mmt Vs. Ovyron (G4 D5 BG2)

[drewdrew]

I didn't keep it, it was saved to the engine's learn file so from the distance the engine would know to prune the line and speed up the solving of ...Ke6, and that was it.

It's an anecdote, I don't really care if people believe me or not because I know it happened

Well, there certainly are very simple mate problems that Stockfish still can't handle in reasonable time. For example, Stockfish-dev still struggles with this oldie, which I attribute to Bernhard Bauer from this forum:

[d]rk6/p1r3p1/P3B1Kp/1p2B3/8/8/8/8 w - - 0 1

info depth 51 seldepth 70 multipv 1 score cp -32 nodes 2713568835 nps 2570122 hashfull 1000 tbhits 0 time 1055813 pv e6d5 b8c8 d5a8 c7c4 e5g7 c4a4 a8b7 c8d7 g6h6 b5b4 h6g5 b4b3 b7f3 a4a6 g5f4 d7d6 g7d4 a6a2 f4e3 a7a5 e3d3 a5a4 d3c4 a2c2 c4b4 b3b2 d4b2 c2b2 b4a4 b2b6 f3e2 d6c6 e2d1 c6d5 a4a3 d5c4 d1e2 c4c3 a3a4 b6b4 a4a5 b4b8 e2b5 c3d4 b5e2 d4c5 a5a4 b8b7 e2f3 b7b4 a4a5 b4b1 a5a4 c5d4 f3e2 d4e5 e2d3 b1d1 d3e2 d1e1

With a higher number of threads, SF appears to find this reasonably easily though the depth at which it first finds it seems to vary quite a bit from run to run:
info depth 47 seldepth 17 multipv 1 score mate 5 nodes 984081754 nps 469953082 hashfull 22 tbhits 17322204 time 2094 pv e6d7 b5b4 g6f7 b4b3 f7e8 h6h5 e8d8 g7g6 e5c7

I'm intentionally using default settings. In particular, one thread (deterministic) search. Still can't find mate-in-five ...

info depth 60 seldepth 79 multipv 1 score cp -32 nodes 10948636149 nps 2661592 hashfull 1000 tbhits 0 time 4113566 pv e6d5 b8c8 d5a8 c7c4 e5g7 b5b4 a8d5 c4c5 d5f3 c5a5 g6h6 c8d7 h6g6 a5a6 g6g5 a6a2 g5f4 d7d6 g7f8 d6c7 f8c5 b4b3 c5d4 a7a5 f4e3 c7d6 e3d3 a2h2 d3c4 h2h4 f3d1 a5a4 c4c3 b3b2 d1c2 a4a3 d4g7 d6d5 c3b3 h4h1 g7b2 a3b2 b3b2 d5c4 c2g6 h1h2 b2a3 h2f2 g6e8 c4c3 a3a4 f2e2 e8g6 e2e6 g6f7 e6b6 a4a5 b6b8 f7e6 c3d4 a5a4 b8b6 e6f5 b6b2 a4a5 b2b8 f5h3 d4e3 a5a4 e3f2 h3e6

You're using like 200 threads?

More like 400, yes.

[zullil]

I didn't keep it, it was saved to the engine's learn file so from the distance the engine would know to prune the line and speed up the solving of ...Ke6, and that was it.

It's an anecdote, I don't really care if people believe me or not because I know it happened

Well, there certainly are very simple mate problems that Stockfish still can't handle in reasonable time. For example, Stockfish-dev still struggles with this oldie, which I attribute to Bernhard Bauer from this forum:

[d]rk6/p1r3p1/P3B1Kp/1p2B3/8/8/8/8 w - - 0 1

info depth 51 seldepth 70 multipv 1 score cp -32 nodes 2713568835 nps 2570122 hashfull 1000 tbhits 0 time 1055813 pv e6d5 b8c8 d5a8 c7c4 e5g7 c4a4 a8b7 c8d7 g6h6 b5b4 h6g5 b4b3 b7f3 a4a6 g5f4 d7d6 g7d4 a6a2 f4e3 a7a5 e3d3 a5a4 d3c4 a2c2 c4b4 b3b2 d4b2 c2b2 b4a4 b2b6 f3e2 d6c6 e2d1 c6d5 a4a3 d5c4 d1e2 c4c3 a3a4 b6b4 a4a5 b4b8 e2b5 c3d4 b5e2 d4c5 a5a4 b8b7 e2f3 b7b4 a4a5 b4b1 a5a4 c5d4 f3e2 d4e5 e2d3 b1d1 d3e2 d1e1

