Still not as fast as xQc versus MoistCr1tikal.
Objectives
- Uses easiest difficulty
- Genre: Board
Difficulty Choice
Skill Level 1 was chosen for a variety of reasons. First and foremost is because the lower-right side button on the controller is the "Force Move" button, which forces the computer to make whatever move it has so far deemed best. The button only works once the computer has found at least one "suitable" move to make; the first "suitable" move is the one which would be selected by the computer on Skill Level 1. Therefore, there is no point in playing on a higher skill level when you could just press the "Force Move" button as soon as it is allowed. On Skill Level 7, the computer will not move at all unless the "Force Move" button is pressed. I could still select Skill Level 7 just for the fun of it but that would lose time to menuing and I did not want that.
Secondly, say there's a movie which completes an entire game on Skill Level 6 and does not allow pressing the "Force Move" button. It would be massively boring. The fastest checkmate I've found with Stockfish 12 makes the computer think for 7 hours, 12 minutes, and 54 seconds over the course of 30 painstaking moves. See the discussion for more info on this. In other words, Skill Level 1 is chosen not only because the computer thinks much, MUCH faster but it also makes lousier moves as I'll soon get into.
Routing
The goal is not necessarily to checkmate in the least number of moves but to deliver the checkmate that makes the computer "think" the least. For instance, the computer thinks for a cumulative 1:10 over eight moves in this TAS. I have also found a six-move checkmate but the computer thinks for 1:22 and is therefore not faster. Shorter "thinking" times are achieved by keeping the position simple. Giving the computer few options and not attacking with a lot of pieces simultaneously is a good way to minimize thinking times.
RNG
Moves are NOT determined stochastically. The computer's response to a position at a given skill level will be the same every time.
The Computer's Weakness?
The first thing to test for would be to see if the computer falls for any quick mates. I was surprised by how well the computer was able to defend every permutation I threw at it. Eventually, I found a way to make it crack: the computer values potential pawn promotion much higher than it should. In the case of this TAS, Black values getting a pawn to the second rank more than it values its own king! In the end, we have a game which looks like a longer variation of the Scholar's Mate.
Game Summary and Q/A
| Move | White | Black | Black's Time | Commentary |
|---|---|---|---|---|
| 1 | e4 | e5 | 0:07 | King's Pawn Opening. |
| 2 | Nf3 | d5 | 0:16 | White puts pressure on Black's e5 pawn. Black ignores this and opts for the Elephant Gambit, establishing a pawn duo in the center and attacking White's e4 pawn. |
| 3 | Nxe5 | dxe4 | 0:24 | Both sides capture. |
| 4 | Bc4 | Be6 | 0:38 | White eyes up the f7 square. Black defends this with his light-square bishop!? |
| 5 | Bxe6 | fxe6 | 0:46 | Both sides trade bishops, but Black's pawn structure is now compromised. |
| 6 | d3 | exd3 | 0:56 | White sets up a distraction, offering to improve Black's pawn structure, which Black is eager to capitalize on. |
| 7 | Qf3 | dxc2 | 1:08 | The White Queen posts up on the f3 square which is no longer defended... but Black, who is obsessed over the idea of promoting, takes on c2 and does not notice that it's mate in one! |
| 8 | Qxf7# | 1:10 | GG |
Q: Why waste two moves to first play Bishop c4 and then capture Black's Bishop? Why couldn't you just play d3 on move 4?
A: Responding with literally anything else on move 4 prompts Black to play Knight f6. If Black plays this move, we are not giving checkmate anytime soon. Black's bishop must be captured because it is defending the f7 square, which will later be used for checkmate.
Q: Why couldn't you play Queen h5 on move 6 instead?
A: Black would meet that with Knight h6, defending against checkmate. d3 is key to distracting Black and to lure their e4 pawn away from defense of f3.
Piece Movemements
There are also some meta strats going on with the cursor. A piece or square can be selected even if the cursor is on an infinitesimal protion of the square's right or bottom edge, or the bottom-right corner. This allows us to set all our pieces down quite early, and gives an advantage to picking up pieces if they're moving to the right (see moves 5 and 7; their movement distances are shortened due to how far to the right they are picked up).
The cursor becomes active for the player's next turn from the moment the computer starts moving a piece, even though it cannot be seen yet. This is why the cursor is always in the perfect spot when it becomes visible.
Special Options
The (8) key on the left controller allows the player to start as Black instead of White. I've also tested this pretty extensively and no quick mates can really be found. Besides, by going second, you're giving the computer an extra turn to think for which loses a lot of time.
You can actually adjust the skill level on the fly between the computer's turns. I also tested many variations to see if the computer would in some instances make a more dubious move on level 2 than level 1. In some cases, it actually did! But too much time was lost due to extra thinking time so we stayed on Skill Level 1.
Suggested Screenshot
Any frame after the CHECKMATE! message appears
feos: While a few people in the thread liked this movie, we can't assume that the general audience of the site is going to enjoy a Chess TAS, judging by the fact that there are no Chess TASes in Moons (or Stars for that matter). Submission votes are currently: 54% (6/0/5). And gameplay itself differs a lot from what usually excites people in a speedrun.
As an illustration, we have a Monopoly movie that stands quite well in Moons, while its counterpart branch is questionable in terms of entertainment and seems to only be there for legacy reasons. No such examples for a Chess game so far.
Since this movie doesn't represent Moons content, its only option would be going to Vault, but the Vault rules ban board games. Here's a thread where the reasons for that are discussed at length.
Maybe the rule wording isn't quite perfect, but its meaning is that videogame adaptations of real-world board games are banned, not anything with traits resembling board games mechanics. For example, we have published collections of minigames to Vault, even though there's a board-like overworld in them. There were some board games that were accepted to Vault by mistake and then we had to move them to Moons to avoid confusion, because Vault is meant to be simple, clear, strict, and serve as a reliable source of information on what we consider a legitimate speed record. We don't hold movies in there if they break Vault tier rules, and they don't set a precedent if they get there by accident.
Maybe someday in the future rules will change, and then due unrejections will happen, but for the time being, I'm rejecting this movie.