I think he did not use an emulator. He says he played on original console in the description of his commentary
lol. This is such a bizarre performance. The plan behind his route is strange. He's incorporated so many specific events that should be considered universal by now, but ignores so many others. Like, someone who can do this has obviously been studying the game, and probably studied Challenger's TAS. But it's just so weird to blatantly make so many choices for slower patterns at certain places. It wouldn't even make sense to say he's playing conservative in some places, because it's just wrong. And meanwhile, throughout the run, he's doing lots of insane guard skips that go way beyond dangerous. If you have time to practice all these guard skips to the point where you think they aren't going to ruin your run, you should be implementing the easy stuff too. Easy example is the music pause at the magic mirror in Level 5. How does he not do that? And yet he knows about the music swap at the Level 19 super boss fight? And how is he gonna miss one of the kill jumps during the super boss fight, but nail the zombie finish on Level 17? lol. Then in some levels, his precision is remarkable, and in others, he's making ridiculous mistakes, or again, skipping easy tricks as if he doesn't know about them. Very very strange.
It's kinda like Niftski going for a lightning 454, but then using the coin block in the pipe jump room on 8-4.
This japanese guy's run is pretty much incredible, but he could easily be shooting for a sub-35 if he'd just do it right. hahahaha. He's done the impossible stuff, and skipped the easy stuff. XD
Ok, I'm done being critical. And I'm not one to be talking, because I'm unable to do this route currently. I haven't even started trying to learn all those guard skips. My mind is too full right now, trying to remember all the little details for the sub-7 warp route, and the sub-23 warpless route. XD
It makes sense to me that he knows about some tricks but not others, because as far as I can tell discussion regarding this game is pretty scattered. Here, and in some TASVideos forums. "Universal" doesn't mean much in an universe consisting of basically you and challenger, lol.
lol. I guess my point is, there's so many fingerprints from the TAS's in his performance, of things which do require lots of practice to remember and do correctly. So it's just kind of funny to me how this Japanese guy was just kind of selective, and learned a lot of the tricks, but also didn't bother to learn a lot of the others that are readily available from the same sources (the same TAS videos).
With a game like this, learning the right patterns and tricks just saves soooo much time. And when I say "universal," I guess I do mean specifically what is laid out in the various TAS's. Because that's a pretty easy resource today, for anyone in the world.
When I got the WR in March of this year, I had watched the 33:31 TAS and the current 3rd-5th place runs to take what I could from them, but I couldn't do them all, so I developed some slower but safer strats for some guard skips/shortcuts. So maybe this runner didn't know how to do certain strats or failed them during the run and went for others/improvise on the fly. That happened to me in Level 19, I failed one of the jumps because of the nerves of being in WR pace. I should also point out that this runner is also playing in the japanese version of the game, where the jingle skip is not optimal as there's no Konami jingle, and the shortest jingle available is a bit far away. It is still optimal for level 19 skip tho. I agree with GMP that any info of this port, speedrun or not (like development for this port), it's hard to come across even by searching. I didn't know about the speedrun community until last year and I was doing mods for this SNES game for almost three years.
Shauing, have you thought about taking back the record? If so, I have some advice. If you've seen my "spliced" videos of all-20-levels in under 23 minutes, you know I spend a lot of time playing the levels individually. This is actually fantastic practice for a full run. If you take the time to map out individual levels, you can get as close to TAS as possible (for almost all tricks). It's pretty easy, because of the effective 15 fps input timing, as opposed to a game like Super Mario Bros 1, where "frame perfect" really does mean at 60 fps. After I spent days and days and days on a single level, recording video after video, and using the password system to quickly confirm my progress, I could really piece together a lot more detail for a full run, instead of just "winging it" in certain parts.
If you spend a few months building up a private archive of Personal Best videos for each level, that will translate to better performance automatically. I know nerves can be a big deal towards the end of a run, but the timing system in this game is very forgiving, and if you put enough time into it, inaccuracies due to nerves can still fall within the margin of error after enough practice.
After I built up a full archive of recent PBs for every level, then I started do groups of levels, like 1-5, 6-10, etc., and then finally start doing full runs again, when everything is fresh and sharp. Some may argue that this method is not necessary -- but I say based on experience, it is a guaranteed plan for success. lol. One of these days, when I'm not so busy, I will definitely try to do all my PBs for the all-20-levels under 23 minutes glitch route, and see if I can finally put a good run together that will justify redoing the categories, as I proposed previously.
In response to your post, I think you make some good points. Perhaps I overestimated how easy it is for other people to find all the TAS and RTA info for this version of the game.
I might consider to take it back if there's more new strats/tricks to implement (no warps of course, that's a different category); there were a few guard jumps and tricks from that 33:31 TAS that I had to do differently because the timing was quite tight (not frame perfect I believe, but still really tough).
After the last rule change from May 30th, 2023 the runner @Akuma_apn has brought forth some compelling arguments as to why the run should start "on the first frame of the first active screen of the game (after the title screen)" (see ). As