right, but most games start timers when the game content starts, so wouldn't it make more sense to start the game at file select?

No, glitchless categories make perfect sense to me, and obviously they should be separate from glitch...ful categories, but saying some random category wouldn't have a reason to exist is a bit ridiculous to me. If someone's wiling to run a category, within reason obviously (with or without glitches), why should it matter how many stars it is?

Also starting timers at power on IS ridiculous. Why not start at turning on the TV while you're at it?

How is it ridiculous? If someone wants to do a 70-star run and use whatever glitches they want, they can, and if no one else decides to challenge their run with another, then hey, free WR.

Also, care to name other game that starts at console power on? I'm curious to see if another one exists.

I don't deny that this is a great aggregate site for tons of speedrun stuff, but again, if someone posts a run somewhere else that is the fastest time, it's still the fastest time.

As for the "spewing shit" comment above, I think I've only posted in one other thread about the start time to the game, which is still ludicrous imho.

BTW, if you haven't figured it out, I like to challenge the norms, I'm certainly not looking to make enemies (or even friends tbh), but if I see something ridiculous, sometimes I like to poke the hive.

Saying I don't even play when you have no idea if I do or not...

What kind of authority do the people on this site have to determine if something is "world record" or not. Just because it doesn't exist here, does not mean he can't hold a record. If someone did a run and recorded it and put it on youtube and not here, and it was easy to see it was the absolute fastest time, would it count, or would they have to submit it?

I could just as easily say I don't see any of these records in Guinness World Records, therefore they don't count. This website has no monopoly on this subject.

just because it doesn't exist on this site doesn't mean he can't have a world record in it.

I'd say if you want to do a 70 star run with glitches go for it. Free world record if no one else wants to do it.

If you have 100$ to spend might as well buy an N64 and just play on that. Should be able to find one, a copy of SM64, and a decent controller for not much more than that.

I almost exclusively use my ever drive to play N64 these days, and I've never noticed any differences in games vs the real cart.

The only real difference will come with needing to save/reset warp in games, since the ED64 can only reset to its menu, and not the loaded rom, so that does add time, but for games that don't do that, and start their time on file selection/new game, like all of them should anyways, it's a good alternative to the increasing costs of some of these games.

I don't see how anyone could actually verify if a run was done on an ever drive vs an official cart, considering its the exact same rom running on the exact same hardware. It would be 100% indistinguishable.

[I have tried replaying 120 Star TAS with one frame delay, and it got desynced after few minutes.]

I fail to see how a hard coded set of inputs matter here. Are you saying that the frames from the time you select the file, to the time you can advance the opening text box are different, or that it behaves different in some way on a console that's just been powered on vs one that's been on for an hour?

[Starting from power on is the correct way for the official leaderboard, there's nothing subjective.]

It's still subjective since it's based on a consensus of what people THINK should be the "correct" way. (And im not saying that starting on file select is NOT subjective either).

[There's no reason to use file select timing except for it being ¤arguably¤ more convenient. Changing the timing standard would require a lot of work for pretty much no benefit.]

More convenient, saves time instantly (i figured that's what speed running was about), and how many other games start on a system reset? Seems like SM64 is a bit of an odd man out in this case.

Is there any proof that a game having been on after so much time results in a different number of frames having passed between 2 points vs the game being powered on? Obviously the total frame count will differ, but if we count the frames between some point of animations, will there be a difference?

"Correctly" is completely subjective though.

Also Jumpyluff, are you saying that there could be discrepancies in timing if someone were to select a file at X frames on console power, vs someone who waited longer and selected a file at Y frames?

It's the principle of there being no reason to start it on reset over file select. Would a run even be accepted if I submitted one starting at file select?

Start it at file select, then add 8 seconds if you care to figure out how it would compare if you started on reset. It's far more logical to start on file select.

I've been trying to find a real answer to why runs start on reset/power-on for a while. It's clear now that there is no real reason behind it. Tradition is probably the worst reason as well.

I'll always start my time from file select, and if anyone cares enough, they can just tack on the few seconds if they're that anal about it.