TAS review:
The use of rotations and mix of both super fast and perilous play made this TAS a fun watch. I particularly enjoyed the blatant disobedience of the "Copy!" prompt in the "
Sync or Sink" microgames.
Once again, I've compared it to the current WR
like I did for Mega Microgames!. This one uses
an updated script and accounts for regional differences (the intro screens after each soft reset are 52 frames faster in the Japanese version).
For brevity, I've only listed non-trivial segments, excluding menus and segments that are within margin of error (<0.050 s):
Notable time *saved* (-):
Kat & Ana, games: -0.353 (TAS: 1:38.262, WR: 1:38.616)
Kat & Ana, boss: -5.937 (TAS: 1:16.145, WR: 1:22.083)
Dr. Crygor, boss: -3.031 (TAS: 33.468, WR: 36.500)
WarioWatch, games: -4.082 (TAS: 1:35.500, WR: 1:39.583)
Notable time *lost* (+):
Wario, games: 0.428 (TAS: 39.378, WR: 38.950)
Jimmy, boss: 2.164 (TAS: 6.847, WR: 4.683)
Dribble, boss: 0.881 (TAS: 32.715, WR: 31.833)
[Total]: -21.369 (= -24.924 +3.555) (TAS: 27:19.730, WR: 27:41.100)
menu: -11.377
games: -4.024
boss: -5.966
First the positives: the TAS is definitely faster than the WR. Besides faster menu navigation, the time saved is from:
- A much more efficient defeat of Kat & Ana's boss.
- More aggressive and risky play than the WR in Dr. Crygor's boss.
- Sprinting through WarioWatch to take advantage of the only "real" non-boss time save
- (the WR accidentally pausing briefly at the start of Kat & Ana's microgames)
The TAS also has three time losses compared to the WR:
- Jimmy's boss missed the fast cycle, but there's a good reason for that: it was probably impossible in the emulator version this TAS used. For reference, Mikewillplays' post above suggested a tilt value of 31500, which doesn't work in BizHawk 2.9.1 but does in BizHawk 2.11.
- In Dribble's boss, the initial high-speed back-and-forth movement seemed to be too fast and made the initial acceleration slower.
- The Wario games time loss is specifically from the sixth microgame "Step On It"; this microgame is apparently not fixed-time, and the TAS had bad layout RNG that led to a few frames slower finish, which was then magnified because the ending is cycle-based, where time can (mostly?) only be saved in multiples of ~25 frames.
I made some (unoptimized?) movies demonstrating the improvements:
Jimmy's boss (saves
00:02.160),
Dribble's boss (saves
00:00.871),
Step On It (saves
00:00.385)
Nonetheless, in light of the challenges presented by the gyroscope issues, I strongly believe that this movie is acceptable for publication, fulfilling the
current site standards of being "as optimized as [the author] can make it".
Requiring the movie to be redone as an acceptance requirement would be disrespectful of the author's time; this movie, while imperfect, is solidly constructed.