This TAS improves MESHUGGAH's run https://tasvideos.org/3176M by 67 frames. Input from Power On to the beginning of the Keno number reveal is identical to MESHUGGAH's run, as is the input from the end of the Keno number reveal to the final input. What differs is the input during the Keno number reveal.
As per MESHUGGAH's notes, there's a per-reveal timer at $7010 that can be manipulated with input. The minimum value was reached for 14/20 numbers. More specifically:
30 - 0x1E (0) 40 - 0x27 (9) 50 - 0x1F (1) 10 - 0x20 (2) 70 - 0x1F (1) 28 - 0x1E (0) 80 - 0x1E (0) 12 - 0x1E (0) 60 - 0x1E (0) 63 - 0x1E (0) 24 - 0x20 (2) 29 - 0x1E (0) 44 - 0x1E (0) 32 - 0x1E (0) 15 - 0x1E (0) 19 - 0x1E (0) 79 - 0x1E (0) 20 - 0x1F (1) 78 - 0x1E (0) 13 - 0x1E (0)
That leaves 16 frames that can theoretically still be saved during the reveal. Whether that's technically feasible, I leave to someone with more motivation/brains.
nymx: Claiming for judging.
nymx: I have to say, I was also curious about the question of co-authorship. After reviewing the inputs, I found that they are similar in some places—but not in the areas that matter most. Many games will naturally look similar from TAS to TAS, especially during menuing and prompts. However, for this TAS, the important comparison is where the actual optimization takes place.
Between frames 645 and 1945, the inputs in this submission are dramatically different. Since this section contains the core gameplay and optimization, I don’t see how co-authorship would be necessary here. It is also worth noting that we are comparing FM and BK files, with the inputs shifted one frame earlier in BizHawk; this does not affect the conclusion.
Your approach has clearly made a difference, and I congratulate you on the effort.
Accepting over [3176] NES Vegas Dream by MESHUGGAH in 00:50.28.
fsvgm777: Processing.