8th International CSVN Tournament (ICT) 30-31 May/1 June

[Harvey Williamson]

http://www.csvn.nl/index.php?option=com ... &Itemid=28

This tournament kicks off next Friday in Leiden. Full coverage on this forum and all Hiarcs games live on Playchess.

Participants so far are here http://www.csvn.nl/index.php?option=com ... &Itemid=28

[Harvey Williamson]

A quick Leiden warm up at 16+0. Rybka was on Suj's 16 core. At the bottom of the table you will see Hiarcs gets a GM norm while Rybka does not


Erdo's Rundenturnier 2008
(TD: Erdo, Time: 16m + 0s)

1 HarveyHiarcsX,HIARCS 12.118 MP 2589 +13 &;
½ 1 ½ ½ ½ 1 1 ½ ½ 1 7.0 / 10
2 Devil_j9,Rybka 2.3.2a mp 2702 -150 ½ &;
0 ½ 1 ½ ½ 1 1 ½ 1 6.5 / 10 0.00
2 Erdo,HIARCS 12.106 MP 0 1 &;
1 ½ ½ ½ 1 1 ½ ½ 6.5 / 10 0.00
4 Alinghi08,Zappa X 2571 -43 ½ ½ 0 &;
1 1 ½ 0 1 ½ 1 6.0 / 10
5 Ultra-d,Rybka 2.3.2a mp 2576 -84 ½ 0 ½ 0 &;
½ 1 1 ½ 1 ½ 5.5 / 10
6 CompMaster,Toga II 1.4.2JD 2556 -97 ½ ½ ½ 0 ½ &;
1 ½ ½ ½ ½ 5.0 / 10
7 Ultra-blitz,Toga II 1.4.2JD 2609 -190 0 ½ ½ ½ 0 0 &;
½ ½ 1 1 4.5 / 10 0.00
8 Chessfurby,Archimedes 1.0 2326 +121 0 0 0 1 0 ½ ½ &;
½ 1 1 4.5 / 10 0.00
9 Rocky III,Zappa X 2617 -234 ½ 0 0 0 ½ ½ ½ ½ &;
1 ½ 4.0 / 10
10 Kingofengines,Deep Shredder 11 x6 ½ ½ ½ ½ 0 ½ 0 0 0 &;
1 3.5 / 10
11 Solidius,Deep Shredder 11 UC 0 0 ½ 0 ½ ½ 0 0 ½ 0 &;
2.0 / 10
Durchschnitt Elo: 2467 <=> Cat: 9 gm = 7.00 m = 5.00

[turbojuice1122]

Yes, I'm sure that's an unbiased tournament, likely to predict the result of Leiden. Haha, what books were used for the various programs?

I have a tournament of my own going on with a time control of 15'+15" (though it's only on my single-core computer), and after nine rounds, we have the following:

Code: Select all

1. Rybka 2.3.2a 32-bit   2870 +116            7.0
2. Ktulu 8               2670 +229            6.0
3. Naum 3.1              2795 +91             6.0
4. Fritz 11              2820 -5              5.0
5. Shredder 11           2795 +19             5.0
6. Fruit 2.4 Test Beta A 2760 +53             5.0
7. Hiarcs Paderborn      2760 +16             4.5
8. Hiarcs 12             2775 -30             4.5
9. SmarThink 1.10        2675  +82            4.5
10. Rybka 1.2f 32-bit    2830 -105            4.0
11. Toga II 1.4.2JD      2770 -107            3.0
12. Zappa Mexico II      2755 -91             3.0/8
13. Loop 10.32f          2720  -122           2.5
14. Deep Fritz 10        2750  -175           2.0/8

[Harvey Williamson]

there were many of the playchess cookers playing with their best books in a tournament open to all on Playchess so I think it means a lot more that your tournament.

[turbojuice1122]

Suj playing with Rybka on only a 16-core? That seems to give some evidence to the contrary...

[Harvey Williamson]

Suj playing with Rybka on only a 16-core? That seems to give some evidence to the contrary...

If you ask him he will tell you the scaling is rubbish above that. But if you think a 16 core is not enough for Rybka that makes me feel happy.

[turbojuice1122]

No, my point is that if he was trying to win, he would be going with Zappa on a 32-core or 64-core.

[€rdo]

No, Suj only Play with 2 Rybka on 2 x 16 way and with Shredder 11 X64 on another 8 way Skulltrail with 8 x 4 Ghz own Book from Suj !
And now u say if he like too win Than he play with Zappa hahhaha why not Rybka lol !

Good Luck €rdo

[Harvey Williamson]

No, my point is that if he was trying to win, he would be going with Zappa on a 32-core or 64-core.

We said no more than 2 of any engine. There were 2 zappas as you see. But if you can't accept this tournament as honest that is fine with me. I expect it when rybka does not win.

[turbojuice1122]

No, Suj only Play with 2 Rybka on 2 x 16 way and with Shredder 11 X64 on another 8 way Skulltrail with 8 x 4 Ghz own Book from Suj !
And now u say if he like too win Than he play with Zappa hahhaha why not Rybka lol !

Good Luck €rdo

Because, very simply, Zappa is stronger than Rybka on hardware that is as powerful as what Suj has.

[turbojuice1122]

No, my point is that if he was trying to win, he would be going with Zappa on a 32-core or 64-core.

We said no more than 2 of any engine. There were 2 zappas as you see. But if you can't accept this tournament as honest that is fine with me. I expect it when rybka does not win.

No, it's not the fact that Rybka didn't win--it's the fact that even for a small tournament, those results are unrealistic when compared with the data when considering the fact that Rybka on a 16-core with supposedly good book gets beat by 1.5 points. Statistically, it can happen, but it's highly improbable--thus, something else is going on. The other result, by Devil_j9, is quite understandable and within the realm of statistical error, so I don't question that one. Also, we know from past messages and such that obviously Suj doesn't like to make Rybka or the Rybka team look good nowadays, so not doing well in a fairly meaningless tournament like this, but one that has potential PR value, is not a problem.

[turbojuice1122]

Granted, RockyIII's result of being 3 points out of first is also statistically improbable, though I have no idea what hardware was used there.

[€rdo]

Can u put this Message on Rybka Forum also

Bye €rdo

[turbojuice1122]

Rybka lost by 0.5 points.

But the result you quoted concerning Suj's machine had Rybka losing by 1.5 points.

[Harvey Williamson]

Devil_J9 is Suj.