Problems with this position

Kempelen · Post by **Kempelen** » Thu May 28, 2009 6:32 pm

I found this position searching in the forum.

[D]4k3/8/8/8/8/8/4P3/4K3 w

It was posted in a thread about testing transposition table. OK I said, let's test it. What I found is that my engine does not find the solution even TT working fine. Then I give the position to other engines which I know TT work apparently fine. The conclusion is that there were engines which dont found the solution even in 30 depth or even passes 2 minutes:

I.e, Buzz engine:

Code: Select all

31 169 2268 12551036 Kd2 Kd7 Kd3 Kd6 Kd4 Ke6 Ke4 Ke7 Ke5 Kf7 Kd6 Kf6 e4 Kg5 e5 K
f4 e6 Ke3 Kd5 Kd3 Ke5 Kc4 Ke4 Kb3 Kd4 Ka2 Kd3 Ka3 Ke4 Kb3
31 169 2270 12557132 Kd2 Kd7 Kd3 Kd6 Kd4 Ke6 Ke4 Ke7 Ke5 Kf7 Kd6 Kf6 e4 Kg5 e5 K
f4 e6 Ke3 Kd5 Kd3 Ke5 Kc4 Ke4 Kb3 Kd4 Ka2 Kd3 Ka3 Ke4 Kb3

i.e2, Mizar:

Code: Select all

19 238 176 1560515 e1d2 e8e7 d2e3 e7d7 e3f4 d7e7 f4f5 e7f7 e2e4 f7e7 e4e5 e7f7 e
5e6 f7e8 f5f4 e8f8 f4e4 f8e8 e4e5 e8f8
20 238 198 1771848 e1d2 e8e7 d2e3 e7d7 e3f4 d7e7 f4f5 e7f7 e2e4 f7e7 e4e5 e7f8 e
5e6 f8e7 f5e5 e7f8 e5f6 f8e8 f6e5 e8f8

As I see, my search tend to fill with king moves or premature pawn push:

Code: Select all

19    197   1425   8622288 1. Kd2 Kd7 2. e4 Ke6 3. Kd3 Ke7 4. Ke3 Ke8 5. e5 Kf8
6. Kd4 Ke7 7. Kd5 Kd7 8. e6+ Kc7 9. Ke5 Kd8 10. Kf6 Kc7
20    197   2571  15915625 1. Kd2 Kd7 2. e4 Ke6 3. Kd3 Ke7 4. Ke3 Ke8 5. e5 Kf8
6. Kd4 Ke7 7. Kd5 Kd7 8. e6+ Ke7 9. Ke5 Ke8 10. Ke4 Kf8 11. Kd5

Of course, I add pawn passed bonus, pawn position bonus and all those things...

What is exactly needed here to correctly find the solution, a part from a correct TT implementation?

bob · Post by **bob** » Thu May 28, 2009 7:07 pm

Kempelen wrote:I found this position searching in the forum.

[D]4k3/8/8/8/8/8/4P3/4K3 w

It was posted in a thread about testing transposition table. OK I said, let's test it. What I found is that my engine does not find the solution even TT working fine. Then I give the position to other engines which I know TT work apparently fine. The conclusion is that there were engines which dont found the solution even in 30 depth or even passes 2 minutes:

I.e, Buzz engine:
Code: Select all
31 169 2268 12551036 Kd2 Kd7 Kd3 Kd6 Kd4 Ke6 Ke4 Ke7 Ke5 Kf7 Kd6 Kf6 e4 Kg5 e5 K
f4 e6 Ke3 Kd5 Kd3 Ke5 Kc4 Ke4 Kb3 Kd4 Ka2 Kd3 Ka3 Ke4 Kb3
31 169 2270 12557132 Kd2 Kd7 Kd3 Kd6 Kd4 Ke6 Ke4 Ke7 Ke5 Kf7 Kd6 Kf6 e4 Kg5 e5 K
f4 e6 Ke3 Kd5 Kd3 Ke5 Kc4 Ke4 Kb3 Kd4 Ka2 Kd3 Ka3 Ke4 Kb3
i.e2, Mizar:
Code: Select all
19 238 176 1560515 e1d2 e8e7 d2e3 e7d7 e3f4 d7e7 f4f5 e7f7 e2e4 f7e7 e4e5 e7f7 e
5e6 f7e8 f5f4 e8f8 f4e4 f8e8 e4e5 e8f8
20 238 198 1771848 e1d2 e8e7 d2e3 e7d7 e3f4 d7e7 f4f5 e7f7 e2e4 f7e7 e4e5 e7f8 e
5e6 f8e7 f5e5 e7f8 e5f6 f8e8 f6e5 e8f8
As I see, my search tend to fill with king moves or premature pawn push:
Code: Select all
19    197   1425   8622288 1. Kd2 Kd7 2. e4 Ke6 3. Kd3 Ke7 4. Ke3 Ke8 5. e5 Kf8
6. Kd4 Ke7 7. Kd5 Kd7 8. e6+ Kc7 9. Ke5 Kd8 10. Kf6 Kc7
20    197   2571  15915625 1. Kd2 Kd7 2. e4 Ke6 3. Kd3 Ke7 4. Ke3 Ke8 5. e5 Kf8
6. Kd4 Ke7 7. Kd5 Kd7 8. e6+ Ke7 9. Ke5 Ke8 10. Ke4 Kf8 11. Kd5
Of course, I add pawn passed bonus, pawn position bonus and all those things...

