punches are faster, but it will also whiff about 25% of the time, which would instantly end your attempt. I opt for the faster 2nd cancel.
After the cancels, Soda can take one of 2 different routes; throw 2 uppercuts, or throw 2 more hooks. Hell throw the uppercuts about 75% of the time, which is part of the luck we need to continue the strategy, but we also need something else that Ill go into next.
Off of the 2nd cancel, a right dodge is buffered, which is a perfect setup to land a gut punch into Sodas ducking uppercut stance, assuming that he does the fastest delay anyways. Soda can do one of 4 different delays, each 25% likely. When Mac is about centered, you need to buffer a left gut punch, then immediately press Down as the punch goes off. This technique is called the Screwdriver, and it was discovered by Jack Wedge days before I set my last PB in this fight.
A somewhat newer discovery is that by holding Right for the entire duration of when youre doing the Screwdrivers dodge, you will effectively eliminate 2 of the 4 possible delays for Sodas uppercut, making it ~50% likely instead of ~25% each time its used.
Once the gut punch lands, you just need to do a delayed gut punch to cancel his 2nd uppercut, and buffer a star punch out of it. No matter what Soda does after that, the star will land, and since you landed a gut punch while he attempted to throw an uppercut, it enables Soda to be knocked down instantly from a star punch!
Phases 2 and 3 are the same, and they can be handled in a couple of different ways. The traditional way is to counter his first hook with a right jab, then buffer a right gut punch, then perform the Screwdriver just like in phase. I opt for a harder variation that theoretically should save frames over the standard approach though.
I instead cancel the first hook with a right gut punch, then throw a jab into his 2nd hook. You are supposed to buffer a jab and let go of Up on the first possible frame, which allows the jab to be bufferable and land, but Ive opted to let go of all inputs as soon as I land my gut punch, then quickly press Up+A in isolation, which theoretically should be possible to land on the same frame as the buffered one, but doesnt kill an attempt outright if I let go of Up a frame late like the other approach. I may switch to the fully buffered method later when saving those frames become more necessary.
I end up losing quite a few frames from all these adjustments, probably more than if I had just played it safe and done the standard strategy that I got the 46.25 with, and that could be attributed to not being used to these timings yet. However, despite my mistakes, it seems I saved enough time for a meager total net gain of frames, and scraped by with a small PB.
This took about 20 minutes, which you can see in the unlisted attempt video I linked above. My chat was pretty spot on with how easy it would be for me to PB on this fight, haha. Ill definitely beat this, hopefully sooner rather than later.
The world record is a 45.00 by lucandor158, which is leagues more advanced than what Im doing here. Ill have to lower my PB a ton before I ever even consider attempting his strategy, because his fight is absolutely amazing. If I ever switch over to that strategy Ill talk about why its such a monster of a record, but until then...
Thanks for watching!")
