Posts for wuschelbeutel1

Posted: 8/6/2018 5:14 PM
Quote
wuschelbeutel1
Experienced Forum User
Joined: 9/22/2009
Posts: 18
Wyster wrote:
So yesterday was the reveal of Whiteted's highly anticipated "M-Strat", and as part of this he asked me to make a TAS demonstration run: Link to video Final time is 1:35.70 which beats the previous record by 8 seconds.
That's fantastic! How much more do you think you can squeeze out after optimizing? I think Whiteted mentioned that 1:30:xx is possible with some other strats (which I'm not familiar with). I haven't played Goldeneye since my childhood. Does the "large/hut key" unlock (i) the fenced hut where the safe with the bunker plans is located or (ii) the hut where the safe key is located?
Posted: 2/27/2012 6:25 PM
Quote
wuschelbeutel1
Experienced Forum User
Joined: 9/22/2009
Posts: 18
heres a link where you dont have to register: http://www.nicozon.net/watch/sm6438206 what surprises me is that the other (pseudo-)obsoleted run is not even on the snes movie list on tasvideos....why? also, somebody who speaks japanese should contact this person and get him to publish his run on tasvideos....
Posted: 1/16/2012 11:02 PM
Quote
wuschelbeutel1
Experienced Forum User
Joined: 9/22/2009
Posts: 18
All these unwritten glitch constraints cause cognitive dissonance.
Posted: 1/8/2012 3:34 PM
Quote
wuschelbeutel1
Experienced Forum User
Joined: 9/22/2009
Posts: 18
what a beautiful run. this is like childhood christmas eve all over again. when mario goes back to yoshis island 1, its like a epic queen sac, or like a massive workaround in a tsp that has it fruits later in the path - i love it. thanks to everyone involved for the huge effort that goes into making this happen and im glad that due credit was given to nathan for discovering that glitch.
Posted: 12/5/2011 3:56 AM
Quote
wuschelbeutel1
Experienced Forum User
Joined: 9/22/2009
Posts: 18
will this new glitch, which allows the game to end really soon, use input controls exclusively?
Posted: 11/13/2011 10:33 PM
Quote
wuschelbeutel1
Experienced Forum User
Joined: 9/22/2009
Posts: 18
the one that makes non-shelled koopas appear?
Posted: 11/13/2011 2:44 AM
Quote
wuschelbeutel1
Experienced Forum User
Joined: 9/22/2009
Posts: 18
the new run will beat the previous one by over 20.6s? thats very impressive; i thought there were only a few seconds left to squeeze out at this point
Posted: 5/29/2011 3:57 AM
Quote
wuschelbeutel1
Experienced Forum User
Joined: 9/22/2009
Posts: 18
Good indicators for strong submissions: - High rerecord count per time unit (can obviously be distorted by a weak TASer or bogus rerecords) - Subsequent improvements show an approximately exponentially diminishing trend (as is the case with the popular Mario titles). This is analogous to the obj. function graph over runtime in most mathematical program solvers. - Community popularity (high profile titles attract more ideas and attempts by more people) - Engine exploits are exhausted - Low game complexity, low inputs and low run time. 3D titles have more possible input combinations, some games are too simple (like rail shooters or minigame-type games), and in a >1h run there probability of undiscovered glitches, imperfections etc. are really high. I think the second point is the most valid and is heavily correlated with the third point.
Posted: 5/28/2011 6:36 PM
Post subject: The N64 TAS closest to perfection
Quote
wuschelbeutel1
Experienced Forum User
Joined: 9/22/2009
Posts: 18
In which N64 game do you think the optimality gap of the currently published TAS is the tightest? (Relative to the optimal time value not in absolute time) Mario 64 comes to mind, but what are some other candidates?
Posted: 2/16/2011 12:54 AM
Quote
wuschelbeutel1
Experienced Forum User
Joined: 9/22/2009
Posts: 18
Why not submit the run until cradle (to keep your time consistent/comparable with the previous attempts) and post the last two levels for entertainment?
Posted: 1/22/2011 4:58 AM
Quote
wuschelbeutel1
Experienced Forum User
Joined: 9/22/2009
Posts: 18
I was very happy to see new progress. Excellent job
Posted: 2/22/2010 5:33 AM
Quote
wuschelbeutel1
Experienced Forum User
Joined: 9/22/2009
Posts: 18
Slowking wrote:
Nothing of the run is done yet. But yes, it will be sub 1h.
Wow, I didn't expect that kind of progress.
Posted: 2/20/2010 5:10 AM
Quote
wuschelbeutel1
Experienced Forum User
Joined: 9/22/2009
Posts: 18
Thus far, how much time have you guys shaved off the last release?
Posted: 2/5/2010 5:02 AM
Quote
wuschelbeutel1
Experienced Forum User
Joined: 9/22/2009
Posts: 18
Nice one!
Posted: 2/1/2010 5:12 AM
Quote
wuschelbeutel1
Experienced Forum User
Joined: 9/22/2009
Posts: 18
Didn't Henrik get 0.52 before already? (Not trying to deflate the run: it is outstanding). How do you guys go about deciding whether it is possible to squeeze more time out of a level or not? Have you already experimented before and know exactly which levels have an improvement potential or do you just decide to break even with previous times to complete the run (with a still better cumulative time b/c of the previous level's time savings)?
Posted: 12/30/2009 11:34 PM
Quote
wuschelbeutel1
Experienced Forum User
Joined: 9/22/2009
Posts: 18
Man, I was so excited about this run... almost halfway done with so much progress and now it's dead. This is just sad
Posted: 12/27/2009 2:58 AM
Quote
wuschelbeutel1
Experienced Forum User
Joined: 9/22/2009
Posts: 18
This probably sounds dumb (I don't know much about CS & math) but why is it not viable to use a brute force backwards induction method. Doesn't the complexity rise linearly that way? Let's say we consider Mario Bros.. We start at the last controllable frame of the game, try every possible input combination (or those that make sense), find the fastest one, then use that and go back one frame - rinse and repeat. That way all the games should be solved relatively quickly...or where do I go wrong?
Posted: 9/22/2009 6:29 AM
Quote
wuschelbeutel1
Experienced Forum User
Joined: 9/22/2009
Posts: 18
Any progress?