Rio de Janeiro, BrazilThe glitch itself is allowed since it uses the same idea as similar ones used in runs on the leaderboard, althought it isn't faster since the final time of the run is 15.433 (at 30 FPS) because, as Yummy_Bacon said, time starts on the first frame the play/restart button is pressed.
For the Survive category though it wouldn't be allowed once the game gets frozen and it would be a disadvantage with other runs which were done without glitches, specially because you could do it over and over instead of playing the mini-game itself.
If it gets allowed in runs, it'll definitely make 2-3 Skip easier, in theory
There is a 2-3 Skip where you use a bat to get on top of the level, it saves 10s althought it's really RNG heavy and nobody goes for it, about the Dpad stuff, kinda curious, did you unabled the Left+Right option from your Emulator or even with it desactivated you could do this?
Used Somewes.com to see the time it takes to the screen to turn black after the first input and it showed me 0.733, so if someone wants to start the time when it turns black, just put 0.73 on the LiveSplit or if the timing method changes, put -0.73 on LiveSplit, that's it I think.
All runs must be done in one sitting, and recording must contain everything, even the breaks, but timer doesn't stop since game's timing method is RTA
We could try both things, althought even knowing how much time it takes to turn the screen fully black, we'll still framecount it from the first input given since final time somewes/YFT will give to you is based when you set the first frame, sadly it isn't like a LiveSplit or something around that.
I think the timing method could change since it's based on the first input frame, HOWEVER, it's not everybody who does have an inout displayer for that and so it makes retiming really hard and inconsistent, I'd like to suggest if it could starts on the first frame the screen gets black after selecting a file so it would be more consistent like this.
My idea is too add Loadless (Load Removed Time) as a timing method because I notice that some cellphones has slower loadings than others (don't have values yet because didn't retime them precisely, but this is something to be discuted before being added), and loading differences can be an issue due to the disadvantages between them, but again, before that, retiming science should be done and also see if it's worth it since retiming loadless is really hard and it consumes lots of time.
The other thing is about retiming itself, idk what program has been used by the moderators but I've got some different times from the program/site that I do use (which is somewes.com) so I'd like to know what am i doing wrong so I can fix that for future runs (along with the first/final frames as well)
Last thing is I think we could have more ILs, at least 30 levels since we have a category that goes from 1 to 30, but again, just a suggestion
So recentlty I decided to replay Candy Crush Soda to grind and do some level science/searching and notice that all the levels I said that were different from my version are now the same as the version I was comparing to (which was atlas' one) I'm not sure if the game updated and changed everything, but I'm pretty sure the game probably has a "Demo" version which has all these levels from the leaderboard (aside from Atlas' runs) and after sometime it does change to the "real ones"
I think we could split the leaderboard between "Demo" and "Real version" or something like that to avoid level differences since the Demo seems like to have a bunch of "pre-scrypted" levels where you could do the same moves and get the same results
I plan to install in my other phone and do more science about it + uninstall it and reinstall it and see what's going on with it
The videos ahead were recorded on August 23/24 respectively (description for double check), HOWEVER, the change was notice by me days/weeks before, just did not have time to record and share this information in here
The proof of the level changes:
Moderators are working on runs, so please be patient with the run verification delay.
A new category is not needed but the category itself can be separate between Plants and Zombies subcategories
I think milliseconds will be needed, people will probably try to optimize it
Not required but highly recommended and would be appreciate it very much if it had
His run was really weird indeed, I remember when I was grinding the game again, I had splits with his time to compare agaisnt and It seemed to be impossible to save time, no matter what I did, I did the same strats, tried to optimize as much as I could and I was still behind by a lot, even playing on Fusion (which is sighly fastser than BizHawk and it was at the time I was still running on that Emulator), so I agree that he probably used a different version of the game and that it should be moved.
Because the Shindou Version has been separated and it is on this leaderboard https://www.speedrun.com/sm64ique
For timers you can use your phone, but final times are always retimed to avoid misstiming, but it's good to have a timer to keep track on run's pace as well
