https://manifold.markets/harfe/manifold-plays-chess-3-if-we-play-3-86c4fb507db1 bxa5'
https://manifold.markets/harfe/manifold-plays-chess-3-if-we-play-3-9e4ed18ae6ae Rf7

@harfe your random number is: 537237526

Salt: BTVyIVCt4BlujpEM1QA8, round: 2957264 (signature 99e1b843c746adff805d7ceed178a915282d24bdb81ae0b8a5e0a7a0d69ae19806fa5eec34c4318af5245dd487ada7e9190de7c12ebe52fe6c419fd143afb2f8a28bb650e4c37ce33e1c2a39c119217e9fd8428d5a2f5dce111338856d168468)

@FairlyRandom bxa5, Rf7

@harfe you asked for a random integer between 1 and 999999918, inclusive. Coming up shortly!

Source: GitHub, previous round: 2957262 (latest), offset: 2, selected round: 2957264, salt: BTVyIVCt4BlujpEM1QA8.

@FairlyRandom 999999918

Moves by average probability:

0.454323 bxa5

0.206719 Rxc6+

0.046704 b5

0.044430 Rh7

0.044265 Ra7

0.044101 d4

0.044085 Re7+

0.043975 Rd7

0.043422 Rf7

0.006861 Kc3

0.000000 Rb7 (removed)

--------------------

pick a number between 1 and 999999918 (inclusive)

Outcomes by integer range:

[ 1-230364365] bxa5, Rxc6+

[230364366-278796535] bxa5, b5

[278796536-324829926] bxa5, Rh7

[324829927-370689671] bxa5, Ra7

[370689672-416377021] bxa5, d4

[416377022-462047403] bxa5, Re7+

[462047404-507601787] bxa5, Rd7

[507601788-552573865] bxa5, Rf7

[552573866-574497262] bxa5, Rxc6+, b5

[574497263-595301144] bxa5, Rxc6+, Rh7

[595301145-616024132] bxa5, Rxc6+, Ra7

[616024133-636666828] bxa5, Rxc6+, d4

[636666829-657301623] bxa5, Rxc6+, Re7+

[657301624-677882405] bxa5, Rxc6+, Rd7

[677882406-698192178] bxa5, Rxc6+, Rf7

[698192179-715707252] Rxc6+, b5

[715707253-732351059] Rxc6+, Rh7

[732351060-748931811] Rxc6+, Ra7

[748931812-765449964] Rxc6+, d4

[765449965-781961956] Rxc6+, Re7+

[781961956-999999918] other

Hash of the complete table:

9cb0a8d79942721a7e7362fd2b0475b4c006b21a12c07ee1983dffc885a13fb2

silly engines.. so we got our top 3 all are equally fast #11.

@deagol We will probably do better. I’m not that good at chess but I was getting mates in 8 from the previous position against Stockfish level 3.

@TobyBW oh yes these are worst case with perfect play from black

this one’s as fast as Rxc6+, one step behind the top two

oh wait, now that my engine learned of b5 it realized Rxc6+ isn’t a step behind after all, no need to move the rook on next move, 35. b5 does the job

this is as fast as bxa5, but somehow missed by engines 🤷‍♂️

🤪

Added a 3rd good move to hedge against the blunders that are being added lol

@jack found a couple better ones

34 bxa5 Kd6

35 Rh7

involves only 1 rook move before pawn queened.

Seems better than Rxc6+ which involves 2 moves other than advancing pawn to be queened

I think.

@ChristopherRandles no need to move the rook, 35. b5

After 35 b5 I was thinking 35 ... g4 and you can't advance the pawn but if you have consulted engines perhaps this doesn't matter in some way because Rc4 is helpful after pawn has queened. The engines are more likely to be right than me,

@ChristopherRandles Right, I saw that too so there the engine seemed stuck at some depth, a few minutes thinking, and suddenly BAM! 36. d4! it found a mate in #11. It took me a while to understand what was going on, how this different pawn advance could match the other more obvious lines. But yeah definitely an “engine move”

@deagol “mate in 11” counting from before our current move 34, so #8 from the position after:

34. Rxc6+ Kd7

35. b5 g4

36. d4! (#8)

But maybe we should avoid this, much riskier as a blunder on a critical move to kill that advancing g pawn could throw the game.

finally the engines see the light