I just pushed a new version (v3) of my recreational chess project iCE to my website. Sorry, still no SMP but should be somewhat stronger anyway.
Changes:
UCI Options Cleanup The UCI options of iCE are restructured and renamed. Details can be found here: http://macechess.blogspot.de/2015/07/ic ... tions.html
Multi PV support iCE is now supporting Multi PV mode.
Bugfix The ponder protocol implementation is fixed. In Shredder GUI it did not work.
Evaluation: A bit of weight tuning was perfromed. However most changes failed or had only minor impact.
Search Changes: iCE 2 added History Heuristics, Late Move Pruning, Razoring and Counter Move Heuristics. For iCE 3 I optimized the conditions a bit further to make smarter pruning decisions.
Books: I finally touched my book code again. It was necessary as the books contained some errors coming from changes to the internal move format. I also improved the book building framework to use perfromance data of iCE in certain lines.
Time Management: I changed the algorithm for time allocation to not use all available time on the last move before TC.
More details and a history of what I have done over the past months are here on my blog http://macechess.blogspot.de/
You can download it via
http://www.fam-petzke.de/cp_download_en.shtml
iCE 3 gets released
Moderator: Ras
-
tpetzke
- Posts: 686
- Joined: Thu Mar 03, 2011 4:57 pm
- Location: Germany
-
Graham Banks
- Posts: 46256
- Joined: Sun Feb 26, 2006 10:52 am
- Location: Auckland, NZ
Re: iCE 3 gets released
Thanks Thomas.tpetzke wrote:I just pushed a new version (v3) of my recreational chess project iCE to my website. Sorry, still no SMP but should be somewhat stronger anyway.
I'll put it into my Division 2 tournament shortly.
gbanksnz at gmail.com
-
Sylwy
- Posts: 5317
- Joined: Fri Apr 21, 2006 4:19 pm
- Location: IAȘI - the historical capital of MOLDOVA
- Full name: Silvian Rucsandescu
Re: iCE 3 gets released
Thank you !
-
tpetzke
- Posts: 686
- Joined: Thu Mar 03, 2011 4:57 pm
- Location: Germany
Re: iCE 3 gets released
Yes, sorry. I realized that you already started D2 with the old version but I thought I let it out into the wild anyway
-
supersharp77
- Posts: 1268
- Joined: Sat Jul 05, 2014 7:54 am
- Location: Southwest USA
Re: iCE 3 gets released
Thank You........we Love Mace......Ice....Ice 1.0 Ice 2.0 and Ice 3.0 !! 
-
Graham Banks
- Posts: 46256
- Joined: Sun Feb 26, 2006 10:52 am
- Location: Auckland, NZ
Re: iCE 3 gets released
No problem at all.tpetzke wrote:Yes, sorry. I realized that you already started D2 with the old version but I thought I let it out into the wild anyway
gbanksnz at gmail.com
-
Frank Quisinsky
- Posts: 7554
- Joined: Wed Nov 18, 2009 7:16 pm
- Location: Gutweiler, Germany
- Full name: Frank Quisinsky
Re: iCE 3 gets released
Hi Thomas,
thanks for the new ICE version!
I had a lot of fun with ice and your engine.
ice2 plays a lot of games for my Rating System.
6.150 games and the rating are since 1.200 games around 2765.
Short question:
ice2 gave often around +0.5-1.0 a higher eval after opening book moves as other engines. Will be this one the same compare to version 2.0. If I understand you changed here a bit.
Have a good new year Thomas and thanks again for your wonderful engines.
Best
Frank
thanks for the new ICE version!
I had a lot of fun with ice and your engine.
ice2 plays a lot of games for my Rating System.
6.150 games and the rating are since 1.200 games around 2765.
Short question:
ice2 gave often around +0.5-1.0 a higher eval after opening book moves as other engines. Will be this one the same compare to version 2.0. If I understand you changed here a bit.
Have a good new year Thomas and thanks again for your wonderful engines.
Best
Frank
-
tpetzke
- Posts: 686
- Joined: Thu Mar 03, 2011 4:57 pm
- Location: Germany
Re: iCE 3 gets released
Hi Frank,
If you divide the iCE score by 2 you often get a good match to the score of its opponent.
Probably yes. The pawn advantage in iCE is lower compared to other engines. I do not scale the output to match the usual pawn advantage scheme of other engines but output the raw scores. They are the product of my tuning process and the tuner does not care about it.ice2 gave often around +0.5-1.0 a higher eval after opening book moves as other engines. Will be this one the same compare to version 2.0. If I understand you changed here a bit.
If you divide the iCE score by 2 you often get a good match to the score of its opponent.
-
Ralf Müller
- Posts: 127
- Joined: Sat Dec 29, 2012 12:07 am
Re: iCE 3 gets released
Great! I'm very looking forward to test your engine when I'm back at home! 
-
mar
- Posts: 2949
- Joined: Fri Nov 26, 2010 2:00 pm
- Location: Czech Republic
- Full name: Martin Sedlak
Re: iCE 3 gets released
tpetzke wrote:still no SMP but should be somewhat stronger anyway.
Ok expecting only +100 then.
Seriously - congratulations, Thomas!