(Link to video)
There is the "3D Pieces, Level 5" branch published a while ago. Since 3D mode is now legitimized on this site, here is the "3D Pieces (any%)" branch.

Goals

Aims for fastest checkmate, subject to the following rules:
  • Must be 1 player vs CPU, starting from a new game (no SETUP BOARD).
  • 3D mode at all times.
Note: Any CPU difficulty is permitted. Start to force CPU to move is permitted. Technically, glitches are also permitted, but none of the known glitches make it any faster.

Wasn't there a previous TAS on this site with this same category?

How did you manipulate the CPU moves? These moves are clearly not what the CPU plays on Novice.

By pressing start to force CPU to move, you can make the CPU play moves that are even worse than those found on Novice. I found out about this trick from speedrun.com (thanks ShesChardcore!) and I simply used the same strategy here. Pressing start at different times can give a different CPU move.

Table of moves and comments

MoveWhiteCommentBlackComment
1e3 h6CPU move when force move is used right away.
2Qf3Faster than 2. Bc4, 3. Qf3, 4. Qxf7#.Rh7Forced right away. Faster than 2... e5.
3Bc4 h5Faster than 3... Rh8 on 3D mode.
4Bxf7# 1-0
Note that, unlike on speedrun.com which ends time when the game plays the checkmate jingle, in this TAS time is ended on last input. Note that the sequence after the end of input is longer than the entire input itself!

nymx: Claiming for judging.

nymx: After a group discussion with our senior judges, I have come to a conclusion on this submission.
  • Optimization is extremely tight. The only frames for saving, are used to manipulate the "forced move" on the CPU's part. As Fractal stated, this is not really RNG...but taking a move that is being tested, which on different frames...can be bad chess moves. I do find it odd though, that pressing "Start" earlier actually shows a "Book Move"....probably going through many iterations to confirm that current move for countering "White's" turn.
  • Audience feedback was light, yet positive...with all votes being "Yes".
  • Movie's branch, "3D Pieces", is acceptable to that of an any% branch. Having passed on any glitches, which failed to save time, this branch would have obsoleted any existing glitched runs.
Accepting as a "3D Pieces" branch, under a "Standard" publication.
Thanks to ikuyo for your review.
Zapp Brannigan: If we hit that bullseye, the rest of the dominoes will fall like a house of cards. Checkmate.

TASVideoAgent
They/Them
Moderator
Joined: 8/3/2004
Posts: 14776
Location: 127.0.0.1
This topic is for the purpose of discussing #7926: FractalFusion's NES Battle Chess "3D Pieces" in 00:46.62
ShesChardcore
She/Her
Skilled player (1502)
Joined: 2/23/2022
Posts: 131
Location: MN
Good work! I'm glad to see this game getting some love lately from you and adelikat. Even more glad that I helped contribute somehow!
GMP
He/Him
Active player, Editor, Reviewer (358)
Joined: 5/22/2020
Posts: 197
Location: Chennai, India
Extra points for checkmating with the bishop, yes vote!
Post subject: FractalFusion vs Battle Chess | Analysis - tasvideos.org
ikuyo
She/Her
Active player, Judge (499)
Joined: 7/8/2021
Posts: 98
Always excited to see your work! Here's a quick review: Sync confirmed for Bizhawk 2.8, using the USA ROM. The RNG manipulation is really funny as someone with a chess background, I seriously never thought I'd see chess AI worse than freaking Martin-bot yet here we are! Move order and manips are performed excellently, with clear emphasis on minimizing input time by optimizing movement. Overall, a really clean movie and an obvious Yes vote from me.
Editor, Skilled player (1936)
Joined: 6/15/2005
Posts: 3239
By the way, it's not so much RNG as it is "force move" (a number of chess engines have this feature); by pressing start, the CPU is forced to play the best move it has currently calculated. Pressing it as soon as possible makes Battle Chess play a (usually bad) move without having been given time to evaluate the position. Obviously, the move that is forced depends on the timing of when you press start. But, in general, the sooner you press start to force move, the worse the resulting CPU move is.
ikuyo
She/Her
Active player, Judge (499)
Joined: 7/8/2021
Posts: 98
FractalFusion wrote:
By the way, it's not so much RNG as it is "force move" (a number of chess engines have this feature); by pressing start, the CPU is forced to play the best move it has currently calculated. Pressing it as soon as possible makes Battle Chess play a (usually bad) move without having been given time to evaluate the position. Obviously, the move that is forced depends on the timing of when you press start. But, in general, the sooner you press start to force move, the worse the resulting CPU move is.
Oh, my bad! Thanks for the clarification.
Post subject: Movie published
TASVideoAgent
They/Them
Moderator
Joined: 8/3/2004
Posts: 14776
Location: 127.0.0.1
This movie has been published. The posts before this message apply to the submission, and posts after this message apply to the published movie. ---- [5158] NES Battle Chess "3D Pieces" by FractalFusion in 00:46.62
Krayzar
He/Him
Joined: 9/25/2020
Posts: 3
Just wanted to make a slight correction here as the discoverer of the "Start Button" gimmick and the creator of the published 2020 Human theory TAS for Battle Chess on SRC that all the modern runs are based off. The Start button behavior isn't documented in the manual or other resources as a force move button. Reading this through makes it sound like it is. Perhaps it's simply intended but undocumented (and there's certainly evidence that it is - the game stops listening for start after a successful "interruption"), but the manual only mentions Start in the context of starting the game. My human theory TAS attempts to display this but there's an unpublished "non-Human" version I have that demonstrates this in full, simply for fun. FractalFusion and I came up with the exact same solution down to the frame, beyond my extra start presses to demonstrate that the game really does stop listening after a certain point and pressing start doesn't introduce any delays. I much prefer how clean the file looks on this one and am pleased folks are having fun with Battle Chess!