So Cuttyflame just beat my time with one full second using the same routes as me (except level 12 which only saved about 0.3 seconds) i had a hard time to understand how this could be 1 second faster in total, im guessing it's because of loading times between levels or potentially mistimed run. So my suggestion is to instead of using RTA use ingame time by adding up all of the times you can see at the level selection screen.
Following this method my 2:35:56 run would be a 2:14:65 IGT and Cuttyflame's 2:34:52 run would be a 2:14:58 IGT which to me seems alot more accurate.
This would eliminate any timing errors or inconsistent load times, also it would make it easier for mods to time the runs as it only requires adding up the times at the end instead of timing the video frame by frame. Only problem is this only works if the run is deathless so this might not work for any% but in my opinion it should be used in 30exit and the other shorter categories where deathless is possible.
A new rule would be required to show all the level times in the level selection screen at the end of the run. (which both me and Cuttyflame are already doing in our runs)
After review your request, It's a good idea. IGT is more fair than RTA for : Deathless, Low%, Deathless and Low% subcategories in 30 and 60 Exits.
I dont mind this method, it is indeed more accurate.
But saying its faster then counting frame is totally wrong, it took me 40 sec to re-time my current run. It took me 30 min to do the document below and 10 more minute to be sure every time was correct
And there is no timing error when you use the frame system compare to doing math I did this document for you and i can whisper you a link to access editing anywhere if you want, but here are my result
https://docs.google.com/spreadsheets/d/19rpmQK6A2muUdrb-eeX14DkycbjMN1SjtKjH_QLMXhU/edit#gid=0
Igt of my run was 2:08.91 and your igt was 2:08.92
I dont understand why you are calculating with the -9. The method that i was talking about was just adding up all the times at the level selection screen. Im not sure how i can share my excel sheet but here is a screenshot so maybe you understand what i mean https://imgur.com/a/PX8T1
Level selection screen was proven to be wrong. Here https://www.speedrun.com/run/mrq2q17y you can see my time end at 6.18 in the menu, it show 2.81 9-6.18=2.82 thats why i use the minus 9
How does that prove that the level selection screen is wrong? It might aswell be the time remaining which is inaccurate
How do we know whats right then? by frame timing the level, i get 3.036 sec while the ingmae showe 2.82 and the level selection got 2.81
That's interesting i can't seem to get below 2.84 after about 5mins of trying, not sure really what would be the best way of timing it. btw how are you frame timing it? do you start the time when the screen starts fading to black or when the level starts fading in?
On the level i posted earlier i start 1 frame before 8.97 so at 34 frames and ended at 125 so 91 frames
And the game act weird when you start a lvl but move at a specific time. if you watch this run frame by frame, the timer start, end and then start back which gave you a headstart. it only work when you start the level from the screen selection though.
i tried making a script that runs level 1 multiple times. Both ways of timing it ( time remaining and level selection screen ) gave me inconsistent times. level selection varied around 2.63 - 2.64 time remaining varied around 6.33 - 6.37 another thing to note was that the script occasionally failed, bascially meaning that the loading time from starting the level is inconsistent which also means that RTA will also be slightly inaccurate... Honestly if all three methods of timing runs are slightly inaccurate i feel like we should use the level selection times since it only requires adding up the times which you can easily see in the video at the end and it suffers the least if you have bad computer performance
but what about run like thenewchallenger where he doesnt show the ending leaderboard (sorry for the late reply)
I timed the run of NewChallenger in 30 Exits (Deathless) with the true timing method for IGT.
https://docs.google.com/spreadsheets/d/1tCXgw2OVyFiAVlx0qKvGPVGC1MOFHkLOp3-o4TIr3qk/