10 monthes earlier, I have submitted a run that improves most of the times. I did have the feeling it wasn't perfect as there was one thing that was annoying me. I have submitted it because there was a discussion about Super Punch-Out on IRC and I told adelikat about my improvements. He was interested and I've decided to submit it so he could watch it and maybe improve it. The other reason is that I was fully busy with Legend of Zelda, Castlevania and SMB2 (Which was never finished but maybe someday...) and seriously, I didn't really care about that game when I think Zelda and Castlevania movies are great TASes.
The thing that was annoying me is Bear Hugger. Explanations below.
Not that it really matters as I don't care, for those that are interested, that movie is 31 frames faster than adelikat's published movie.

It's important to mention that these times are made on the US NTSC version of the game. We often see people claiming false records, such as georgiek[1] in this submission thread or telling us that someone made some records on some old magazine which the staff of the magazine couldn't verify if some cheat devices could have been used or not. Or people posting about some Youtube video such as MeLeeak's videos, http://www.youtube.com/user/MeLeeak , but they are made with the PAL version which does have favorable different timing unlike US NTSC.

So, you should be aware that if you think or saw a faster time:

1) Make sure it is legit. No cheat devices used or an hacked ROM.
2) Verify if it's done on US NTSC version. The PAL version does have different timings and maybe the Japanese NTSC one too. If all is fine, me, and I am sure adelikat too, would be pleased to see a new strategy.

Also, for those that are unfamiliar with Punch-Out games, take note that delayed punches are intentionals. In this game, it either was done for extra damage like in NES Mike Tyson Punch-Out or/and for giving extra special power bar or/and for extra dizzy counter points, which can only be seen with memory address 7E092B. The dizzy counter is important because as soon a fighter is dizzy, sooner you can put him down.

There are many type of punches in that game. I feel it's needed to list them and what are the advantages and disadvantages for each of them.

There are 4 normal punches. When I say left or right, I am not talking about whether it's opponent's left or right but the player.

Left face punch: The fastest punch but weakest punch.

Right face punch. It is usually slower than the left. However, depending of the opponent, it can be as quick. It just depend of how long the animation of the opponent last after receiving the punch. Also, sometimes, the right face punch can do more damage, also depending of who is the opponent.

Left body punch: It is usually the same speed as the right face punch and do the same amount of damage. Again, it depends of who is the opponent. There are somes that do have a weak stomach. ;)

Right body punch: It is generally the slowest but the strongest of all normal punches. If the opponent's weakness is its stomach, that punch is really useful as one punch is almost equal to 2 left face punch.

There are 3 special attacks which can only be used if the "S", which stands for Special, bar is full.

Right hook: Sometimes referred as Uppercut which I don't agree as the animation doesn't look like an uppercut but a hook. It is a hook delivered at enemy's face.

Left hook: It is faster than right hook and delivered at opponent's body. Sometimes it can do more damage if enemy's weakness is its body.

Also, doing a (left of right) hook gives lot of dizzy counter points.

Rapid punches: Sometime referred as Rapid Hook, if it hits the body or Rapid Uppercuts if it hits the head. Obviously they aren't uppercuts. Whether it's hooks is arguable. But after reading the hook page on wikepedia, I've concluded they aren't hooks.

Anyway, rapid punches aren't much used in this run as it is usually better to do a hook than those because hook gives more damage in a shorter time, and most of the time, rapid punches doesn't give much dizzy counter points and sometimes none at all.

I have retested all the fighters and even retested what I already tested in my previous version.

In previous submissions(Almost all), there was lacking many useful informations for viewers. And also, people weren't credited as much as they should. I'm trying to be as best and fair as possible in this section. If it's missing an important information, just post a message or pm me. In this section, when I'm talking about Jarboo and some other unknown names, I'm talking about this. Also, when I say, "he is the first person to achieve it", it is whether it's tool-assisted or not.

Gabby Jay: The record is 6"26. Gord Ryan is the first to achieve such time. Unless someone find a bug, I doubt it can be improved. Also, that record can only be accomplished by using a pause trick. And only when paused, the special bar can be full without costing in-game time.

Bear Hugger: The record is 12"92. It is a new record, the reason of this submission and the only one improvement. The precedent record was 13"49. So it's 0"57 faster. I didn't like the previous strategy. I fell there was an extra unuseful punch. And I didn't see that fight improved in adelikat's run too. That's why I put that project in my "future but not urgent project list". But today, that thing that was annoying me for monthes is now gone. Also, Jeff Romanko did managed to do a time of 13"47 which makes me suspicious toward current time. I did managed to get a time of 13"43 with older strategy. Then got a time of 13"1x after quicktesting a new strategy which makes me happy ;) . Then after optimizations, I got the desired results.