What is exactly needed here to correctly find the solution, a part from a correct TT implementation?

You should be able to see a queen very quickly. A simple piece of chess knowledge can help, but it is not necessary.

The simple rule for king and pawn vs king, pawn not a rookpawn, is that if the winning king is 2 squares in front of the pawn, and the pawn can't be taken by the opponent, then the position is won. If the king is one square in front of the pawn, the pawn can't be captured, and the winning king has the opposition, then the position is won.

If any of the above fail, the position is a draw. This position is trivially won since the king can reach a position that meets the above very quickly.

If you can't solve this one in a second or so, you still have some search/transposition table issues even though you think it is working correctly...

Crafty can see the mate here, without endgame tablebases. It takes about 2 minutes to find a mate in 22. EGTB's confirm that this is the shortest mate.

Here's some snapshots:

Code: Select all

              depth   time  score   variation &#40;1&#41;
                1     0.02   1.60   1. e4      &#40;600nps&#41;             
                1->   0.02   1.60   1. e41     &#40;850nps&#41;             
                2     0.02   1.54   1. e4 Kd7  &#40;900nps&#41;             
                2->   0.02   1.54   1. e4 Kd7  &#40;1Knps&#41;             
                3     0.02   1.92   1. e4 Kd7 2. e5ps&#41;             
                3->   0.02   1.92   1. e4 Kd7 2. e5ps&#41;             
                4     0.02   1.54   1. e4 Kd7 2. Kf2 Ke6           
                4     0.02     +1   1. Kf2!                        
                4     0.02     +3   1. Kf2!                        
                4     0.02     +M   1. Kf2!                         
                4     0.02   6.56   1. Kf2 Kd7 2. Kf3 Ke6           
                4->   0.02   6.56   1. Kf2 Kd7 2. Kf3 Ke6     

At this point, Crafty's endgame knowledge has classified this as a "won position" because of the rules I gave.

               27     0.49     +1   1. Kf2!                          
               27     0.51     +3   1. Kf2!                          
               27     0.51     +M   1. Kf2!                          
               27     0.53  12.84   1. Kf2 Kf8 2. Kf3 Kf7 3. Ke4 Ke6 4.
                                    e3 Kd6 5. Kf5 Ke7 6. Ke5 Kd7 7. Kf6
                                    Kd6 8. e4 Kd7 9. e5 Ke8 10. Ke6 Kf8
                                    11. Kd7 Kf7 12. e6+ Kf6 13. e7 Kf5
                                    14. e8=Q Kf4

at depth=27, 0.5 secnds, it sees that the pawn can actually promote to a queen...

               41     3.90     +1   1. Kf2!                          
               41     4.14     +3   1. Kf2!                          
               41     4.28     +M   1. Kf2!                          
               41     2&#58;24  Mat22   1. Kf2 Kd7 2. Kf3 Ke7 3. Ke4 Ke6 4.
                                    e3 Kd6 5. Kf5 Ke7 6. Ke5 Kd7 7. Kf6
                                    Kc6 8. Ke6 Kb5 9. e4 Kc5 10. e5 Kc6
                                    11. Kf7 Kb5 12. e6 Kb4 13. e7 Kc3 14.
                                    e8=Q Kd2 15. Ke6 Ke2 16. Qd8 Ke1 17.
                                    Qd4 Kf1 18. Qd2 Kg1 19. Kf5 <HT>

At depth 41 it fails high after 3.9 seconds, and after a long search, announces the mate in 22...

