So... if you've been keeping up with Appel speedrunning, you'll certainly have seen @DapperScratcher's 3'51"533"' run. It's an insane run that rivals even @MichaelChan's in skill!

However, after a bit of digging, a few people, myself included, have realized that the timing is a bit off with it, at least compared to @MichaelChan's run.

First, I'll give what any good argument requires - hard facts. Let's take a look at the splits of the two runs, not counting load times.

@MichaelChan - 18.9 + 27.2 + 13.8 + 25.3 + 15.4 + 31.7 + 36.6 + 35.4 + 15.4 = 219.7 seconds.

@DapperScratcher - 18.9 + 27.5 + 13.8 + 25.3 + 15.5 + 31.4 + 36.9 + 34.9 + 15.3 = 219.5 seconds.

Notice something off here? On the leaderboards, @DapperScratcher's run is an entire 1.366 seconds slower than @MichaelChan's run, even though her splits are clearly 0.2 seconds faster.

If her time was only slightly slower than @MichaelChan's, this could be attributed to slight lag, and I wouldn't push the issue any further. However, this is a way too much for me to just let it slide.

Honestly, I'm not here to point fingers at anyone. @MichaelChan's cheating scandal was already resolved long ago, and I'm not looking to open up that rabbit hole again. All I know is that @DapperScratcher deserves**** to be at the top of the leaderboard.

Therefore, I propose a new timing system. One that will remove the disadvantages of lag and make sure that every run is timed fairly, based on their splits and nothing else.

Instead of timing the runs with vidtimer.com or any other website, we could simply add up each speedrunner's splits as well as their load times. Courtesy of @MichaelChan's own video,* we know that each load time is 1.62 seconds. If we multiply that by eight to get 12.96 seconds, and add that to the splits of the two top speedrunners...

@MichaelChan - 219.7 + 12.96 = 232.66 seconds.

@DapperScratcher - 219.5 + 12.96 = 232.46 seconds.

Now, let's convert that to minutes using simple math.

@MichaelChan - 3'52"66"'

@DapperScratcher - 3'52"46"'

So, there you have it. Yes, this means everyone's runs will be slowed down. However, keep in mind everyone means everyone, so everything will still remain equal between the speedrunners, while being much more fair, especially to speedruners like @Swskrei, who has a huge disadvantage because of lag.**

Mods, tell me what you think! If you don't have the time to retime everything, I'd be happy to.

*@MichaelChan's video, if anyone was wondering:

**Speaking of, @Swskrei, could you put up your run using something other than clipchamp? It seems the link for it has expired.

Sorry, I forgot clipchamp removed shared videos.

This is another piece of evidence I found: Although the timer matches the ingame timer at 18.9, it's 0.2 behind by level 2. From the beginning to the end of level 2, Michael's timer takes 27.02, while the ingame timer is obviously not that.

Similarly, level 9 takes about 15.5 seconds. The amount of time passed in Michaels timer is 15.2 seconds. There is a slight discrepancy here...

Interpret it as you want, but after level 1 the two timers don't match.

Hmm... interesting. Not sure exactly what that means, but it's certainly something to keep in mind.

Keep in mind - 12.96 isn't easily achieved either. It depends on how fast you go through the menu. Even if every individual level being fully optimised, if you don't go through the menu transitioning fast enough, it's still going to be slower than some runs.

With that stated, just another thing to help people get faster times - exterior autoclickers could slow this down due to lag, it's best to either do it manually or package it.

And if the timer, for some reason, livesplits isn't trustworthy, the weirdest proving method is by 'timing the timer'. With that said, we all need to acknowledge that livesplits timing method is janky and has two modes - in-game and real-time, which is also what happens to be a speedrun.com selection thing. That's another thing, hope this helps (and maybe helps future speedrunners achieve faster times)

Oh and, yeah I don't care about that number one spot anymore after it's there. All I'm saying is if the truth is really so that they're faster... yeah.

Well, @DapperScratcher's run has perfect load times, so there's nothing to worry about there. I'm not sure if she used an autoclicker, we can ask her later, I suppose.

And... livesplits really isn't the worry here, since neither run is timed using that. The mods use vidtimer.com to make sure there's no discrepancy between the timers. The only problem is that there still seems an issue with the timing, somehow :(

Well, with the splits alone, @DapperScratcher's time is certainly faster; there's no question about that.