Piston Hurricane: The record is 5"49. The first guy to my knowledge to had achieved such time is Gord Ryan. Most people, including older TASes, could only perform a 5"50 as the 5"49 strategy wasn't available. In my previous version, I did find the way to do it. It was quite obvious... I wonder how we missed it.

Bald Bull: The record is 8"24. adelikat is the first one to achieve such time. In my precedent version I did also improve that fight, mainly the first part of the fight. At second part, either we do an uppercut then a punch to put him down a second time like I did. Or do the opposite, do a punch then finish it with the uppercut like adelikat did. Adelikat's choice is the best. If we do an uppercut, the animation lasts long. So it costs many in-game time if the fighter's energy doesn't go down to 0. Finishing him with an uppercut is payful as the timer stops several in-game milliseconds before the animation is over. I think the animation is cut-off by a little too. That may explain why adelikat gains some frames over my version when theoritically, it should have been equal. Now, I know I did an obvious error.

Bob Charlie: The record is 5"50. It is another one which I don't know who is the first person to achieve such time. So many people could do that time on console that it may have been done during SNES era.
^[1] georgiek wrote: "i said it a year ago i will say it again bob charlie can be beaten in the 4:xx seconds bracket i have done it on console he has random factors can someone please spend hours on this i promise you it is possible i said it a year ago but no one has done this yet slackers!"

My answer:

Bob Charlie doesn't have much more randomness as Super Mario Bros 1.

Left face punch is the fastest punch in the game, see earlier for punches details, except the right face punch that was needed for doing extra damage, all punches are left face punch. And 2 uppercuts. Also punches at stomach aren't beneficial.

I think the current strategy is as optimal as possible. Someone would have said 5"49, I would have some doubts but after retesting it, I am confident with the current strategy. I think the chances of being improved is nearly impossible.

Dragon Chan: The record is 7"30. It was firstly done by me in my precedent but unpublished submission. With the PAL version, I think it might be possible to achieve a great time, maybe sub 7 seconds. During the process of making my precedent version, I discover a good strategy but it is missing 1 frame to achieve it. There is no way to make it workable in this version. :(. Anyway, 7"30 is not bad when it has been long time at 7"49.

Masked Muscle: The record is 6"33. I don't know between Jeff Romanko and Juan Sebastian Arévalo that made the first record but I congratulate them. It's possible to put him down later, ~0.3 game timeunit, and wakes up with less energy, sufficient to put him down with only one uppercut instead of one punch and uppercut after first countdown. It would have saved some time. But he starts with a different pattern that is really unfavorable. I think that pattern is affected by either you start punching at his face or his body during the first part of the fight. If it's the case and I am 98% sure it is, I doubt a new strategy could be find for him.

Mr. Sandman: The record is 13"13. Me and Genisto were the first ones to achieve such times. Note it is now the longest time. Bear Hugger isn't no more. So, "Good Night!"

Aran Ryan: The current record is 8"45. Many people could achieve that time on console and I don't have a single clue who was the first to have done it. It is also one that can be done at full speed by using the strategy in this run. It doesn't mean you will do 8"45 but you will be close.

Heike Kagero: The best time is 10"60. Many people had done that time.

Mad Clown: The record is 9"05. I did find the current strategy in my previous run. It is one that I am proud. I took several hours finding that strategy. I didn't know I could find such big improvements when adelikat had already improved it.

Super Macho Man: The record is 9"43. Done by me and Genisto, it hasn't been improved since then.

Narcis Prince: Fastest time so far is 8"43. Found by me. It is also another one which I am proud. I found an unexpected strategy involving 4 rapid punches which is cool to watch.

Hoy Quarlow: 8"35 is the current record. It is has been done by many people and I don't know who is the first guy that made it but something that I am sure, it can't probably be beaten.

Rick Bruiser: 10"67 is the best time I had found in my precedent version. That's another one that I am proud because I remember Genisto who has past several hours and hours doing the 10"71 version which were thought to be optimized. I also took several hours to improve it. :P

Nick Bruiser: The record is 7"52. I am not sure if it's me or Genisto that obtain that time first. I know adelikat got 7:54 but Genisto has also done a 7:5x version not so long time after our first version. I am not sure if he did 7"52 or 7"54. It wasn't published because he thought he could improve it furthermore then is "Hard Drive disk" got corrupted and he loses the run :(. But I saw the video with my own eyes. I know it's not fake. However the technique to do the 7"52 movie is something that I have used for several fighters and was discovered in my latest version. If Genisto would have found it, I am sure he would also told me other fighters could have been improved. So, I am not sure. Anyway, I couldn't improve it furthermore.

I "rethank" adelikat and Genisto for their work in previous TASes. I also thank people who indirectly help me doing this run.

NesVideoAgent: Hi! I am a robot. I took a few screenshots of this movie and placed them here. Here goes! Feel free to clean up the list.

adelikat: Nice improvement, accepting for publication.