Jan Brouwer · Post by **Jan Brouwer** » Thu May 28, 2009 7:45 pm

Rotor also finds a mate (albeit mate in 23) in about 2 minutes, and has no special endgame knowledge for this endgame, except for passed pawns:

Code: Select all

info depth 1 seldepth 2 nodes 1 nps 10 score cp 110 hashfull 0 pv e2e3
info depth 1 seldepth 2 nodes 2 nps 20 score cp 132 hashfull 0 pv e2e4
info depth 2 seldepth 3 nodes 12 nps 120 score cp 122 hashfull 0 pv e2e4 e8d7
info depth 3 seldepth 4 nodes 38 nps 380 score cp 134 hashfull 0 pv e2e4 e8d7 e4e5
info depth 4 seldepth 5 nodes 129 nps 1290 score cp 134 hashfull 0 pv e2e4 e8d7 e4e5 d7e6
info depth 5 seldepth 6 nodes 385 nps 3850 score cp 132 hashfull 0 pv e2e4 e8d7 e1f2 d7e6 f2e3
info depth 6 seldepth 7 nodes 768 nps 7680 score cp 132 hashfull 0 pv e2e4 e8d7 e1f2 d7e6 f2e3 e6e5
info depth 7 seldepth 8 nodes 1483 nps 14830 score cp 122 hashfull 0 pv e2e4 e8d7 e1f2 d7e6 f2e3 e6e5 e3d3
info depth 7 seldepth 8 nodes 2311 nps 23110 score cp 132 hashfull 0 pv e1f2 e8d7 f2e3 d7e6 e3f4 e6d5 e2e4 d5d4
info depth 8 seldepth 10 nodes 3269 nps 32690 score cp 142 hashfull 0 pv e1f2 e8d7 f2e3 d7e6 e3f4 e6f6 e2e4 f6e6
info depth 9 seldepth 12 nodes 5282 nps 52820 score cp 164 hashfull 1 pv e1f2 e8d7 f2e3 d7e6 e3d4 e6d6 e2e4 d6e6 e4e5
info depth 10 seldepth 13 nodes 7355 nps 73550 score cp 164 hashfull 1 pv e1f2 e8d7 f2e3 d7e6 e3d4 e6f5 e2e4 f5f4 e4e5 f4f5 d4d5


...


hashfull 24 pv e1f2 e8d7 f2e3 d7e7 e3d4 e7e6 e2e4 e6d6 e4e5 d6e6 d4e4 e6e7 e4f3 e7f7 f3
e3 f7g7 e3d4 g7f8 e5e6 f8e7 d4e5 e7f8 e5d6 f8e8
info depth 24 seldepth 28 nodes 443286 nps 680930 score cp 216 hashfull 25 pv e1f2 e8d7 f2e3 d7e7 e3d4 e7e6 e2e4 e6d6 e4e5 d6e6 d4e4 e6e7 e4f3 e7f7 f3
e3 f7g7 e3d4 g7f8 d4d5 f8e7 e5e6 e7f6 d5d6 f6g7 d6d5
info depth 25 seldepth 28 nodes 556284 nps 584331 score cp 884 hashfull 27 pv e1f2 e8d7 f2e3 d7e7 e3e4 e7e6 e2e3 e6f6 e4d5 f6e7 d5e5 e7d7 e5f6 d7c8 e3
e4 c8d7 e4e5 d7e8 f6e6 e8d8 e6f7 d8d7 e5e6 d7c6 e6e7 c6d5 e7e8q
info depth 26 seldepth 28 nodes 600540 nps 549945 score cp 884 hashfull 27 pv e1f2 e8d7 f2e3 d7e7 e3e4 e7e6 e2e3 e6f6 e4d5 f6e7 d5e5 e7d7 e5f6 d7c8 e3
e4 c8d7 e4e5 d7e8 f6e6 e8d8 e6f7 d8d7 e5e6 d7c6 e6e7 c6d5 e7e8q


...