I definitely agree that load times are needed for a perfect run. Only, they aren't really a measure of skill rather than how good your computer or autoclicker is.

As to what I was saying earlier, less time passes in Michaels timer than the ingame timer. I understand that Appel's timer is far from perfect, but if its used for level runs I don't see why it should be used for FG%. (as in added up with load times) If we really wanted to time our runs perfectly we would use an out of game timer for every category.

At this point getting the world record means getting extremely lucky with your timer. That isn't something thats fair or beneficial for the community.

I agree with the new timing method. Doing the sum off all the split times and then adding the 12.96 seconds that are spent in the loading screen. But i don't know if everyone would agree to it.

And with the topic about DapperScratcher's and Michael's runs.

I always thinked that Michael's runs are all kinda benefited by the lack of lag. Because Michael Chan has a very good computer (i suppose) his run have a lot of less lag.

Definitely putting him in advantage to other runners.

For the moment i will not say anything more until everybody says their opinion. To be able to make a more precise opinion on this topic.

Oh, and about, autoclickers. I tested them to see the result and it doesn't result in perfect load times. Some of the clicks aren't registered so it mostly slower by around 2 frames than a perfect loadtime. So probably DapperScratcher hasn't used an autoclicker.

And the last. But not least fact that i know. Is that Scratch games actually run at around 31 FPS instead of 30FPS. But the timer is coded to be at 30FPS. So, the reason why out-game timers always display a little less than the ingame timer is probably that.

For example, you can look that, in every run using Livesplit, when you touch the flag in level 1. The timer puts around 18.73 and around that. Instead of 18.9.

I just mention this for giving more knowledge about the topic of the ingame / outgame timer.

Another majour issue is that the timing ends when you touch the flag but you can touch a flag and the level not end (this is due to flag hitboxes and player collectible hitboxes)

Those problems aren't as important because they'll only slow you down by a few frames.

For load times, not everyone gets perfect load times. Using 12.96 as the standard is an easy way to get a good time if somebody is really bad at load times. I don't use an autoclicker; I use my touchpad and press down until the next level loads. It'll be pretty hard to calculate load times if a run makes a mistake...

Agreed. Honestly, I always thought it was unfair if someone has a worse computer or a worse mouse, they have an inherent disadvantage. Load times aren't a measure of skill.

@Luis_sm, maybe we should have some sort of community vote, or at the very least, some sort of vote between the mods to decide if we should put the new timing system into place. You say you "don't know if everyone would agree." Well, this is an easy solution to that.

Unfortunately, the majority of the mods have nearly left the community... @sushisrx desired me to be a mod (Probably not the best choice, considering my computer) but there aren't enough mods online to do that anyways!

Another thing I was considering is if the player should pause between each run, perhaps to verify that nothing... suspicious was happening. I still think people should attempt the current version of FG%, just in case we decide to change the system again.

@Sir_Farcelot yeah i think we could do a community vote. But i will wait until more mods find out this forum and say their opinions.

The only problem, is that, Kanuburgetz, Chiken_11 and sushisrx aren't very online.

And, the biggest problem of this, is that, as @Swskrei mentioned, most of the mods have left the community, and the biggest porblem is, THAT WE CAN'T EVEN ADD ANY MODS!

So, as a current moderator, only Super-Moderators can add more mods, and the Super-Moderators are Chiken_11 and sushisrx. Which left the community. So, we need to contact with them to tell their opinion about this / Add more mods.

True, we should wait until this becomes a bigger topic. Do you know any ways to contact them other than speedrun.com?

Ah, that is quite unfortunate. @sushisrx wanted to give his modship to @swskrei... I suppose when he or @Chiken_11 comes back, we can ask them to add more mods who are more active.

@sushisrx was the person who suggested having me be a mod, but for some reason he didn't do it. @Chiken_11 is actually online today! I was about to say a few months, but I apologize for assuming you (if you respond) weren't online.

@Luis_sm, is it possible to actually make a NEW Appel speedrun page to fix @Swskrei's moderator suggestion? I think that he is perfectly capable of being a moderator as I've seen in his own speedrunninng page and believe that he's poses all the virtues needed for the job. Also, if @kanuburgetz, @Chiken_11, and @sushisrx aren't always online, you can add them in as moderators along with the newer ones. The runs might be a problem but you could ask all the users to add their older runs to the newer version. Another idea to this problem is to contact the main mods and ask if there can be more, to fix the speedrunning dilemma.

Anyways, thanks for reading