Houdini just massacred Rybka with Black :-)

[Houdini]

Correct, therefore I asked. Let's assume for a second that Houd is just R4 itself. Then you can easily conclude that if you have a tuned Houd that it must be a little bit better in a short tournament. But this is as if you would argue that R4 (tuned) is a bit better than itself. Which sounds a bit illogical.

What I mean is, if you take clones that fight its mothersource it may happen that the mother will slightly lose in short terms.

You very much sound like a bot programmed to spout semi-intelligent random sentences .

There have been several long time control Houdini 1.5 vs. Rybka 4 matches played by Paul Watson, Ingo Bauer, Sedat Canbaz, Pal Larkin and Martin Thoresen (the list is not exhaustive), the results of which can all be found on this forum. They all point to one and the same conclusion: Houdini 1.5 is about 50 Elo stronger at long time control than Rybka 4.

If you actually watch the games that are played, you will notice that moves, evaluations and PVs are quite different. For a remarkable example, see Pal Larkin's recent thread.

Robert

[Dr.Wael Deeb]

Correct, therefore I asked. Let's assume for a second that Houd is just R4 itself. Then you can easily conclude that if you have a tuned Houd that it must be a little bit better in a short tournament. But this is as if you would argue that R4 (tuned) is a bit better than itself. Which sounds a bit illogical.

What I mean is, if you take clones that fight its mothersource it may happen that the mother will slightly lose in short terms.

You very much sound like a bot programmed to spout semi-intelligent random sentences .

There have been several long time control Houdini 1.5 vs. Rybka 4 matches played by Paul Watson, Ingo Bauer, Sedat Canbaz, Pal Larkin and Martin Thoresen (the list is not exhaustive), the results of which can all be found on this forum. They all point to one and the same conclusion: Houdini 1.5 is about 50 Elo stronger at long time control than Rybka 4.

If you actually watch the games that are played, you will notice that moves, evaluations and PVs are quite different. For a remarkable example, see Pal Larkin's recent thread.

Robert

You very much sound like a bot programmed to spout semi-intelligent random sentences

Semi-intelligent random sentences
You are so overestimating our beloved friend Rolf

[Dr.Wael Deeb]

The two participants of the Elite Match are the top two finishers from Division I

Correct, therefore I asked. Let's assume for a second that Houd is just R4 itself. Then you can easily conclude that if you have a tuned Houd that it must be a little bit better in a short tournament. But this is as if you would argue that R4 (tuned) is a bit better than itself. Which sounds a bit illogical.

What I mean is, if you take clones that fight its mothersource it may happen that the mother will slightly lose in short terms.

[Rolf]

Houdini 1.5 vs. Rybka 4 matches played by Paul Watson, Ingo Bauer, Sedat Canbaz, Pal Larkin and Martin Thoresen (the list is not exhaustive)

If you actually watch the games

First a friendly hello from my side, Robert.

Wouldnt you agree that H1.5 came out after R4 already existed some months? What I want to say is that you could tune your version against a stable entity that was never changed. Could you confirm this? If you agree then you might also agree that this same procedure was always to be seen. That a number one that isnt changed is a bit helpless against tuned programs.

Yes, I see the games of this tournament and against your prog R4 never really had a chance IMO. But you made a lot of draws against far weaker competition than R4. Doesnt this raise questions for you too? H1.5 is tuned against R4.

Now the sensitive point, Robert. For me there is no doubt that you, Robert, are a splendid programmer, no doubt about it. Also in astronomy you are a genius. But the evidence speaks against you if it comes to the question if H1.5 is sober. If it were, why all the testers dont accept you? If you were sober why you didnt compete in Japan for the Wch?

I have no programming experience at all. I'm a very old psychologist. I have no axe to grind with you or any other Belge superprogrammer. I'm just a user of chess software for 30 years and a chessplayer. I'm interested in the question of betrayal in chess and identity of personalities in virtual environments and psychiatric clinics.

I have no reason therefore to treat you as an outcast. For me you are a nice guy. Question is if you could prove your identity with other means than CC programming because on that field I couldnt follow what you would say.

Perhaps you heard of the Hippo family also residing somewhere in Belgium or Russia. They didnt even have concrete names. If you know what I mean. Question is if I would find you if I would seek you even if you helped me? If yes, then why should we just meet each other?

