Haha, this is pretty amusing. I've ran the board pretty loose since I was just added by ZS45 to help him out, but in the end I suppose I've taken over most of the operations. The rule requiring video proof for emulator runs is something I wrote some years back when the game was added, and then I subsequently seemingly forgot about it. I'll go through and retroactively reject the ones you've listed since that's what should have been done to begin with. It's not enough that the times are within reason (and I personally believe the submissions); the rule has been there since the start.
as an fyi: you don't have to click follow to have access to the forum dedicated to the game. the full list of subforums is hidden by default on the main forum view, but you can select to show all (which lists a stupid amount of series/games). alternatively, you can go to the specific game leaderboard and find the link to the forum on the sidebar to the left.
kaiba, hakumei to mikochi, flying witch, kuma miko, in this corner of the world, moomin (1990-1991 run)
Been thinking of attempting a better level 0 start run, might as well go for it on 1.0 for the input buffering alone! The difference in level increases shouldn't have any effect on that run if I understand correctly.
I've added the game with the base modes as categories: https://www.speedrun.com/tetrispsp I think Marathon (level 1-15) is the best category for this. 10,000 points is simply too short for any meaningful competition. I also added the variant modes as Sprint subcategories, since the common denominator is that they all have as a goal to clear 40 lines.
I don't think it's necessary to add any more categories, although I suppose I would be open for something like 100,000 points or clearing Marathon from level 15 as a shorter category.
For what it's worth (although I don't run the game), I'd be in favor of a ban considering the precedent (banning it in score attacks and tournament play). It doesn't change the state of the game beyond shortening the run, at the cost of spending more time setting up the glitch. If we assume that the glitch would be allowed, I can only imagine it'll be a point of annoyance for runners to have to set up the glitch between resets, just for the sake of a measly ~10 piece head start. It would also heighten the bar for entry somewhat because it's no longer so simple as to just start the game and play in order to grab those free few seconds.
Surely not directly comparable but in Super Monkey Ball, in order to reduce load time at the start, the first level is loaded in Practice mode before menuing over to start the actual run. An example can be seen at the start here https://www.speedrun.com/supermonkeyball/run/yl96j52y
An example in a more related game, Dr. Mario for NES, determines piece sequence based on the frame you start the game. This can essentially eliminate the randomness of the run. Some discussion can be found here https://www.speedrun.com/drmariones/thread/derds
I found it by accident during a race, after that there was a small effort in trying to find other places to utilize it. Unfortunately, it seems to be limited to Trashing the Camp due to Terk's... suboptimal hitbox, and the one (or two) other spots where you can fall through the floor doesn't lead to the goal.
I tried it a few times during a test run on N64 earlier today. If I recall correctly it has been done in a TAS, but I couldn't get the timing right in my <10 attempts today.
You bring up a valid point. Even though the core of the game is the same between N64, PS1 and PC, they do play out differently in various ways. I also like your argument that splitting the leaderboard up between platforms is likely to spark more competition on the PS1 and N64 side; that would be great.
I think the only real argument against this is that it'll get messy with 2 categories (any%, 100%) and 9 subcategories for each (easy/n64, easy/ps1, easy/pc, medium/n64, and so on). The reality of it is that most categories would end up vacant or not worthy of being taken very seriously.
I honestly don't mind setting up the leaderboard like this, but I'll direct some other people to the post before I get to work on it.
Edit: I'm stupid, I've split it up in a more sensible way which TheGreatestRoman pointed out to me. While I was at it, I changed 100% to "all letters/maps", the coin requirement has simply made it too high of a barrier for entry.
Your assumptions were correct, I had already uploaded when I noticed it and thought it was okay anyway. In hindsight (obviously), I decided it would be for the better that the cut was clean, so I reuploaded it!
The emulator I used was Gambatte. Regarding the beginning of the video, ffmpeg had a little hiccup it seems. I still have the original recording so I'll upload a version with the start intact!
Edit: I've resubmitted the run with a better video,
Yeah, it seems to be RNG. I think I've had better success with a certain timing, but I also think that might be placebo.
I'm fairly certain that's perfectly fine in the case of the GB/GBC. I don't think there have been any capture mods released for those handhelds and I also assume such a thing would make them rather cumbersome to play on.