info depth 40 seldepth 50 nodes 37763716 nps 692124 score cp 984 hashfull 605 pv e1f2 e8d7 f2e3 d7e7 e3e4 e7e6 e2e3 e6f6 e4d5 f6e7 d5e5 e7d7 e5f6 d7c8
 e3e4 c8d7 e4e5 d7e8 f6e6 e8d8 e6f7 d8c7 e5e6 c7c6 e6e7 c6d5 e7e8q d5d4 f7e6 d4e4 e8d7 e4f4 d7d4 f4f3 e6e5 f3g3 d4f4 g3g2 e5d4 g2g1 d4e4 g1g2 f4f3 g2h
2
info depth 41 seldepth 52 nodes 78577300 nps 742851 score mate 23 hashfull 633 pv e1f2 e8d7 f2e3 d7e7 e3e4 e7e6 e2e3 e6f6 e4d5 f6e7 d5e5 e7d7 e5f6 d7c
8 e3e4 c8d7 e4e5 d7e8 f6e6 e8d8 e6f7 d8c7 e5e6 c7c6 e6e7 c6d5 e7e8q d5d4 e8a4 d4c3 f7e6 c3d2 a4d4 d2e2 e6e5 e2f3 d4d3 f3f2 e5f4 f2e1 d3c2 e1f1 f4f3 f1
g1 c2g2
info hashfull 636
info string nodes&#58; 79622253 q-nodes&#58; 38% pv nodes&#58; 15.65% try 10% attacks 49% req capt 20% req check 3%
info string illegal ic&#58; 2% illegal oc 6% fail-high&#58; 96% hash&#58; 97% pawn hash&#58; 100% king eval cuts&#58; 32% null cuts&#58; 9%
bestmove e1f2

Kempelen · Post by **Kempelen** » Thu May 28, 2009 8:00 pm

bob wrote: If you can't solve this one in a second or so, you still have some search/transposition table issues even though you think it is working correctly...

Do you think it is possible a search inestability in this kind of possitions?

bob · Post by **bob** » Thu May 28, 2009 8:02 pm

Kempelen wrote:
bob wrote: If you can't solve this one in a second or so, you still have some search/transposition table issues even though you think it is working correctly...
Do you think it is possible a search inestability in this kind of possitions?

I do not see how. You should not be doing any null-move searching. Extensions are almost non-existent until you get a queen on the board. Any material capture turns this into a dead draw. So I am not sure where you might encounter any of the usual instability issues...

P. Villanueva · Post by **P. Villanueva** » Thu May 28, 2009 8:45 pm

I've tried your position with KMT Chess (under 2000 ELO)

