Always thought DS games were capped ad 30 fps for some reasons, but NSMB does seem smooth.

Other precise 'tricks' I know are: -Jumping off a switch without landing on the surface, activating or not, depending on which frame you press jump, seems not possible on boss switches (I guess 1 frame each, 1 before landing on the switch for not trigger the switch, same frame to trigger it). -Shooting fireballs missing koopas, on the same plane surface (don't have an idea on how it's precise) -Double/triple jumping without holding the run button, preserving all the momentum you have (~1-2 frame window,quite sure smaller than blue shell double/triple jumps) -Double/triple jumping bouncing off an enemy/falling from a surface (same as the last one, I guess the game checks if Mario has the the right conditions to have a normal double/triple jump window that's around 10 (??)) -Jumping off those similar to launchpads things in 4-Ghost House, freezing some elements on the level (remember it being precise, ~1/2 frame window). Getting flawless mid-stair without any slow down at full running speed (stairs like the ones at the end of 1-1) -Skipping cliffs without the fire flower and losing momentum (I think 1 frame before landing on cliffs??) -All the tricks relative to fireballs (they all have the same window, I guess around ~3/4 frames) skipping cliffs skipping launchpads losing no momentum on slopes (making levels like 2-3 a few frames faster) -Doing the 8-castle trick a little before the flag (very precise, don't have an idea)

Not completely sure about any of these, also don't know about about vertical-mini jumps too I guess TASers are the people who know the most about this stuff.

I meant to use those loading times as DS times (w loads) and the actual time (w/o loads). But now that Walrus got all the loading times, it seems a more viable option.

Found an easier setup for the Bowser Jr. battle a while ago but I forgot to upload it .-. It's about 0.5 slower than current faster strats.

Going to upload soon and post it here

There's no problem about said accuracy, I guess tenths are enough, and ty for EMU times too, that's a really big step. Mmh, the thing about adjusted time is that you can't sort runs by a variant, it's not a real problem but I'd prefer using another solution, like: -using with / without loading times columns (meh, kinda confusing) -adding time to EMU/VC runs and stating on the description of the run what would have been the times with different consoles -something else

I'm not sure I'd be able to do some races, anyway, here's my skype -> Selfishigheels

I guess we're just going to add times to VC/EMU runs.

http://imgur.com/Mq3OBQf Variants don't work as I wanted, you can't give a default order by variants' values.

@koffingrockz Solid proofs are always the best thing to solve it, even if his arguments may seem absurd.

I'm testing some stuff on leaderboards, nothing official

There's no problem in using even milliseconds in leaderboards, I'll add that.

@MyLittleWalrus Is that a round 45.50 or you just didn't estimated with enough accuracy? Just a friendly question. Also, your spreadsheet is private currently.

Thanks for making any% a good category (ﾉ◕ヮ◕)ﾉ¤:･ﾟ✧

Risky but nice.

I wouldn't allow super stars, unlike 1-Ups, they directly change the gameplay itself.

True that, we choose to "eventually" pick star coins, so that's not a strict low%.

Hearing others' opinion will help.

I'll try that. I love you.

Thank you so much for taking your time to solve this, things on leaderboards will definitely be changed. Ye, I think it'd be cool to provide the data you have.

Glad this is a consistent thing.

Oh, it's actually said in the title, sorry.

Having low% in IW is fine. Not really the same for IL: it would coincide a lot with normal any% ILs. I'd personally leave another "low%" variant as it's been done for banktoads in Super Mario Galaxy 2.

So I guess what you want here he's a different low% than we already have. Go ahead defining better your idea, still it'd probably be too arbitrary, at least for now.

@LightOfEon Ty, still a lot of optimization ahead for Any% NSE, a perfect run would be roughly 1:01:45 (I promise video proof is coming).

And yeah, I'll put standards on category rules (all of them,main and misc, at least for now).

I'd really like to know what Jumpyluff thinks about all of this.

I do believe your runs are legit. But accepting every WR runs by trusted people w/o video proof is not possible. I'd like to avoid the situation NSMBW leaderboards have been passed through a little ago.