- AirXonix is an old game where you trap balls and fill the needed amount of area. The goal is to fill all levels.
- In certain levels I have tested multiple paths to find the best outcome. I manipulated the hit detection for the fastest ball trap, and in 1 part of the tas, cleaner did not destroy my path somehow, so I just thought to leave it there. I guess if the ball is close to a wall, it doesn't have an immediate hit detection so you can basically go trough the ball. I was aiming to finish all the difficulties under half an hour, I did that successfully.
- This movie is required, it is installation of AirXonix --> https://tasvideos.org/UserFiles/Info/638556810044004211
eien86: Claiming for judging.
feos: Trimmed the end of the movie. It contained empty inputs as well as OS shutdown. We need the movie to stop at some kind of an ending, letting the music play.
eien86: AirXonix is yet another Qix-inspired game where the player needs to surround marbles that keep bouncing around. The game also contains enemies whose purpose is thwarting the player's goal, and powerups meant to improve the player's abilities. This movie completes all 'chapters' in order of difficulty.
I believe the goal of '100%', as specified by the author, does not apply here since it refers the number of levels completed, and not any other metric (powerups collected, enemies killed, etc). Since the completion of all levels is to be expected by any solution of this game (I don't think an Doom-style episode-per-episode solution makes much sense here), this should be deemed simply any%.
Then there is the question of optimality:
- There is no analogous run in speedrun-dot-com to compare this against, as the most players only compete for the best solution on Easy mode. The sum of best solutions for Easy, Classic, and Modern adds up already to a larger time than this movie, which also includes Hard and Extreme, so one can infers this is much faster than the RTA solutions out there.
- There was a discussion in the submission's topic about optimality. Arguably the author themself has admitted to this being suboptimal or 'not good'.
- Others have pointed out this looks like superhuman play
In my opinion, this movie looks flawless, in the literal meaning of the word. It lacks visible mistakes or egregious slowdowns and reaches a consistency far beyond what we can expect from human play. That notwithstanding, it seems clear this solution still has a lot of room for improvement. However, the path forward isn't clear: a lot of research work only on pathing needs to be done. I don't think botting would even help here (the exploration space is astronomical).
I enjoyed watching this movie and believe the wider community will too. Accepting to Standard
Thanks for Randomno for the independent sync confirmation and encode. I based my judgement based on that. Thanks also feos for verifying sync and trimming the end of the input.
feos: Processing...