Does perft include underpromotion?

[Chan Rasjid]

My engine failed some perft test.

I visited sites that explain perft but there was no explicit mention about under promotion. Since it is legal to have 4 promote types, I assume there would be 4 moves per promotion for perft.

Is it correct?

Thanks
Rasjid

[vladstamate]

Hi,

Yes, perft includes underpromotions since those are legal moves too.

Regards,
Vlad.

[Chan Rasjid]

Hi,

Yes, perft includes underpromotions since those are legal moves too.

Regards,
Vlad.

Thanks.

Now's the problem - trickest of part to seek needle in a haystack or under the ocean floor.

Rasjid

[michiguel]

Hi,

Yes, perft includes underpromotions since those are legal moves too.

Regards,
Vlad.

Thanks.

Now's the problem - trickest of part to seek needle in a haystack or under the ocean floor.

Rasjid

Use perftdiv to find the needle

Miguel

[Chan Rasjid]

[d]8/7p/p5pb/4k3/P1pPn3/8/P5PP/1rB2RK1 b - d3 0 28

I hope the diagram shows.

Perft(6) = 38633283
//ok crafty - 38633283

My perft fails in this position. Just may be someone is familiar with this test given by a site "Sharper" and have some hints of likely move generation errors.

My perft:-

perft 1 4, nps(0)
perft 2 94, nps(0)
perft 3 2669, nps(2869000)
perft 4 54726, nps(1509050)
perft 5 1539195, nps(2518313)
perft 6 31753730, nps(4667662)

Thanks,
Rasjid

[Chan Rasjid]

Hi,

Yes, perft includes underpromotions since those are legal moves too.

Regards,
Vlad.

Thanks.

Now's the problem - trickest of part to seek needle in a haystack or under the ocean floor.

Rasjid

Use perftdiv to find the needle

Miguel

BTW your number of posts is 2292 and mine is 229.

This is real WORK

Rasjid

[JVMerlino]

Your needle is right at the top.

Perft 1 is wrong, because there are five legal moves (not four, as shown by your output). You probably aren't counting the en passant capture at d3, which is part of the FEN.

jm

[JVMerlino]

Your needle is right at the top.

Perft 1 is wrong, because there are five legal moves (not four, as shown by your output). You probably aren't counting the en passant capture at d3, which is part of the FEN.

jm

And just to verify, I made a version of my engine that doesn't generate en passant captures, and my perft 1 through perft 5 agree with your numbers. So that is clearly most (if not all) of the problem.

However, my perft 6 is 31753706, whereas yours is 31753730 (difference of 24). But that could be another en passant effect?

jm

[Chan Rasjid]

Your needle is right at the top.

Perft 1 is wrong, because there are five legal moves (not four, as shown by your output). You probably aren't counting the en passant capture at d3, which is part of the FEN.

jm

Seems there are people around who knows how to read chess board diagrams and perft printout

I just recently implemented perft and uci with setboard from fen. Probably the en passant part is not tested yet. So I will check.

Your observation helps greatly. My level of arithmetic reaches up to addition, subtraction and multiply. "div" would be serious.

I did not expect my move generator to have problems as it is "OLD" and have played many games w/o a hitch.


Many Thanks,
Rasjid

[JVMerlino]

Your needle is right at the top.

Perft 1 is wrong, because there are five legal moves (not four, as shown by your output). You probably aren't counting the en passant capture at d3, which is part of the FEN.

jm

Seems there are people around who knows how to read chess board diagrams and perft printout

I just recently implemented perft and uci with setboard from fen. Probably the en passant part is not tested yet. So I will check.

Your observation helps greatly. My level of arithmetic reaches up to addition, subtraction and multiply. "div" would be serious.

I did not expect my move generator to have problems as it is "OLD" and have played many games w/o a hitch.

Many Thanks,
Rasjid

Glad to help! It's hard to say, but I don't think that there would be many games which would make it clear to you that your engine wasn't handling en passant properly. Just because an engine isn't choosing an en passant move, it doesn't mean that it isn't generating those moves properly.

I would guess that the only time it would come up in an obvious way would be in this instance (analyzing perft data) or in a game in which an en passant move was the only legal move AND your engine declared an invalid checkmate/stalemate. And I just can't see that happening very often.

jm