Code: Select all

ply  score    nodes    time    bf  principal variation
  1    145        7     0cs     *  e2e4
  2    135       24     0cs     *  e2e4 e8d7
  3    154       89     1cs  3.57  e2e4 e8d7 e4e5
  4    144      208     2cs  2.94  e2e4 e8d7 e4e5 d7e6
  5    145      476     2cs  2.31  e2e4 e8d7 e1d2 d7e6 d2d3
  6    140      988     3cs  2.18  e2e4 e8d7 e1d2 d7e6 d2d3 e6e5
  7    140     1871     4cs  1.98  e2e4 e8d7 e1d2 d7e6 d2d3 e6e5 d3e3
  8    150     3860     5cs  1.98  e1d2 e8d7 d2d3 d7e6 d3d4 e6d6 e2e4 d6e6
  9    169     7001     6cs  1.93  e1d2 e8d7 d2d3 d7e6 d3d4 e6d6 e2e4 d6e6
                                   e4e5
 10    169     9445     8cs  1.56  e1d2 e8d7 d2d3 d7e6 d3d4 e6d6 e2e4 d6e6
                                   e4e5 e6f5
 11    169    12964    10cs  1.36  e1d2 e8d7 d2d3 d7e6 d3d4 e6d6 e2e4 d6e6
                                   e4e5 e6f5 d4d5
 12    169    17569    12cs  1.36  e1d2 e8d7 d2d3 d7e6 d3d4 e6d6 e2e4 d6e6
                                   e4e5 e6f5 d4d5 f5f4
 13    179    25323    15cs  1.40  e1d2 e8d7 d2d3 d7e6 d3d4 e6d6 e2e4 d6c7
                                   e4e5 c7c6 d4c4 c6d7 c4d4
 14    197    36756    17cs  1.45  e1d2 e8d7 d2d3 d7e6 d3d4 e6d6 e2e4 d6e6
                                   e4e5 e6f5 d4d5 f5f4 e5e6 f4e3
 15    192    75255    22cs  1.72  e1d2 e8d7 d2d3 d7e6 d3d4 e6d6 e2e4 d6e6
                                   e4e5 e6f5 d4d5 f5g6 e5e6 g6f6 d5d6 f6f5
 16    192    89545    24cs  1.56  e1d2 e8d7 d2d3 d7e6 d3d4 e6d6 e2e4 d6e6
                                   e4e5 e6f5 d4d5 f5g6 e5e6 g6f6 d5d6 f6f5
 17    212   111579    27cs  1.22  e1d2 e8d7 d2d3 d7e6 d3d4 e6d6 e2e4 d6e6
                                   e4e5 e6e7 d4d5 e7d7 e5e6 d7c7 d5c5 c7d8
                                   c5d6 d8e8
 18    212   141538    31cs  1.26  e1d2 e8d7 d2d3 d7e6 d3d4 e6d6 e2e4 d6e6
                                   e4e5 e6e7 d4e4 e7d7 e4d5 d7e8 e5e6 e8f8
                                   d5d6 f8e8
 19    212   179624    37cs  1.27  e1d2 e8d7 d2d3 d7e6 d3d4 e6d6 e2e4 d6e6
                                   e4e5 e6e7 d4e4 e7f7 e4d5 f7g7 d5d6 g7f7
                                   e5e6 f7f8 d6d7 f8g7
 20    212   230495    43cs  1.28  e1d2 e8d7 d2d3 d7e6 d3d4 e6d6 e2e4 d6e6
                                   e4e5 e6e7 d4e4 e7f7 e4d5 f7f8 e5e6 f8e7
                                   d5e5 e7d8 e5d6 d8e8
 21    222   976855   139cs  2.33  e1d2 e8d7 d2d3 d7e6 d3e4 e6d6 e4f5 d6e7
                                   f5e5 e7f7 e5d6 f7g7 e2e4 g7f8 d6e6 f8g7
 22    237    1774K   238cs  2.77  e1d2 e8d7 d2d3 d7e6 d3e4 e6d6 e4f5 d6e7
                                   f5e5 e7f7 e5d6 f7g7 e2e4 g7f8 e4e5 f8e8
                                   d6e6 e8f8 e6f6 f8g8 f6e7 g8h7 e5e6 h7g6
                                   e7d6
 23    237    2343K   306cs  1.55  e1d2 e8d7 d2d3 d7e6 d3e4 e6d6 e4f5 d6e7
                                   f5e5 e7f7 e5d6 f7g7 e2e4 g7f7 e4e5 f7e8
                                   d6e6 e8f8 e6d7 f8g7 e5e6 g7f6 e6e7
 24    267    5953K   769cs     *  e1d2!
 24    347    8290K     13s     *  e1d2!!
 24    942      14M     23s  2.87  e1d2 e8d7 d2d3 d7e6 d3e4 e6d6 e4f5 d6e7
                                   f5e5 e7f7 e5d6 f7e8 e2e4 e8d8 e4e5 d8c8
                                   d6e6 c8c7 e6f7 c7c6 e5e6 c6b5 e6e7 b5c4
                                   e7e8q
 25    942      14M     23s  2.53  e1d2 e8d7 d2d3 d7e6 d3e4 e6d6 e4f5 d6e7
                                   f5e5 e7f7 e5d6 f7e8 e2e4 e8d8 e4e5 d8c8
                                   d6e6 c8c7 e6f7 c7c6 e5e6 c6b5 e6e7 b5c4
                                   e7e8q
 26    942      16M     25s  1.05  e1d2 e8d7 d2d3 d7e6 d3e4 e6d6 e4f5 d6e7
                                   f5e5 e7f7 e5d6 f7e8 e2e4 e8d8 e4e5 d8c8
                                   d6e6 c8c7 e6f7 c7c6 e5e6 c6b5 e6e7 b5c4
                                   e7e8q c4d3
NPS&#58; 494411
Time&#58; 11999cs
Branching factor&#58; 1.89
TT hits&#58; 6%
Depth&#58; 26

It fails high at 7.6 seconds, and finish that iteration at 23 (it finds the pv between 13 and 23 seconds).

There is no endgame knowledge for this position, there isn't passed pawn push extensions (both should make the search faster).

System: AMD TURION 64x2 mobile 1'6 GHz, 2GB RAM

hgm · Post by **hgm** » Thu May 28, 2009 8:54 pm

Micro-Max 4.8m (gcc compile) needs 24 ply and 24 sec to find the promotion, with 48MB hash table (always replace). The only knowledge in there is that a Pawn on the 6th is worth about 140, on the 7th about 200.