[Houdini]

Rolf,

I don't feel like participating in your psychology experiment. If you're really interested, you'll find the answers to most of your questions on this or on the OpenChess forum.

More information about the Hippo family in Belgium can be found at this web site.

Robert

[Damir]

right on the spot...

[beram]

Dear Robert...Dear Rolf....
perhaps Robert you should be warned for the fact that Rolf is the most annoying CC troll creator, who ever lived, who has made insinuating his first nature. So I would like to dismantle his first intrigue:

Wouldnt you agree that H1.5 came out after R4 already existed some months? What I want to say is that you could tune your version against a stable entity that was never changed. Could you confirm this? If you agree then you might also agree that this same procedure was always to be seen. That a number one that isnt changed is a bit helpless against tuned programs.
This is not true it plays very well against all programs not specifically against R4. *See list below taken from Ingo Bauers IPON list

Yes, I see the games of this tournament and against your prog R4 never really had a chance IMO. But you made a lot of draws against far weaker competition than R4. Doesnt this raise questions for you too? H1.5 is tuned against R4.
This has nothing to do with tuning, when you let Houdini play hundreds of games against weaker programs there also occur a lot of draws.

Code: Select all

   1 Houdini 1.5a             3009 2700.0 (2223.5 : 476.5)
                                   100.0 ( 58.0 :  42.0) Deep Rybka 4             2954
                                   100.0 ( 60.5 :  39.5) Stockfish 2.01 JA        2919
                                   100.0 ( 61.5 :  38.5) Stockfish 1.9.1 JA       2899
                                   100.0 ( 65.5 :  34.5) Critter 0.90 SSE42       2891
                                   100.0 ( 78.0 :  22.0) Komodo64 1.3 JA          2829
                                   100.0 ( 80.0 :  20.0) Naum 4.2                 2822
                                   100.0 ( 74.0 :  26.0) Komodo 1.2 JA            2804
                                   100.0 ( 77.5 :  22.5) Deep Shredder 12         2800
                                   100.0 ( 80.0 :  20.0) Gull 1.1                 2789
                                   100.0 ( 84.5 :  15.5) Deep Fritz 12 32b        2779
                                   100.0 ( 78.0 :  22.0) Deep Sjeng c't 2010 32b  2778
                                   100.0 ( 86.5 :  13.5) Protector 1.4.0 x64      2758
                                   100.0 ( 85.0 :  15.0) spark-1.0 SSE42          2754
                                   100.0 ( 82.0 :  18.0) HIARCS 13.2 MP 32b       2750
                                   100.0 ( 87.0 :  13.0) HIARCS 13.1 MP 32b       2730
                                   100.0 ( 86.0 :  14.0) Zappa Mexico II          2714
                                   100.0 ( 87.0 :  13.0) Protector 1.3.2 JA       2701
                                   100.0 ( 90.5 :   9.5) Deep Onno 1-2-70         2684
                                   100.0 ( 88.5 :  11.5) Hannibal 1.0a            2682
                                   100.0 ( 89.0 :  11.0) Deep Junior 12           2678
                                   100.0 ( 91.0 :   9.0) Toga II 1.4 beta5c BB    2670
                                   100.0 ( 90.5 :   9.5) Strelka 2.0 B            2669
                                   100.0 ( 92.0 :   8.0) Umko 1.1 SSE42           2633
                                   100.0 ( 95.5 :   4.5) Loop 13.6/2007           2615
                                   100.0 ( 90.5 :   9.5) Jonny 4.00 32b           2614
                                   100.0 ( 92.5 :   7.5) Crafty 23.3 JA           2600
                                   100.0 ( 92.5 :   7.5) Spike 1.2 Turin 32b      2582
   2 Deep Rybka 4             2954 3800.0 (2944.5 : 855.5)
                                   100.0 ( 42.0 :  58.0) Houdini 1.5a             3009
                                   100.0 ( 46.5 :  53.5) Houdini 1.03a            2949
                                   100.0 ( 53.5 :  46.5) Stockfish 2.01 JA        2919
                                   100.0 ( 59.5 :  40.5) Stockfish 1.9.1 JA       2899
                                   100.0 ( 59.5 :  40.5) Critter 0.90 SSE42       2891
                                   100.0 ( 56.0 :  44.0) Stockfish 1.7.1 JA       2886
                                   100.0 ( 69.5 :  30.5) Komodo64 1.3 JA          2829
                                   100.0 ( 71.0 :  29.0) Naum 4.2                 2822
                                   100.0 ( 69.5 :  30.5) Critter 0.80             2819
                                   100.0 ( 69.5 :  30.5) Komodo 1.2 JA            2804
                                   100.0 ( 78.5 :  21.5) Deep Shredder 12         2800
                                   100.0 ( 76.5 :  23.5) Critter 0.70             2790
                                   100.0 ( 76.0 :  24.0) Gull 1.1                 2789
                                   100.0 ( 76.5 :  23.5) Deep Fritz 12 32b        2779
                                   100.0 ( 76.5 :  23.5) Deep Sjeng c't 2010 32b  2778
                                   100.0 ( 80.0 :  20.0) Gull 1.0a                2767
                                   100.0 ( 82.0 :  18.0) Protector 1.4.0 x64      2758
                                   100.0 ( 75.0 :  25.0) spark-1.0 SSE42          2754
                                   100.0 ( 75.5 :  24.5) HIARCS 13.2 MP 32b       2750
                                   100.0 ( 78.0 :  22.0) HIARCS 13.1 MP 32b       2730
                                   100.0 ( 79.5 :  20.5) spark-0.4                2714
                                   100.0 ( 82.0 :  18.0) Zappa Mexico II          2714
                                   100.0 ( 81.5 :  18.5) Protector 1.3.2 JA       2701
                                   100.0 ( 84.5 :  15.5) Deep Onno 1-2-70         2684
                                   100.0 ( 91.0 :   9.0) Onno-1-1-1               2683
                                   100.0 ( 87.0 :  13.0) Hannibal 1.0a            2682
                                   100.0 ( 83.5 :  16.5) Deep Junior 12           2678
                                   100.0 ( 89.0 :  11.0) Deep Sjeng WC2008        2674
                                   100.0 ( 84.0 :  16.0) Toga II 1.4 beta5c BB    2670
                                   100.0 ( 87.0 :  13.0) Strelka 2.0 B            2669
                                   100.0 ( 79.5 :  20.5) Deep Junior 11.2         2666
                                   100.0 ( 91.0 :   9.0) Umko 1.1 SSE42           2633
                                   100.0 ( 94.5 :   5.5) Loop 13.6/2007           2615
                                   100.0 ( 91.0 :   9.0) Jonny 4.00 32b           2614
                                   100.0 ( 89.5 :  10.5) Twisted Logic 20100131x  2600
                                   100.0 ( 90.0 :  10.0) Crafty 23.3 JA           2600
                                   100.0 ( 91.0 :   9.0) Spike 1.2 Turin 32b      2582
                                   100.0 ( 98.0 :   2.0) Crafty 23.1 JA           2547