With a higher number of threads, SF appears to find this reasonably easily though the depth at which it first finds it seems to vary quite a bit from run to run:
info depth 47 seldepth 17 multipv 1 score mate 5 nodes 984081754 nps 469953082 hashfull 22 tbhits 17322204 time 2094 pv e6d7 b5b4 g6f7 b4b3 f7e8 h6h5 e8d8 g7g6 e5c7

I'm intentionally using default settings. In particular, one thread (deterministic) search. Still can't find mate-in-five ...

info depth 60 seldepth 79 multipv 1 score cp -32 nodes 10948636149 nps 2661592 hashfull 1000 tbhits 0 time 4113566 pv e6d5 b8c8 d5a8 c7c4 e5g7 b5b4 a8d5 c4c5 d5f3 c5a5 g6h6 c8d7 h6g6 a5a6 g6g5 a6a2 g5f4 d7d6 g7f8 d6c7 f8c5 b4b3 c5d4 a7a5 f4e3 c7d6 e3d3 a2h2 d3c4 h2h4 f3d1 a5a4 c4c3 b3b2 d1c2 a4a3 d4g7 d6d5 c3b3 h4h1 g7b2 a3b2 b3b2 d5c4 c2g6 h1h2 b2a3 h2f2 g6e8 c4c3 a3a4 f2e2 e8g6 e2e6 g6f7 e6b6 a4a5 b6b8 f7e6 c3d4 a5a4 b8b6 e6f5 b6b2 a4a5 b2b8 f5h3 d4e3 a5a4 e3f2 h3e6

You're using like 200 threads?

More like 400, yes.

Seems like overkill to solve mate-in-5!

[Uri Blass]

I didn't keep it, it was saved to the engine's learn file so from the distance the engine would know to prune the line and speed up the solving of ...Ke6, and that was it.

It's an anecdote, I don't really care if people believe me or not because I know it happened

Well, there certainly are very simple mate problems that Stockfish still can't handle in reasonable time. For example, Stockfish-dev still struggles with this oldie, which I attribute to Bernhard Bauer from this forum:

[d]rk6/p1r3p1/P3B1Kp/1p2B3/8/8/8/8 w - - 0 1

info depth 51 seldepth 70 multipv 1 score cp -32 nodes 2713568835 nps 2570122 hashfull 1000 tbhits 0 time 1055813 pv e6d5 b8c8 d5a8 c7c4 e5g7 c4a4 a8b7 c8d7 g6h6 b5b4 h6g5 b4b3 b7f3 a4a6 g5f4 d7d6 g7d4 a6a2 f4e3 a7a5 e3d3 a5a4 d3c4 a2c2 c4b4 b3b2 d4b2 c2b2 b4a4 b2b6 f3e2 d6c6 e2d1 c6d5 a4a3 d5c4 d1e2 c4c3 a3a4 b6b4 a4a5 b4b8 e2b5 c3d4 b5e2 d4c5 a5a4 b8b7 e2f3 b7b4 a4a5 b4b1 a5a4 c5d4 f3e2 d4e5 e2d3 b1d1 d3e2 d1e1

With a higher number of threads, SF appears to find this reasonably easily though the depth at which it first finds it seems to vary quite a bit from run to run:
info depth 47 seldepth 17 multipv 1 score mate 5 nodes 984081754 nps 469953082 hashfull 22 tbhits 17322204 time 2094 pv e6d7 b5b4 g6f7 b4b3 f7e8 h6h5 e8d8 g7g6 e5c7

I'm intentionally using default settings. In particular, one thread (deterministic) search. Still can't find mate-in-five ...

info depth 60 seldepth 79 multipv 1 score cp -32 nodes 10948636149 nps 2661592 hashfull 1000 tbhits 0 time 4113566 pv e6d5 b8c8 d5a8 c7c4 e5g7 b5b4 a8d5 c4c5 d5f3 c5a5 g6h6 c8d7 h6g6 a5a6 g6g5 a6a2 g5f4 d7d6 g7f8 d6c7 f8c5 b4b3 c5d4 a7a5 f4e3 c7d6 e3d3 a2h2 d3c4 h2h4 f3d1 a5a4 c4c3 b3b2 d1c2 a4a3 d4g7 d6d5 c3b3 h4h1 g7b2 a3b2 b3b2 d5c4 c2g6 h1h2 b2a3 h2f2 g6e8 c4c3 a3a4 f2e2 e8g6 e2e6 g6f7 e6b6 a4a5 b6b8 f7e6 c3d4 a5a4 b8b6 e6f5 b6b2 a4a5 b2b8 f5h3 d4e3 a5a4 e3f2 h3e6

You're using like 200 threads?