Code: Select all

go
-1     -1        1          1 e2e3
 0     74        1          2 e2e3
 1     81        1          9 e2e4
 2     75        1         56 e2e4
 3     81        1        137 e2e4
 4     77        1        510 e2e4
 5     81        1       1049 e2e4
 6     79        1       1924 e2e4
 7     79        1       2935 e2e3
 8     79        1       4662 e2e3
 9     81        3       8922 e2e4
10     81        4      15298 e2e4
11     81        5      24090 e2e4
12     81        7      37851 e2e4
13     82       14      78987 e1d2
14    139       22     127746 e1d2
15    140       37     216192 e1d2
16    139       53     311341 e1d2
17    148      109     665291 e1d2
18    153      136     827118 e1d2
19    183      165     993999 e1d2
20    196      247    1465897 e1f2
21    196      343    2027235 e1d2
22    194      788    4645882 e1d2
23    213     1340    7890272 e1f2
24    823     2435   14472865 e1f2

Kempelen · Post by **Kempelen** » Fri May 29, 2009 2:39 pm

bob wrote: If you can't solve this one in a second or so, you still have some search/transposition table issues even though you think it is working correctly...

This position is became me crazy...... have used a lot of hours between today and yesterday and dont find anything bad.
The only suspicious thing is tree is usually busher than other engines. If I put king i.e in d4 or e4 it find the solution, althought it needs same time, so I think is samething which makes the tree to grow more than usuall. It is strange because in normal situations my engine works fine, even in tactical suite like wac it solves 290 of them.

I need a little help here. I get a 11% hash hits for a 20depth search for the position. What hash hit ratio are other getting here?

Any other tip which can help me?.....

thanks in advance.

Desperado · Post by **Desperado** » Fri May 29, 2009 3:05 pm

Hi Fermin,

try to switch of any selection in the search(futility,nullmove,lmr...).
Trans and alphbeta should be enough for solving the position(finding the promotion at least).

Second, you may test first the following position w: Ke6,Pe5 b: Ke8...
if this works well put white king and pawn back, rank by rank.

if the initial position doesnt work something strange is going on,ESPECIALLY if SELECTIVE-SEARCH is off.

if you state your winning position with a "winning offset" like

Kinpos==pawnpos+2infront -> score+=winningscore(20000 for example)
also switch that off, because this may obviate the promotion...

Good luck

MattieShoes · Post by **MattieShoes** » Fri May 29, 2009 4:53 pm

For a 20 ply search:

Code: Select all

# Hash&#58;
#         Probes            Hits          Misses         Cutoffs
#        5720795         5695810           24985         3528920
#         100.0%           99.6%            0.4%           61.7%
#
# Hash table data&#58;
#            Size&#58; 16777216 entries
#          Filled&#58;    24022 &#40;0.1%)
#         Current&#58;    24022 &#40;100.0%)
#   ALPHA   nodes&#58;     8070 &#40;33.6%)
#   BETA    nodes&#58;    15070 &#40;62.7%)
#   EXACT   nodes&#58;      882 &#40;3.7%)
#   Has best move&#58;    13606 &#40;56.6%)
#     Depths&#58;
#    0&#58;       4468 &#40;18.6%)
#    1&#58;       3666 &#40;15.3%)
#    2&#58;       2574 &#40;10.7%)
#    3&#58;       3023 &#40;12.6%)
#    4&#58;       2117 &#40;8.8%)
#    5&#58;       2192 &#40;9.1%)
#    6&#58;       1296 &#40;5.4%)
#    7&#58;       1356 &#40;5.6%)
#    8&#58;        704 &#40;2.9%)
#    9&#58;        729 &#40;3.0%)
#   10&#58;        463 &#40;1.9%)
#   11&#58;        463 &#40;1.9%)
#   12&#58;        301 &#40;1.3%)
#   13&#58;        260 &#40;1.1%)
#   14&#58;        168 &#40;0.7%)
#   15&#58;        144 &#40;0.6%)
#   16&#58;         46 &#40;0.2%)
#   17&#58;         35 &#40;0.1%)
#   18&#58;         10 &#40;0.0%)
#   19&#58;          6 &#40;0.0%)
#  20+&#58;          1 &#40;0.0%)

It finds the promotion on ply 24.

Problems with this position

Problems with this position

Re: Problems with this position

Re: Problems with this position

Re: Problems with this position

Re: Problems with this position

Re: Problems with this position

Re: Problems with this position

Re: Problems with this position

Re: Problems with this position

Re: Problems with this position