[noctiferus]

Sometimes_They_Come_Back

[Rolf]

would like to dismantle his first intrigue

Thanks for the listing by Ingo where many progs appeared in at least two versions. Which is what I was saying. R4 remains while others change and tune. H1.4 wasnt stronger than R4 yet. But then it got a cure and it became lame 50 points stronger. Again - while R4 wasnt changed at all.

You could also take the usual results in tournaments. THERE to everybody's surprise R4 is winning all along against everybody. Reason? It was also prepared for a tournament and forcedly all the competition were confused.

But having said that, I must insist that the listing by Ingo isnt holding water at all because Ingo is too shy to prove his claimed results with the gamescores. A novelty in the long 500 years of chess history when the gamescores had been saved.

It's still interesting. We have programmers without a name and address. We now have tournaments without gamescores.

And we have a figure called Houdini who has the talent to communicate with spiritists out of the closet of his coffin - after his death!

It was always a pleasure for me when I could present this magic hero of his times. Look at the following picture, where I am quoted as Rolfe which is still my true name as artist.

http://de.wikipedia.org/w/index.php?tit ... 0527004343

[Dr.Wael Deeb]

Sometimes_They_Come_Back

Rolf is Rybka's team first ranked hellhound and he's here for a reason after the second ranked one lost the debate after the appearance of Fabien and left the forum AFAIK....
Dr.D