Search found 104 matches

by Cheney
Wed Jul 22, 2020 5:54 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: When a Node Type Changes...
Replies: 2
Views: 1475

Re: When a Node Type Changes...

Thank you :) That statistic has been used and what lead me to wanting to measure each node type. I have collected data on fail highs, first move fail highs, and fail highs within the first 5 moves. I have also collected this data per move type - hash, capture, promotions, etc. I have not collected t...
by Cheney
Tue Jul 21, 2020 7:33 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: When a Node Type Changes...
Replies: 2
Views: 1475

When a Node Type Changes...

Hi, I decided to study a little bit more into Knuth's node types to measure the frequency in which a type does not do what it was expected to do and I am now trying to wrap my mind around the basic data. I've read a few dozen posts, thought I read at one time what it means when a node type changes, ...
by Cheney
Thu Jul 09, 2020 5:25 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Questions on Razoring and Code Structure for Pruning
Replies: 23
Views: 5100

Re: Questions on Razoring and Code Structure for Pruning

OK, thanks! I did attempt that before the search and when it failed miserably (getting deeper but playing crazy lines) I figured I was not understanding the process. I tried various options, like trying it only when futility pruning was not allowed, static depths, smaller margins, etc. I did not inc...
by Cheney
Thu Jul 09, 2020 4:26 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Questions on Razoring and Code Structure for Pruning
Replies: 23
Views: 5100

Re: Questions on Razoring and Code Structure for Pruning

Ty :) The easy answer :) - I have reviewed the Crafty code and did not see razoring. I believe Dr. Hyatt removed that; I have seen comments on other posts where it was removed. I forgot about main.c and all the details there, going to read it, thanks. Normally the margin for futility pruning is so l...
by Cheney
Thu Jul 09, 2020 2:00 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Questions on Razoring and Code Structure for Pruning
Replies: 23
Views: 5100

Questions on Razoring and Code Structure for Pruning

Hi, I have recently added in extended futility pruning (up to 3 plys) and am having some positive success. I am reading about Razoring, have tried laying in some flavors of code, tests are not positive. I also decided to look at some other code and now wonder about more than razoring but also code s...
by Cheney
Wed Jul 08, 2020 12:09 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Futility Pruning Issues
Replies: 5
Views: 1502

Re: Futility Pruning Issues

Problem solved :) After going back to the standard implementation of FP and Extended FP, I realized a bug I fixed but never tested under the standard deployment - it was only tested using the qsearch deployment. The bug was to flag any move that generated check - this was performed when generating a...
by Cheney
Tue Jul 07, 2020 7:59 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Futility Pruning Issues
Replies: 5
Views: 1502

Re: Futility Pruning Issues

After reading your points, I see a definite mistake. I am performing the qsearch for each move I want to prune and yet I am not making that move; thus, I am repeating the same qsearch for every prunable move in that node. I guess where I was trying to go with the qsearch validation was to avoid prun...
by Cheney
Tue Jul 07, 2020 1:51 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Futility Pruning Issues
Replies: 5
Views: 1502

Re: Futility Pruning Issues

Ty for the input :) The interesting thing about that move is it had a high history score. One test I performed without the qsearch validation was to only prune moves not matched by any sorting method. In other words, since history is the last score applied then if the score is zero, look to prune th...
by Cheney
Tue Jul 07, 2020 12:14 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Futility Pruning Issues
Replies: 5
Views: 1502

Futility Pruning Issues

Hi, I have been trying to implement Futility pruning (FP) for some time now and have not had much success. I have read almost every post linked on CPW for this subject. I have tried all kinds of variations but when my base engine plays the one with FP, the FP version loses. However, when I compare t...
by Cheney
Mon Jul 06, 2020 5:34 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Testing Move Order Quality
Replies: 9
Views: 2002

Re: Testing Move Order Quality

Ty Cardoso :) Pruning is the next item on the chopping block. Before I went there, I wanted to make sure I had a sound process for move ordering. I also wanted to develop a way to flag moves as important (to not prune or reduce). As for this sort ordering, I was able to use the data and shift move t...