No need for more than 1 thread to solve this mate in 5 very fast.

FEN: rk6/p1r3p1/P3B1Kp/1p2B3/8/8/8/8 w - - 0 1

AnMon 5.75:
1+ 00:00 2 6 -5.55 1.Bxc7+ Kxc7
1 00:00 8 24 -4.97 1.Bxc7+ Kxc7 2.Kxg7
2 00:00 35 107 -4.97 1.Bxc7+ Kxc7 2.Kxg7
2+ 00:00 141 412 -4.72 1.Bd5 h5 2.Kxh5
2 00:00 165 482 -2.55 1.Bd5 h5 2.Kxh5
3 00:00 222 649 -6.37 1.Bd5 b4 2.Bxa8 Kxa8 3.Bxc7
3+ 00:00 225 628 -5.08 1.Bxc7+ Kxc7 2.Kxg7
3 00:00 246 657 -4.97 1.Bxc7+ Kxc7 2.Kxg7
4 00:00 534 1k -5.13 1.Bxc7+ Kxc7 2.Kxg7 h5
4+ 00:00 604 2k -5.01 1.Bd5 h5 2.Kxh5 Kc8
4 00:00 982 2k -1.99 1.Bd5 Kc8 2.Bxa8 Kd8
5 00:00 2k 4k -0.76 1.Bd5 Kc8 2.Bxc7 Kxc7 3.Bxa8
6 00:00 2k 6k -0.69 1.Bd5 h5 2.Bxc7+ Kxc7 3.Bxa8 b4
7 00:00 5k 12k -0.33 1.Bd5 Kc8 2.Bxc7 Kxc7 3.Bxa8 Kb8 4.Bb7
8 00:00 10k 23k -0.01 1.Bd5 b4 2.Bxa8 Kxa8 3.Bxc7 h5 4.Kxg7 b3
9 00:00 28k 590k -0.01 1.Bd5 b4 2.Bxa8 Kxa8 3.Bxc7 h5 4.Kxg7 b3 5.Kf6
10 00:00 74k 785k -0.01 1.Bd5 b4 2.Bxa8 Kxa8 3.Bxc7 h5 4.Kxh5 b3 5.Be5 b2 6.Bxb2
11 00:00 139k 889k -0.01 1.Bd5 b4 2.Bxa8 Kxa8 3.Bxc7 h5 4.Kxh5 b3 5.Be5 b2 6.Bxb2 Kb8
12 00:00 766k 1,258k -0.01 1.Bd5 b4 2.Bxa8 Kxa8 3.Bxc7 h5 4.Kxh5 b3 5.Be5 b2 6.Bxb2 Kb8 7.Bxg7
13 00:01 1,757k 1,389k -0.01 1.Bd5 b4 2.Bxa8 Kxa8 3.Bxc7 h5 4.Kxh5 b3 5.Be5 b2 6.Bxb2 Kb8 7.Bxg7 Ka8
14 00:03 5,864k 1,532k -0.01 1.Bd5 b4 2.Bxa8 Kc8 3.Bxc7 Kxc7 4.Kxg7 b3 5.Kxh6 Kb6 6.Be4 Kxa6 7.Kg7 b2 8.Bb1
15 00:08 13,764k 1,548k -0.01 1.Bd5 b4 2.Bxa8 Kc8 3.Bxc7 Kxc7 4.Kxg7 b3 5.Kxh6 Kb6 6.Be4 Kxa6 7.Kg7 b2 8.Bb1 Kb6
15+ 00:09 13,994k 1,547k +M7 1.Bd7 b4 2.Kf7 h5 3.Ke7 b3 4.Bf4 h4 5.Be5 b2 6.Kd8 b1Q 7.Bxc7+
15+ 00:09 13,996k 1,547k +M6 1.Bd7 b4 2.Kf7 h5 3.Ke7 b3 4.Bf4 h4 5.Be5 b2 6.Kd8 b1Q 7.Bxc7+
15+ 00:09 13,996k 1,547k +M5 1.Bd7 b4 2.Kf7 h5 3.Ke7 b3 4.Bf4 h4 5.Be5 b2 6.Kd8 b1Q 7.Bxc7+
15 00:09 13,997k 1,547k +M5 1.Bd7 b4 2.Kf7 h5 3.Ke7 b3 4.Bf4 h4 5.Be5 b2 6.Kd8 b1Q 7.Bxc7+

