LCT II Question

[jsgroby]

I am working with some code to analyze the epd file for LCT II. My question is, based on the following entry in the epd file:

1k1r3r/pp2qpp1/3b1n1p/3pNQ2/2pP1P2/2N1P3/PP4PP/1K1RR3 b - - bm Bb4; id "LCTII.POS.02"; c0 "Lilienthal - Botvinnik, Moskau 1945";

What exactly does the Bb4 mean, is that the solution that the engine needs to find within 10 mins for ELO scoring?

I picked up the epd file from here:

http://chessprogramming.wikispaces.com/LCT+II

Thanks
Jeff

[jsgroby]

I am working with some code to analyze the epd file for LCT II. My question is, based on the following entry in the epd file:

1k1r3r/pp2qpp1/3b1n1p/3pNQ2/2pP1P2/2N1P3/PP4PP/1K1RR3 b - - bm Bb4; id "LCTII.POS.02"; c0 "Lilienthal - Botvinnik, Moskau 1945";

What exactly does the Bb4 mean, is that the solution that the engine needs to find within 10 mins for ELO scoring?

I picked up the epd file from here:

http://chessprogramming.wikispaces.com/LCT+II

Thanks
Jeff

Just to be clear, the code is reading in the file and doing the searches for the best position as it should, but my results are not coming back the way I think they should, which is why I am asking about that Bb4 in the string. Whether or not that is the best move that the engine should be finding?

[ZirconiumX]

bm - best move.
am - avoid move.

Matthew:out

[jsgroby]

bm - best move.
am - avoid move.

Matthew:out

Thank you.

jeff

[jsgroby]

Ok, new question, in order to score your engine after running the LCTII test suite against it, do you find your average score then multiply that against 35 and add 1900? That is how I understood the article in the chess programming wiki.

Please advise.

Thanks
Jeff

[ZirconiumX]

Not quite.

Points per position
30 points, if solution is found between 0 and 9 seconds
25 points, if solution is found between 10 and 29 seconds
20 points, if solution is found between 30 and 89 seconds
15 points, if solution is found between 90 and 209 seconds
10 points, if solution is found between 210 and 389 seconds
5 points, if solution is found between 390 and 600 seconds
0 points, if not found with in 10 minutes

You take the total number of points and add 1900.

Matthew:out

[jsgroby]

Not quite.

Points per position
30 points, if solution is found between 0 and 9 seconds
25 points, if solution is found between 10 and 29 seconds
20 points, if solution is found between 30 and 89 seconds
15 points, if solution is found between 90 and 209 seconds
10 points, if solution is found between 210 and 389 seconds
5 points, if solution is found between 390 and 600 seconds
0 points, if not found with in 10 minutes

You take the total number of points and add 1900.

Matthew:out

Again, thanks for clarifying...

[jsgroby]

Not quite.

Points per position
30 points, if solution is found between 0 and 9 seconds
25 points, if solution is found between 10 and 29 seconds
20 points, if solution is found between 30 and 89 seconds
15 points, if solution is found between 90 and 209 seconds
10 points, if solution is found between 210 and 389 seconds
5 points, if solution is found between 390 and 600 seconds
0 points, if not found with in 10 minutes

You take the total number of points and add 1900.

Matthew:out

Again, thanks for clarifying...

With regards to LCT II results, do most of you engine programmers use LCT II results to get an idea of the ELO rating for your engine? Are the results of LCT II considered relatively reliable?

Thanks
Jeff

[Ferdy]

No, and I don't know.

The sts sets could be made as a good candidate to approximate a rating. It has multiple moves and points and varied themes. The other criteria to be factored in will be the idea from LCT, that is the time to get the move selected, not only for best move but for other moves as well.