Amazingly the Low% record got untied again, to 29 seconds.
- The new record skipped the next second barrier, missing 30 completely somehow
- The runner @Sharkking1515 starts way further up the stairs (gains visibility) way later than any other runner of the category thus far.
- The game freezes erratically before the run for a long time, then again at 0:31, which is really odd.
- They don't leave the house as soon as possible, when the owner comes back, they wait around in the house for a second or two. Despite these factors, it's still an enormous record and I have no idea how these factors didn't reduce the time. I don't think it's intentionally cheated but it's surely not a valid game state, if you retime the run from first visibility, manually, you get a time of 31.103 seconds (if we're being really generous).
There's just no way the real length of this run is 29 seconds, although I don't understand how the game calculated such a time
the first time i watch the run i thought the run was a fake run or something. the only part that is weird for me is the alt tab at 0:31 and maybe the game not responding at the start of the run and i don't know how the game calculate the time but yeah 0:29 is a weird time I verified it so ppl can see it and tell me what they think because for me it's a really weird run
Yeah sorry I should explain that point about calculating the game time a bit better, go to this tool https://slashinfty.github.io/yt-frame-timer/ Then follow the instructions with the first frame being the first frame with a visible game, and the last frame being the first frame after they have escaped.
If you do the same thing with other runs, the resultant time is generally the same when rounded to the second, or very close.
I think a message was deleted here but I was going to respond with: This is indeed probably the result of a poorer computer, which for some reason the in-game time is greatly affected by. The solutions to this problem are to reject runs that have stutters or severe buffering before the run starts, or more practically, reject runs that don't align their IGT with the RTA. Because we could overload our computers to get more unrealistically lower and lower times in theory with this, and that's not acceptable Hopefully @omelettejoyeuse sees this
yeah good point there and yeah I think he deleted his post. I think i'm gonna reject the run so thing gonna get back to normal
I'm afraid I'm returning to this thread to bring up IGT again;
The (now former) WR by @Collencyte has an IGT of 26 seconds, no worries, the RTA from when the player was able to move around looks to be (roughly) 28 seconds, a 2 second disparity.
The new WR in all items by @Dario18j7 has a disparity of about 8 seconds, between the rough RTA and the IGT. This is obviously a huge difference elapsing but I'm not sure what the best course of action is - this is a leaderboard for a fun little, heavy RNG, indie game, but it feels like some of the best times have been achieved dubiously.
To make it clear, I'm not saying anyone has done anything wrong, just about the opposite, these are fully legitimate times within the rules but it's a confusing thing. I can't make a side-by-side right now but if anyone wanted to do so it might help understand the issue.
I was very hesitant to upload the run or not when I noticed the disparity in time. I knew beforehand that there was other similar case in this game. I don't really know why this happens, this is not the first time I upload a run of this game (I've uploaded 9, can see them all on my YouTube channel). I'm not sure to what extent it is correct to consider the run as true, but I'm glad to have climbed to the top 1 after a year of speedrunning this game.
A possible solution could be to stop using the time the game tells you when you reach the door (maybe sometimes wrong), and start using methods used in other games.
No worries about uploading the time, we're glad people are interested in doing that haha. Just such a shame the IGT coninually proves to be so poor. I do wonder if an improvement on the IGT might be possible with autosplitting, via LiveSplit for instance, but it'd make the leaderboard less accessible to new runners...
I don't have any problem with it being for fun, but right now it seems you can have a WR in a hugely contested category, with a time of 25 seconds, whilst being 6 seconds (or more than 20%) slower. Surely you'd want to at least find out why this is happening?
maybe the IGT is tied to hardware or something, i mean someone could decomp the game with some unity decomp tool to see i guess
That would be a great step on trying to understand the issue. I might apply 'Occam's Razor' and try emailing Redefinition Games via their info email on the website, it's been years since any post information so I doubt I'll have much success but it's worth a go. Either of us or someone else can try decomping the game too.