FEN: rk6/p1r3p1/P3B1Kp/1p2B3/8/8/8/8 w - - 0 1

Movei00_8_438:
1 00:00 48 786 -3.56 1.Bxc7+
2 00:00 66 857 -3.56 1.Bxc7+ Kxc7 2.Kxg7
2 00:00 75 974 -3.55 1.Bd7
2 00:00 79 1k -3.26 1.Bd7
2 00:00 85 1k -3.56 1.Bxc7+ Kxc7 2.Kxg7
2 00:00 92 989 -3.55 1.Kh5
2 00:00 95 1k -3.56 1.Bxc7+ Kxc7 2.Kxg7
2 00:00 103 1k -3.55 1.Bd5
2 00:00 112 1k -3.26 1.Bd5
2 00:00 127 1k +0.28 1.Bd5 b4 2.Bxc7+ Kxc7 3.Bxa8
2 00:00 197 2k +0.28 1.Bd5 b4 2.Bxc7+ Kxc7 3.Bxa8
3 00:00 238 238k +0.28 1.Bd5 b4 2.Bxc7+ Kxc7 3.Bxa8
3 00:00 529 529k +0.28 1.Bd5 b4 2.Bxc7+ Kxc7 3.Bxa8
4 00:00 669 335k +0.49 1.Bd5 Kc8 2.Bxc7 Kxc7 3.Bxa8
4 00:00 1k 640k +0.49 1.Bd5 Kc8 2.Bxc7 Kxc7 3.Bxa8
5 00:00 2k 106k +0.44 1.Bd5 b4 2.Bxa8 b3 3.Bd5
5 00:00 3k 154k +0.44 1.Bd5 b4 2.Bxa8 b3 3.Bd5
6 00:00 4k 185k +0.54 1.Bd5 Kc8 2.Bxc7 Kxc7 3.Bxa8 Kb6
6 00:00 5k 259k +0.54 1.Bd5 Kc8 2.Bxc7 Kxc7 3.Bxa8 Kb6
7 00:00 7k 286k +0.73 1.Bd5 Kc8 2.Bxc7 Kxc7 3.Bxa8 Kb6 4.Bb7
7 00:00 10k 353k +0.73 1.Bd5 Kc8 2.Bxc7 Kxc7 3.Bxa8 Kb6 4.Bb7
8 00:00 13k 382k +0.66 1.Bd5 Kc8 2.Bxc7 Kxc7 3.Bxa8 b4 4.Kxg7 b3 5.Bd5
8 00:00 21k 491k +0.66 1.Bd5 Kc8 2.Bxc7 Kxc7 3.Bxa8 b4 4.Kxg7 b3 5.Bd5
9 00:00 29k 507k +0.36 1.Bd5
9 00:00 46k 539k +0.25 1.Bd5 Kc8 2.Bxc7 Kxc7 3.Bxa8 Kb6 4.Bb7 b4 5.Kf5
9 00:00 61k 605k +0.25 1.Bd5 Kc8 2.Bxc7 Kxc7 3.Bxa8 Kb6 4.Bb7 b4 5.Kf5
10 00:00 86k 501k +0.24 1.Bd5 Kc8 2.Bxc7 Kxc7 3.Bxa8 b4 4.Bd5 Kb6 5.Bc4 Kc5 6.Be6
10 00:00 130k 577k +0.24 1.Bd5 Kc8 2.Bxc7 Kxc7 3.Bxa8 b4 4.Bd5 Kb6 5.Bc4 Kc5 6.Be6
11 00:00 177k 605k +0.14 1.Bd5 Kc8 2.Bxc7 Kxc7 3.Bxa8 b4 4.Bd5 Kb6 5.Bc4 Kc5 6.Bb3 Kb5
11 00:00 268k 676k +0.14 1.Bd5 Kc8 2.Bxc7 Kxc7 3.Bxa8 b4 4.Bd5 Kb6 5.Bc4 Kc5 6.Bb3 Kb5
12 00:00 408k 659k -0.14 1.Bd5 Kc8 2.Bxc7 Kxc7 3.Bxa8 Kb6 4.Bb7 b4
12 00:00 612k 749k -0.14 1.Bd5 Kc8 2.Bxc7 Kxc7 3.Bxa8 Kb6 4.Bb7 b4
13 00:01 1,202k 735k -0.44 1.Bd5
13 00:02 1,586k 769k -0.23 1.Bd5 Kc8 2.Bxa8 Re7 3.Kf5 g6+ 4.Kf6 Kd7 5.Bd5 g5 6.Bb8 Kd8 7.Be5
13 00:02 1,928k 814k -0.22 1.Bd7
13 00:02 1,928k 814k +0.07 1.Bd7
13 00:02 1,941k 818k +M5 1.Bd7 b4 2.Kf7 b3 3.Ke7 b2 4.Kd8 b1Q 5.Bxc7+
13 00:02 1,975k 825k +M5 1.Bd7 b4 2.Kf7 b3 3.Ke7 b2 4.Kd8 b1Q 5.Bxc7+

