Frank's Red Hots

[Dann Corbit]

Frank made a suggestion to look for mates that would be good study candidates in his contest.
So far, most of the positions seem to be such that the key move is obvious to strong programs or there are many winning moves.
However, it is quite a large collection so I did not finish analyzing yet. Anyway, here is the current collection and you can have some fun with it:

[MOBMAT]

Would you like us to report results here?

[Dann Corbit]

Would you like us to report results here?

I have my 3970x analyzing right now, so if you want to analyze it, do it for fun.
Quite frankly, there are some beautiful mates in there and some are hard to find.
I have been watching the analysis for the last hour or so.
Currently, I am using SF with depth=40 for a first pass.

[MikeB]

first one I looked at

picked this one bm was Q takes with a check
[d]1k1q1n1R/1p2Q3/3pnP2/pP1Nr3/P3P1B1/8/2r5/3R3K w - - bm Qxd8+; ce 32744;

I got mate in 12 with Bxe6 in less than second ...

info depth 35 seldepth 26 multipv 1 score mate 12 nodes 21829010 nps 106482975 tbhits 28135 time 205 pv g4e6 e5h5 h1g1 c2c8 e6c8 d8e7 f6e7 h5h8 e7e8q f8g6 e8g6 h8c8 g6d6 b8a7 b5b6 a7a6 d5c7 c8c7 b6c7 b7b6 c7c8q a6a7 d6c7

found mate in 11 later
info depth 117 seldepth 22 multipv 1 score mate 11 nodes 45581131925 nps 122591220 hashfull 1000 tbhits 9137443 time 371814 pv g4e6 e5h5 h8h5 f8e6 e7e6 c2c5 f6f7 d8c8 e6c8 c5c8 d5e7 c8d8 b5b6 d6d5 h5d5 d8h8 h1g1 b8a8 d5d8 h8d8 d1d8

this was picked because it was a night promotion

[d]r2q1r1b/2pb1pPk/n2p4/1p1Pp2P/2P1P3/pPN1BP2/P6Q/1K4RR w - - bm gxf8=N+

info depth 81 seldepth 22 multipv 1 score mate 11 nodes 4207372544 nps 98841180 hashfull 846 tbhits 0 time 42567 pv g7f8n d8f8 h5h6 d7g4 g1g4 h8f6 h2h5 f6d8 g4g7 h7h8 h5f5 f8g8 g7g8 h8g8 h6h7 g8h8 e3g5 d8g5 f5g5 b5b4 g5f6

later on,Qg3 was found to be shorter
info depth 108 seldepth 20 multipv 1 score mate 10 nodes 62057297270 nps 118963019 hashfull 1000 tbhits 0 time 521652 pv h2g3 d8f6 h5h6 d7g4 g3g4 f6g6 g4h4 b5c4 g1g6 f7f6 g7f8q a8f8 g6g7 h8g7 h4g4 g7h8 h1g1 a6b4 g4f5

this one had the lowest CE
[d]3q1k2/r5p1/b1P1p1R1/p3N2Q/1p1P4/1P5P/2B2r1K/6R1 w - - bm R6g2; ce 32700;
info depth 82 seldepth 28 multipv 1 score mate 14 nodes 9998382053 nps 108076596 hashfull 990 tbhits 10547533 time 92512 pv g6g2 f2g2 h2g2 d8d5 g2h2 d5d4 g1d1 d4f2 h2h1 a6d3 c2d3 f8e7 e5g6 e7f7 g6f4 f7f6 h5g6 f6e5 g6e4 e5d6 e4e6 d6c5 d1c1 f2c2 c1c2 c5b6 f4d5

[MOBMAT]

Would you like us to report results here?

I have my 3970x analyzing right now, so if you want to analyze it, do it for fun.
Quite frankly, there are some beautiful mates in there and some are hard to find.
I have been watching the analysis for the last hour or so.
Currently, I am using SF with depth=40 for a first pass.

How about for multiple best mates? are you checking for those? I'm just looking for a niche way to help out.

[Dann Corbit]

My first run is single pv, and I really appreciate your efforts.
Your enthusiasm for computer chess is admirable

[Frank Quisinsky]

Hi Dann,

I hope you will find a lot of good material.

Interesting storry to collect such a strong test-database!
But it's many work if you will try to make it perfect!

Best
Frank

[MOBMAT]

My first run is single pv, and I really appreciate your efforts.
Your enthusiasm for computer chess is admirable

I'm taking a break from programming chess engines, but NNUE has renewed my interest.
I'll look for multi-PV solutions.
Are you a fan of hot sauce?

[Dann Corbit]

My first run is single pv, and I really appreciate your efforts.
Your enthusiasm for computer chess is admirable

I'm taking a break from programming chess engines, but NNUE has renewed my interest.
I'll look for multi-PV solutions.
Are you a fan of hot sauce?

Frank's Red Hot is one of my favorites.

[Dann Corbit]

The EPD file with analysis and bare EPD with bm:


I am not sure if any will make it as test problems. Some are hard to drive to mate but finding a winning move seems easy.