I also noticed that Hayabusa's Any% (NOOB) WR has frame drops as well.

Before the reset:

• Frame 1529 is skipped

After the reset:

• Frame 469 is skipped
• Frame 472 is skipped
• Frame 479 is skipped
• Frame 1071 is skipped
• Frame 1077 is skipped

This category is much longer and less optimized, and this run has frame delays that are undoing the frame drops, so I don't know if these drops had a significant overall impact on the time. I only checked up to frame 1200 for this run so there's probably more drops in there.

Howdy,

I noticed that the runner RyuHayabusa89 managed to set a new WR for the Any% time, which was surprising because there is no apparent way to save even a single frame's worth of time in that category—it is fully optimized so far as I can tell.

So, I framestepped through Hayabusa's entire Any% run and found that there are dropped frames.

Mods, you can confirm this for yourself using the FCEUX framecounter visible in the bottom left:

• Frame 399 goes right to frame 401, skipping frame 400
• Frame 405 goes right to frame 407, skipping frame 406

There are also the following frames skipped, but these also have a single frame of delay, so likely no advantage is gained from these drops (from what I can tell):

• Frame 401 goes right to frame 403, skipping frame 402
• Frame 407 goes right to frame 409, skipping frame 408
• Frame 410 goes right to frame 412, skipping frame 411
• Frame 419 goes right to frame 421, skipping frame 420

Some of these skips have an additional frame of delay between animation, but some of them do not.

This is not the first time I've seen a run performed on an old version of FCEUX that resulted in an impossibly fast time. We ran into this same issue with The Great Gatsby due to FCEUX being inaccurate and had to ban old versions of FCEUX as a result.

Accounting for the 33ms gained by frames 400 and 406 being skipped entirely would make the run have a time of 10s 049ms but there is more likely than not another frame skip in there that I cannot find.

I believe RyuHayabusa89's time should be updated to be 10s 067ms as those frame drops make it difficult to accurately determine the true time, and there does not appear to be any possible way to go faster than the 10s 067ms time.

That does sound good regarding the RNG. If it's persistently random even without a hard reset then I see no reason to enforce a reset if the player dies and goes back to the main title menu.

I do agree that having to watch the intro over and over really sucks, but without including the reset how would we know a save state wasn't used?

In theory a player could load a save state right before a death, die, have the game go back to the menu, start a new run, and have the RNG be impacted, such as the positioning of the bats and slorbs and the direction they are moving.

I don't know if it would be possible to lock in the RNG that way, but I also don't know that it's not possible. It's only certainly impossible if the reset is included.

At the very least I think we should do some tests to confirm, to the best of our abilities, that RNG will not be locked in without a reset. My concern is that RNG is decided on initial boot which would lock in the RNG after a death, which would allow a player to keep playing until they find favorable RNG after a death.

I think there should be separate categories for NOOB and OOB rather than combining them into one. The routes are different enough to justify separate categories, in my opinion, and personally I'd prefer to not be doing OOB tricks.

Having separate categories will also help future-proof should there be any other OOB exploits discovered in the later stages.

Fyi I've been tinkering with Low% and I'm confident that it's not possible.

I did manage to get through Train with only 2 coins and 1 drink but there was nothing I could do to void them. There's also the 3 coins in the City at the Hydrant Boost which would be unavoidable.

We could maybe have a category for minimal pickups rather than zero pickups, if there's any interest. From what I can see the game should be beatable with around 5-7 coins and 1-3 drinks. Could be a fun category.

If kuma wants to make an official, proper one we could always just move over to that.

I went ahead and made a Discord server: https://discord.gg/9puD9sMJ

This certainly doesn't need to be an official one or anything, but for now we'll have a place for us to post our different strats for each other to learn from.

If it were just that initial loading of the title screen then, yeah, there would be no fairness issue, but it could still affect RNG in ways we don't expect.

It's best to avoid save states entirely with games that have RNG.

Bit late here, but I can see three potential problems with allowing save states:

The first would be compatibility. Save states used on a desktop emulator are no the same as EverDrive N8 save states. In the interest of fairness multiple types of save states would need to somehow be made with identical timing.

The second issue is verifying on what exact frame the save state begins. A player could adjust the save state to be a few frames ahead which would give them an advantage. How would we check for this?

The third is RNG. The crabs at the end of the game are RNG and the use of save states may lock in the crab RNG such that it isn't different every time the way it is meant to be.

I agree with zarc0nis and kuma: we should start all runs with a reset. No save states.

If somebody is willing to test other versions to confirm they do not provide an advantage that sounds great.

But without confirming they don't I believe it's best to assume they do given we've confirmed that one out-dated version does not lag correctly. There isn't a compelling reason to be using an out-dated emulator and with a game this optimized these slight inaccuracies can easily make the difference between 2nd place and a new WR.

If 2.5.0 does not provide an advantage then it would be safe to say all versions that come afterward are also viable emulators for accurate/legit runs.

Please see my thread regarding this: https://www.speedrun.com/the_great_gatsby/thread/l0cid/1#mhb53

The time muten_pizza achieved is not legitimate due to an inaccurate emulator.

After investigation it is imperative that the emualtor FCEUX version 2.2.3 be banned, and quite realistically all versions of FCEUX other than the latest as well, which is the 2.6.2 version.

The time submitted by muten_pizza had an exceptionally fast Stage 1 framecounted to 54.067 which is faster than all three of the Stage 1 times that are currently podium.

We looked into it and after having muten_pizza switch to mesenRTA and perform the identical route the timesave was immediately lost and he was achieving the expected times that the rest of us also get. The version of FCEUX that was being used, for whatever reason, is not accurate enough to lag correctly and will give the player an unfair advantage.

For the reason outlined above I and zarc0nis both strongly believe we should only allow FCEUX version 2.6.2 to be used.

zarc's a good Dodge Ball mod so I'll throw my hat in in support of that, and should the game continue to explode in popularity—I doubt it, honestly—I'm happy to help out with mod shenanigans as well.

And I will say that, for the most part, if I see the split timer at the end of the run reads one way but the submission is a lower time in milliseconds that I assume you all have done the frame count yourselves. I am more than happy to trust you all with your own frame counts

I appreciate the trust, and we're all homies and wouldn't try to cheat, but we're all human first and we make mistakes. I've accidentally submitted bad runs to Dodge Ball that needed to be rejected because of mistakes, and I've mistimed things before.

For example: That happened to Deadguy. He had a 4:10.920 run but accidentally submitted a 4:10.092 which was accepted. Deadguy would never try to cheat but mods should still doublecheck the numbers.

It's extremely important for mods to framecount every single submission to ensure the claimed timing is in fact accurate.

That glitch saves time on the final screen by allowing you to jump into the green light without having to slow down to land on the pillars. The top runs by me, Thee_Deadguy, and CrabbyCrab101 all use this strat to save time there if you want to see it performed.

Otherwise, so far as I can tell, it isn't useful because going down there will make it more difficult to kill enemies which will cause the game to lag. A significant portion of speedrunning this game is lag reduction which is predominately done by killing enemies.

We have that in Super Dodge Ball and I really love it so I'm down for this.