yeah thats easy enough, cannot get the turnskip tho.. even when I get air I've already hit the offroad beforehand
#rMC NG 3lap
hazy hemlock
austere gate
can you scroll up and try referencing some of my earliest ghosts? iirc i got it while being further back. or maybe it was before i sent anything in which case i can try to look
it mostly felt like trial and error until i found a drift trajectory that sort of worked most of the time. but it wouldn't surprise me if certain forward positions always resulted in a stick
hazy hemlock
yeah I think thats the case. I just did rta attempts from midway through the drift trying different positions, this is already 14ms behind because of the wider line and still touches the offroad very slightly. pretty sure anything further ahead will touch it a lot more and can't clear the gap
austere gate
this turnskip is way too high but it's a better basis than yours
btw if we're ~0.5 f behind at the tunnel, we don't get airtime at one of the slope transitions. so idk how to possibly make that work without matching the position
hazy hemlock
yeah it's a bit cooked, although Alego got this tunnel (which would make up the .5f) with a relatively different position from the current tunnel.
I say we either put the 961 in the bag and try again in a few months time or put this on hold until someone wants to to do in the near future. I know you've also got exams soon so it's not gonna be us.
austere gate
yeah i have not been handling my time very well lmfao i really should fuck off the face of earth for two weeks
hazy hemlock
I'm for option 1 personally, just wanna get this run outta my head
austere gate
as i said already i am happy with publishing the 961
hazy hemlock
aight I'm gonna encode this in a bit then
against Thomas 10-12f ahead then? I think 6 is a little short
hazy hemlock
worn steeple
honestly no frames ahead
🔥
need everyone to visually see how much this run saved over the previous, on minimap
austere gate
i don't want anyone to watch thomas's run for more than 5 seconds it's too ugly
austere gate
i will publish the video in 6 hours
tardy crest
First rMC ng 3lap to exclusively use wheel inputs
I wonder how many categories are still left that still don’t exclusively use wheel inputs
Obviously every BKT from before December 2020, but definitely several after as well
oblique nova
i will publish the video in 6 hours
Is this TAS close to the theoretical limit it can be pushed to? I don’t see how any more time could be saved tbh
Yall kind of optimized the crap out of it mad props
fast atlas
Well, a lot of ng tases nowadays are close to the theoretical limit at the time they’re posted, but the limit can change in the future
oblique nova
Well, a lot of ng tases nowadays are close to the theoretical limit at the time ...
That’s true, I’m sure some new strat breakthrough will be found in the future but for now it looks pretty dang optimal
worn steeple
couldn't you say that about any category though?
"it's optimal for now but a new strat could be found in 5 years"
oblique nova
I guess so yeah
I’m just saying, this track seems very optimized and I’m impressed, especially with how many spots you guys had to save like a frame to get good QM or whatnot
granite skiff
Is this TAS close to the theoretical limit it can be pushed to? I don’t see how ...
potentially but unsure (and also unwilling)
using a proper wheelie trigger at the end of lap 1 saves ~10ms i think but completely fucks qm and afaik we’re unsure if it can be fixed
oblique nova
stuff like that makes me think this can’t be optimized too further since it’ll mess up things later on in the run. With how you guys have it here it’s like the “best route” where qm and everything goes right, if that makes sense
oblique nova
I personally think each TAS you guys make is a gift, constantly working and finding new strats, I think it’s all more than special if you ask me
austere gate
idk why ejay and foss are being so callous rn
austere gate
every tas that we publish is as good as we think it can be. proving that a section can be optimized is so much easier than proving that a section can't be optimized. nobody can say for sure that some tas is perfected, we've all been wrong many times.
oblique nova
idk why ejay and foss are being so callous rn
It’s fine lol, I just like voicing my appreciation for you guys and the work you do
austere gate
but also every track is being wrung out of its potential very hard in ng. the way i look at it, is all of the runs being made are very strong baselines, and maybe in the future we'll see crazy small optimizations, or some new thing that nobody saw coming. whether that will ever happen, no one can know
worn steeple
what was wrong with what foss said though
austere gate
the new double trick that marth found could've been something that was found a long time from now. we're lucky it happened now instead of later
austere gate
what was wrong with what foss said though
not unwilling. i already know that the qm for the tunnel won't work if we're a few ms behind. one of the bounces required for turning will not happen
tardy crest
define exclusively
There’s no section of the TAS that was made without full input range code (for example, laps 1 and 2 of the previous BKT were made before Thomas rejoined the community and found out about full input range)
worn steeple
There’s no section of the TAS that was made without full input range code (for e...
I only started using wheel inputs in like mid 2023