[Ovyron]

Indeed non-Stockfish engines have proven useful for Zenmastur's challenge, specially those that show moves with steady progress up to mate and their alternative moves mate slower

Progress update: I have solved ...Kg6 to a mate score now working on ...Qf6 instead of ...Qf5.

[Uri Blass]

Indeed non-Stockfish engines have proven useful for Zenmastur's challenge, specially those that show moves with steady progress up to mate and their alternative moves mate slower

Progress update: I have solved ...Kg6 to a mate score now working on ...Qf6 instead of ...Qf5.

The question is if stockfish is the best engine to mate faster.

I would like to see mate rating list for practical mates when starting positions may be positions from human games when one player resigned and engines with rating above 2500 start from them and play 2 games.

if engine A can force mate in 18 moves and engine B can force mate in 16 moves then engine B get 2 points.
If the result is not 1-1 then the conclusion is that the position is a bad position or one of the engine has a serious bug

In case 1 you do not use the position in the test.
In case 2 you say that the engine has a bug and you do not use it because engine should be able to win a position that humans resign assuming that they did not resign too early.

[Ovyron]

I don't see what's the appeal to mate faster? Best moves take care of winning instead of drawing or drawing instead of losing, but on a decided game it's not clear to me why would you want to mate ASAP.

Slow mates make no difference and this was my main complaint about the challenge. Some person could be really good at finding the fastest mates, but really bad at reaching positions were the opponent is likely to blunder, or bad at avoiding blunders themselves.

[Dann Corbit]

I don't see what's the appeal to mate faster? Best moves take care of winning instead of drawing or drawing instead of losing, but on a decided game it's not clear to me why would you want to mate ASAP.

Slow mates make no difference and this was my main complaint about the challenge. Some person could be really good at finding the fastest mates, but really bad at reaching positions were the opponent is likely to blunder, or bad at avoiding blunders themselves.

If you have a mate in 4 and a mate in 40, most of the time, the mate in 4 isn't just faster, it is also much safer. You have 36 less opportunities for a faux pas.

On the other hand, if you sacrifice your queen for a mate in 10 and there is a grinding mate in 30 without giving up material, then we have a problem. The sacrifice mate is probably a lot more beautiful. But there are a lot of places to go wrong, especially for a human.

So I think it is never an easy decision which is best, without looking at the actual problem carefully.

[Ovyron]

If you have a mate in 4 and a mate in 40, most of the time, the mate in 4 isn't just faster, it is also much safer. You have 36 less opportunities for a faux pas.

Okay, I can see the appeal for mating quickly in traditional chess, specially in blitz where I've had a mate in 4 but not enough time to play it on the board, so I lose.

But this is correspondence chess, with 0% chance of a faux pas.

Under these circumstances I believe Zappa Mexico II* at Depth 15 is enough to play a mate 100% of the time against any defense for Zenmastur's position and then any sane person would just go and do that which takes minutes instead of trying to find the fastest mate.

*) Zappa Mexico II has extraordinary hash management and with just 128MB ram it's capable of remembering important variations and backsolving them to the root without Stockfish's "I just forgot how white was mating" nonsense.

[jp]

The question is if stockfish is the best engine to mate faster.

I would like to see mate rating list for practical mates when starting positions may be positions from human games when one player resigned and engines with rating above 2500 start from them and play 2 games.

Do other engines make (supposedly proven) mate announcements, though?

[Ovyron]

Yes, nothing is special about Stockfish regarding mate announcements. Unless there's a bug, mate announcements are maximal (opponent can't delay it further, when announced it's real).