Search found 244 matches

by MichelG
Wed Sep 27, 2017 08:19
Forum: Draughts, Computer, Internet
Topic: Database Dragon
Replies: 4
Views: 3386

Re: Database Dragon

The only endgame database that are currently available, you can download from https://mdgsoft.home.xs4all.nl/draughts/endgame.htm I think you already have those; it's only a subset of the most important endgames. Most of the others are drawish endgames with lots of kings and would only slow down the...
by MichelG
Wed Sep 27, 2017 08:12
Forum: Draughts, Computer, Internet
Topic: dragon 4.5
Replies: 52
Views: 30887

Re: dragon 4.5

Michel, which processor are you thinking about to buy in the (near) future. Also DB Generation scales quite good with the number of cores. 10+ i think, probably intel. I would like to know the performance of Ryzen though on the draughts engine and some other projects. DB generation is mostly limite...
by MichelG
Mon Sep 25, 2017 20:07
Forum: Draughts, Computer, Internet
Topic: dragon 4.5
Replies: 52
Views: 30887

Re: dragon 4.5

Based upon these results Dragon 4.6.1 vs scan 3.0: +1 -38 =547 (46.8%, -15 elo), my ELO calculation yields 22? So not sure if I have done something wrong. Also interesting to see that these results are based upon a 1 min/game, whereas tournaments (like Olympics) are based upon 20 min/game. Last but...
by MichelG
Mon Sep 25, 2017 09:00
Forum: Draughts, Computer, Internet
Topic: dragon 4.5
Replies: 52
Views: 30887

Re: dragon 4.5

I fixed the memory leak issue, and played some matches against scan again. Dragon 4.6.1 vs scan 2.0: +12 -9 =452 (50.2%, +2 elo) Dragon 4.6.1 vs scan 3.0: +1 -38 =547 (46.8%, -15 elo) (Not quite a scientific test; i was using the computer with the matches running, and they don't have the same number...
by MichelG
Thu Sep 14, 2017 08:06
Forum: Draughts, Computer, Internet
Topic: dragon 4.5
Replies: 52
Views: 30887

Re: dragon 4.5

What are the results against Scan 3.0? Nobody has published anything. Fabien. I did a test last night and I am getting results similar to Krzysztof (39-3-186 for scan 3.0) I believe there is a problem with memory management somewhere; i can see dragon swapping to disk like crazy, and nodes per seco...
by MichelG
Thu Aug 31, 2017 18:47
Forum: Draughts, Computer, Internet
Topic: dragon 4.5
Replies: 52
Views: 30887

Re: dragon 4.5

Dragon 4.6 is now available as beta-test from: https://mdgsoft.home.xs4all.nl/draughts/download/SetupDragon%20v4.6.0.msi - bugfixes and improvements in the engine everywhere - engine equal to 2017 tournament minus some big tables - experimental feature: dragon can now import/copy&paste jpg or png im...
by MichelG
Sun Jul 02, 2017 18:55
Forum: Draughts, Computer, Internet
Topic: Computer Olympiad 2017 Draughts
Replies: 62
Views: 26689

Re: Computer Olympiad 2017 Draughts

Congratulations to Fabien Letouzey again! Results: 1. Scan 33 2. Sjende Blyn 32 3. Dragon Draughts 32 4. Moby dam 27 5. Tornado 22 6. TDKing 21 7. GWD 19 8. Deep Nikita 13 9. Cerberus 12 10. Slagzet.com 7 11. Bit S.a.y. 0
by MichelG
Tue Jun 06, 2017 09:26
Forum: Draughts, Computer, Internet
Topic: AMD Ryzen
Replies: 21
Views: 11352

Re: AMD Ryzen

One more datapoint for the draw discussion.... Latest match dragon vs scan 2.0 gave +22 -25 =941 (from dragon's perspective), at 1 minute/game, 6 cores. That's 95% draw. I used 3-ply ballots for this match. One thing i noted that not all openings are balanced. For example after 1. 33-28 17-21 28-23 ...
by MichelG
Thu Jun 01, 2017 21:48
Forum: Draughts, Computer, Internet
Topic: EndGame Databases
Replies: 289
Views: 134187

Re: EndGame Databases

BertTuyt wrote:I tested also hyperthreading (so 16 threads), and now the time is 22.3 seconds, so 23.1 MPositions/second.

Edit: Small improvement which yield 20.6 sec (aka 25.0 MPos/sec).

Bert
Assuming it scales, that's fast enough to do 5x5 in a day. 6x6 would take a month and fill up a 8 TB harddisk :-)
by MichelG
Thu Jun 01, 2017 21:14
Forum: Draughts, Computer, Internet
Topic: AMD Ryzen
Replies: 21
Views: 11352

Re: AMD Ryzen

Michel, you are right that with short time scales, the non draw rate increases. But with normal tournament conditions, the programs who utilize all processing power will draw more and more. I guess I could win against Bolt running the 100m, when his legs are tied, but under normal conditions , no w...
by MichelG
Thu Jun 01, 2017 13:46
Forum: Draughts, Computer, Internet
Topic: AMD Ryzen
Replies: 21
Views: 11352

Re: AMD Ryzen

And now..... The Intel i9, with affordable 8 core CPUs, and a 18-core extreme model. So up to another game, as normal Draughts is going into draw mode... Which game do you suggest? I think chess is the only other game left where MC and/or ML haven't taken over (yet...) There are still some challeng...
by MichelG
Wed Apr 19, 2017 09:24
Forum: Draughts, Computer, Internet
Topic: Search Algorithm
Replies: 711
Views: 236508

Re: Search Algorithm

It all seems to go to draw and perfection pretty fast at this level. Maybe it is time for everyone to switch to breakthrough-draughts :-) Surely there is room for competition there. I'll do some independent test, but that needs to wait a couple of weeks for my computer is working out some new algori...
by MichelG
Wed Apr 19, 2017 09:19
Forum: Draughts, Computer, Internet
Topic: The unofficial world championships in draughts international 2017
Replies: 38
Views: 18996

Re: The unofficial world championships in draughts international 2017

Hi Krzysztof,

Why are you using 4 threads for some programs and 2 for others in this tournament?

Michel
by MichelG
Mon Mar 27, 2017 13:02
Forum: Draughts, Computer, Internet
Topic: Search Algorithm
Replies: 711
Views: 236508

Re: Search Algorithm

Especially to verify if we really see diminishing returns, or that we see a tail with constant ELO differences. Diminishing returns are mathemaically inevatable. Draughts games have a finite length. Even with LMR en propcut, you can prove that infinite search will reach the game-theorical value. A ...
by MichelG
Wed Mar 22, 2017 12:53
Forum: Draughts, Computer, Internet
Topic: Search Algorithm
Replies: 711
Views: 236508

Re: Search Algorithm

Hi Bert, How long did it take to produce up til 22 ply? It would be nice to see the datapoints at higher ply as well :-) I don't think you can conclude there is an exponential fit here. Any other curve would fit as well, even a strait line would fit. At a later time, i will play dragon against scan ...