The Switch version of SMB3 is now allowed for all 4 categories of Super Mario Bros. 3 (any%, any% no WW, any% warpless, 100%).

Note that the Switch has a framerate very slightly slower than the NES and you'll lose a few seconds over the course of a 100% run due to it (the difference is quite negligible unless you're going for the WR).

In the case of any%, note that the wrong warp setups possible on NES (that you can see all over the LB currently) are the standard we use to determine which version should be allowed or not. For example it was found out that WiiVC behaved differently in that regard and as a consequence was banned for this category.

What it means is that with the informations we gathered so far, it seems that the Switch behaves similarly to the NES, and NES setups have been found to work on the Switch. This however might be reassessed in the future if new data comes up, with Switch specific possibilities for example. If that were the case, the version might become banned for this category alone, but we would still grandfather in older runs using "NES compatible setups" so don't be worried.

All in all, if you're planning on playing any% on the Switch version and are noticing something in 7-1 that seems to differ from the NES, or a setup possibility that nobody else seems to use, please contact us (on Discord) and we'll look into it if necessary.

We're still trying to figure out a couple of things about that version. Hopefully we'll make an announcement about it soonish.

You're going at the last shell seemingly randomly with your tail whips. The movement to the koopa from the pipe and then the two tail whips are a bit precise, you can't just hit it at any timing and expect to get the good value, unless you get lucky. I suggest trying to copy what successful runs with this setup do.

NESClassic is fine for any% no WW, any% warpless and 100%, and yes the clips are still doable.

It's banned for any% though.

I have to say I don't have any knowledge about this specific emulator/core. Maybe there's informations on tasvideos, or hopefully someone else will know.

About the rest: streaming is not required but depending on your claim (like if you're gonna submit a top time) we can ask for additional evidence of the grind you did, stuff like that. So typically if you're only recording locally and are going for a top time, keep your recordings. Streaming is obviously a plus in that regard.

The main and only thing in the recording is to be able to see the entire gameplay clearly and uninterrupted. With sound preferably. Again a run without sound might or might not be rejected depending on the claim.

Facecam is irrelevant. Input display is highly preferred if you're on emulator. A timer is not required but it helps us a lot for verification & retiming purposes so if you can include one, that's good.

Sorry for the lack of replies to the thread, so far we have no idea about the accuracy of that version regarding the LBs. i think it will probably be fine for no WW, warpless and 100% at least (but that's a personal opinion) but it's hard to say at this point for any%. It might end up banned like VC and all for that category if it doesn't handle memory stuff similarly to the NES (given that the normal setups seem to work, it might be an accurate enough emulation though).

Basically it still needs testing. If you're planning on being in the LB, i'd recommend not going hard at it for any% at least (and instead use one of the already accepted platforms). For other categories, it'll prolly be fine (haven't personally looked into it but given what i read/heard).

the problem is not the emulator you use but the rom you use, which is PAL (runs at 50hz instead of 60 so the game is slower. but in the case of smb3 the jumps (& drops in this case) are shorter as a side effect)

you wanna play on NTSC (US)

Ah sorry I didn't bother giving a detailed explanation in the rejection comment because I genuinely thought you were just submitting as a joke (particularly considering it wasn't even a PB).

Basically when submitting runs, the gameplay should be clearly visible at all times (which isn't really the case here, far from it). Additionally, we wouldn't accept runs with cosmetic mods for example. And although this isn't such a mod, the end result is the same. So yeah in short the game should look like the actual game.

Oh sorry you put the names in the other order so I didn't realise (I mean, i knew you were on the right in the video, just didn't realise which name was who). Should have noticed with the times in the title.

About the theory, it's already been confirmed by or4ng33xp0 on discord. The tl;dr is that the timer for the koopa to wake up decreases twice faster if you're not holding it (every 2 frames versus every 4 frames), until 16 frames before it starts shaking. At that point (16 frames before the shake), it decreases every frame for any scenario. So if you want to optimize the slow setup, you want to grab the shell just before it starts shaking (or when it starts shaking) and then quickly jump into the pipe.

RNG in w1: ~+1.5s RNG in 8-tank: +undetermined (maybe around 0.5s compared to best RNG? but kind of a wild guess, never timed that precisely) mistake in 8-2: ~+1s

As far as strats go, humanly viable is often hard to define for a game like smb3:

In world 1: there's a faster route (excluding 1-2 slide) with a leaf grab in 1-3 that can save up to 0.4s (but is way WAY harder to pull off or save any time with). Also as mentioned 1-2 can be done faster than pspeed with slide jumping and perfect + lucky inputs in every direction (the TAS strat) but gl with that RTA. I used that strat as a backup when I was not getting pspeed and almost never saved time with it.

In world 8: 8-1 and 8-f in my PB are definitely good enough for what I was going for but they can both be done a bit tighter, 8-1 in particular. The fort was very good, just didn't use the treadmill pspeed strat (which pretty much only saves time if you keep pspeed flawlessly through the door, and not as much as people think).

In BC it's theoretically possible to get pspeed in the statue room even with the TAS route, but good fucking luck with that RTA. Also, Bowser can be one cycled but nobody's ever pulled that off in any run of any category (been done in practice but you need a specific RNG, you need some really REALLY nice fireballs, and you need to commit to it). Basically BC in my PB is pretty much as good as it's gonna get (can always save a bit of time on Bowser with an even better kill obv but mine was good).

