Search found 493 matches

by D Sceviour
Tue Aug 25, 2015 3:09 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Hash Refutation
Replies: 35
Views: 6027

Re: Hash Refutation

Robert Hyatt asked, ...the "no entries in the remaining hash table" implies (to me at least) that nothing was searched? H.G. Muller paraphrased the problem very well, You now basically know nothing about the remaining moves, and would try to find an alternative cut move amongst them by blindly plung...
by D Sceviour
Tue Aug 25, 2015 1:55 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Hash Refutation
Replies: 35
Views: 6027

Re: Hash Refutation

H.G.Muller wrote, The self-deepening prevents that, by guaranteeing that any fail high is searched to the full node depth, rather than just the iteration depth. Yes, that should result in a complete score. However, the idea with hash refutation is not to increase node counts with full depth searches...
by D Sceviour
Tue Aug 25, 2015 11:52 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Hash Refutation
Replies: 35
Views: 6027

Re: Hash Refutation

Hello H.G. Muller,
Your analysis looks very good.

The self-deepening looks interesting. However, I do not understand "when you get a fail high during IID". IID is supposed to return a fail-high since we are looking for the best move in the move ordering.
by D Sceviour
Tue Aug 25, 2015 11:27 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Hash Refutation
Replies: 35
Views: 6027

Re: Hash Refutation

Hello Robert Hyatt, I am glad you are taking an interest in my chess engine project. Schooner's output is as follows: White(1): analyze Search Iteration = 1 PV [1] 20 d6-c6 d7-c6 Search Iteration = 2 PV [2] 20 d6-c6 d7-c6 g2-g7 PV [2] 35 h5-h6 g7-g6 Search Iteration = 3 PV [3] 43 h5-h6 g7-g6...
by D Sceviour
Mon Aug 24, 2015 11:50 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Hash Refutation
Replies: 35
Views: 6027

Hash Refutation

The problem of tree thrashing or tree explosion has been mentioned in a number of postings. Sometimes the problem is due to a missed mate in one. Here is an approach to the problem of tree explosions. In the example, start with a known mate in 14: [d]6r1/2rp1kpp/2qQp3/p3Pp1P/1pP2P2/1P2KP2/P5R1/6R1 w...
by D Sceviour
Sun Aug 23, 2015 10:18 pm
Forum: Computer Chess Club: General Topics
Topic: Schooner Version 1.4
Replies: 29
Views: 8845

Re: Schooner Version 1.4.1

After 6 rounds, Schooner shares third place in the 54th CCRL Amateur D6. Thank you to Graham Banks and Gabor Szots for getting the engine into the tournament and running quickly.
by D Sceviour
Fri Aug 21, 2015 12:03 pm
Forum: Computer Chess Club: General Topics
Topic: Schooner Version 1.4
Replies: 29
Views: 8845

Re: Schooner Version 1.4.1

Here are the patches for Schooner1.4.1 :

1. Fixed WARNING ERR bug for polyglot book
2. Improved xboard force commands. Arena .abk books should work now.
3. Added a simple sc.ini file for hash and book.
by D Sceviour
Thu Aug 20, 2015 7:25 pm
Forum: Computer Chess Club: General Topics
Topic: Schooner Version 1.4
Replies: 29
Views: 8845

Re: Schooner Version 1.4

Graham, The program works as promised from the initial post providing no changes are made. However, the current version 1.4 requires a polyglot "book.bin" to be installed in the same directory or the engine will crash. There is a fix for this and will be updated along with some improvements to the x...
by D Sceviour
Thu Aug 20, 2015 4:22 pm
Forum: Computer Chess Club: General Topics
Topic: Schooner Version 1.4
Replies: 29
Views: 8845

Re: Schooner Version 1.4

OK, ShortLines.abk uses the force command to send a sequence of moves. Since there were three moves, the side to move was supposed to change. 2015-08-20 15:36:24,644-->1:force 2015-08-20 15:36:24,852-->1:c2c4 2015-08-20 15:36:24,852-->1:e7e5 2015-08-20 15:36:24,852-->1:c4c5 Unfortunately Schooner do...
by D Sceviour
Thu Aug 20, 2015 3:31 pm
Forum: Computer Chess Club: General Topics
Topic: Schooner Version 1.4
Replies: 29
Views: 8845

Re: Schooner Version 1.4

Gabor, thank you. It did not take long to find a way to get the engine to crash. Apparently, the interface takes the following WARNING ERR very seriously and aborts the process: 2015-08-20 15:36:24,624<--1:#Polyglot Book "book.bin" not found - WARNING ERR # 2 Deleting the WARNING ERR line from the c...