All in all, with current strats, a high 10:41 would probably be the holy grail (but to be clear, if the two games can even be compared, that would be way beyond getting a high 4:55 time in SMB1 in terms of difficulty). Like you'd need perfect RNG to begin with and to give you an idea, in my entire no WW grind, I've never seen 4 movements of 1 in w1 and 0 hand at the same time (so i'm not even including the boom bro RNG on top of that). Not. A. Single. Time.

Meanwhile if we did a theory TAS with absolutely everything that is "humanly possible" (which is a very very loose concept here), I suspect it would be around 10:37 or something ridiculous like that.

And to throw a last number: during my grind, the best possible time I was expecting to get was a mid 10:42. And that would have been the fluke of the decade already.

So in your run, it takes approx 10s for the shell to start moving. In his run, approx 6s. Interestingly, I also compared with another run (bjw's run if you want to check) and it's approx 8s in it.

Now to be perfectly clear, that's something that is completely new to me (and probably every other player), but after watching those three examples, I think I figured out what caused it: you grab the shell earlier than the two other players after hitting it, almost instantly. bjw grabbed it after a bit, and aldyn grabbed it super late (almost too late).

So it would seem that the timer for the koopa to wake up goes faster if you're not holding the shell??? I'm gonna post that in discord and see if someone can test that theory.

Just to make sure, you're playing on original console with an original cart? If so I've had that same issue before. I think it has to do with some memory values being manipulated by previous attempts and that don't dump on reset (when you soft reset the console, it doesn't properly reset all the values). First thing you can try is to hard reset the console (turn it off and on). Some people say to wait a bit after turning it off but I don't know how accurate that is.

That fixed the problem when I had it (and in general when you fail getting the wrong warp, I recommend always hard resetting the console before the next attempt because of stuff like that). If the problem persists (we've seen one case like that in the past I think), you could try playing another game first (for a bit, any game, the idea is to scramble the values if possible) then put SMB3 again and see if it's fixed.

Edit: I should mention that when the pipe doesn't bring you all the way like that, it has nothing to do with the rest of the setup. The pipe glitch should bring you all the way regardless of the rest of the gameplay (like you can try doing the pipe glitch right as you enter 7-1 and it should bring you all the way down).

@Supernickid Basically the idea is that we use original releases on the NES to define what the norm/ruleset for the categories are. What it means is that anything that wouldn't be possible on NES in terms of glitches and all is banned. For 100%, warpless and any% no WW, versions like VC are perfectly fine because the differences don't actually affect what goes on in these runs to an extent that makes a difference. For any%, as you know and mentioned, some VC exclusive setups are possible. So now to answer your question and emphasize the point: VC isn't banned because some faster setups are possible (some might exist but we don't really know right now), just that it's working too differently for this category. Faster or slower isn't really relevant, it's just too different for the purpose of any%.

(I think that was probably clear to you already but in case you or someone else was still wondering)

It's not strictly required but it's certainly highly prefered. In general we accept runs like that and ask the runners to get sound for the next submissions. Depending on some situations they might get rejected though (someone having no sound for a run very high in the LB out of the blue or something equally shady).

If you want to organize something SMB3 related, I'd recommend you join the actual SMB3 discord instead: https://discord.gg/FbnsB . Lots of players are there and you'll reach/discuss with potential entrants way more easily.

As long as we can see the whole gameplay well enough and without interruptions, it's fine yes (with sound if possible).

I just sent a message to becored myself, linking to this thread.

Hey so I received messages from garadas regarding beco's runs having been reuploaded on Youtube:

garadas21: hey there! i found two of beco's gimmick run on youtube garadas21: both WRs garadas21: i saw that you're one of the mods garadas21: could you add them both? tnx! garadas21: any%: garadas21: best ending:

Imo the records should definitely be added to the LB again but that being said, seeing as the runs are on someone else's youtube (the guy is not taking credit for the runs to be clear, he just reuploaded them unlisted so they're not lost), it would probably be a nice idea to send a message to becored to see if he wants to reupload them himself on one of his accounts, whether it's twitch or youtube, and submit them again himself.

RNG but if you build pspeed properly, there is not a single pattern that will kill you (even while holding right). The jumps are tight though (both the jump to build pspeed and the jump in front of the bro if he walked towards you).

Usually the first course of action should be to find a reference frame for the start & end that can be easily pinpointed without sound.

For example in SMB3, the first frame Mario is visible in the overworld is frame #258 (at the NES framerate). So we pinpoint the first frame of the run by substracting 258 frames from that one. The ending has a clear visual cue in the same way.

Thing is in the case of Strider, it's not really gonna be possible for mods to do that sort of work without the source video to begin with. You're not gonna be able to pinpoint a frame and see a frame counter from twitch directly. Pinpointing is possible on Youtube if the video has the correct framerate but framecounting is still gonna be annoying. The reason Super Mario Bros. 1 mods are able to framecount any run 5:02 or better is that you can know the exact amount of frames in the run based on Bowser patterns & a couple other things. But that really specific approach can't be applied to many games out there.

So to get back to Strider, once a method to find the start & end frames accurately has been found, the best course of action would probably be to ask the runners to do the framecount themselves with their local recording. Which is one more reason to stick to the runs at the very top of the LB for the framecount, because asking runners at the middle of the LB to worry about that would end up being a logistical nightmare and would deter runners too possibly. If someone cares enough and is good enough to get a top 3 time, they should be able to framecount easily given that a proper method has been found.