Back to forum

Ben Milton    (2009-01-05)

I hereby challenge all to 2 games one as black and one as white, to whoever believes they can beat Rybka 3 using tuned Perfect 15 book and depth 23 after opening. This is ofcourse to win my E-points. Any one interested at all?

Thibault de Vassal    (2009-01-05 16:34:12)
Rybka 3 alone ?

You mean Rybka 3 playing by itself ? Could be a very interesting test match...

Wayne Lowrance    (2009-01-05 17:51:35)

I am not sure if I understand your conditions. do you mean moves by your opponent must be made at depth 23 (don't know how you propose enforce that). Anyhow I have not time for such a match. Wayne

Ben Milton    (2009-01-05 18:30:54)

No, I meant I will be using the conditions above, and the opponent can do whatever in their power to beat me.

Thibault de Vassal    (2009-01-05 18:34:24)
Depth 23

It would have no sense to propose such a challenge. No, only Rybka moves (played by Ben) will be limited to depth 23.

I think I would accept the challenge (for at least 100 Epoints) if the full analysis by Rybka is posted after the end of the game :)

Normajean Yates    (2009-01-05 19:12:44)
Is BenMilton thinking of move mirroring?

Then he will lose both games, because move mirroring is illegal here :)

Thibault de Vassal    (2009-01-05 20:47:24)
I don't think so...

Of course in such a match the full analysis by the engine should be published at the end of the games.

Marc Lacrosse    (2009-01-05 23:17:16)
First flaw is in the book

There are many published lines crushing perfect15.
I am not sure that sedat already corrected all the known ones for the coming perfect16.
Either you publish your ctg at the beginning of the games or you cannot affirm that you run a "modified" perfect15 without providing proofs that you are not playing yourself during the opening phase
So i fear it's a flawed challenge
Apart from this I suppose that any player among top-50 here is ready to play against a pure R3. If we were not ready to accept this, then this would mean that correspondence chess has come to its end.


i must admit that in case your opponent plays postman chess with his engines, times have become tougher for those who try to demonstrate something, and much tougher since rybka 3 appeared. But so far there are still players who keep achieving 70% against opponents who probably all use top engines ...

Ben Milton    (2009-01-06 03:22:29)

Im not sure what exactly move mirroring is. Also as i said my opening book is perfect 15 book BUT it is tuned. Thibault Id have put 100 E-points on the line if i had it. At the moment i only have 12. Does anyone believe they can beat the conditions i proposed and are willing to have a game?

Wayne Lowrance    (2009-01-06 05:11:28)

Well Ben as I said I am not inclined to accept your challenge, but if It could be proved your using only r3 + your best book, you would not have much success against strong centaurs Wayne

Don Groves    (2009-01-06 05:18:20)
I don't understand why a challenge...

Why don't you simply enter the Gold or Silver waiting list and play? The methods you propose are entirely legal here and don't require advance notification.

It would be interesting to see Rybka's analysis published as Thibault suggested but that goes for any worthy game played here.

Ben Milton    (2009-01-06 11:08:51)

Would someone please tell how to be a strong centaur player? I have been playing online computer assisted chess for 2 years now and still have not learned how to be stronger than rybka 3 by itslef. I use Fritz11 GUI. Any reccommendations? would be appreciated

Thibault de Vassal    (2009-01-06 14:14:28)
How to beat Rybka 3 ?

Hi Ben, the main -enormous- advantage you have over Rybka 3 is of course that you know by advance what she may play at least in "some" cases. At correspondence chess, you have to create a trap according to the horizon's effet (don't remember if this is the correct term) or analysis depth. Well, it may represent several weeks of analysis though, to understand such engine's weaknesses, then to incitate her to follow you in a good line, knowing the book she uses.

Playing against Rybka 3 in a freestyle chess tournament will be even harder, only someone who perfectly knows the engine & has a very good understanding of the game may hope to have a good score (over 60%) against the engine IMO.

Finally a good centaur would have quite good chances to win this match in my opinion.

Ben Milton    (2009-01-06 15:24:33)

Thank you very much, but the problem is even though i might know what the opponent might play, i still dont know what move to play so it leads to a trap, the best i can do is to go as deep as possible. Also i heard a few things about "IDeA" which is a tool centaurs use to go deeper but unfortunately Fritz 11 does not have that...Any how if youd be willing we can have the games and i am willing to give you all my e-points (12) if you win and if you dont i dont want any e-points in return. I jus want to see the results. How does it sound?

Thibault de Vassal    (2009-01-06 18:59:13)

That's an interesting challenge, with or without epoints. As for me, that's a pity I have no time enough yet, maybe later (in a few months).

Normajean Yates    (2009-01-08 04:48:31)
your challenge would have takers if...

if you could improve upon this:

Milton, Ben (AUS) [member # 4431]

ELO : 1597

Next page : History

Statistics for rated correspondence chess games :

Won : 0
Lost : 6
Draw : 0
Unfinished : 6

So, as thing stand, your challenge sounds trollish, if you will excuse my saying so. But I think someone has to.

I am in two minds whether to post this: I am violating the golden rule of usenet "don't feed the troll". :/

Ben Milton    (2009-01-08 05:08:32)

The reason to my rating is that my opponents were mostly human at this level and could not compare with me a centaur player and therefore i lost interest and let my games be lost with time forfeit. If you are not interested in the conditions i mentioned before, then no one forced you to play me, and there is nothing trollish about my challenge.

Thibault de Vassal    (2009-01-08 12:23:20)
No ratings...

Just an idea, this challenge could be played here in this forum, move after move, with the conditions mentioned above (& Rybka 3 analysis in real time, no need to hide it after all). 2 unrated games, 40 days + 40 days / 10 moves. What do you think ?

Tano-Urayoan Russi Roman    (2009-01-09 08:31:05)

What is the purpose of this challenge? Do you accept Thibault conditions in the post above this one? and who you will proof the engine is playing alone with book and you are not leading it as Marc Lacrosse pointed?

Thibault de Vassal    (2009-01-09 08:44:35)
Why ?

The purpose is only to show if Rybka 3 can be or not a strong correspondence chess player already, nothing more. Ben has no interest to try to cheat and it would be a big work to change the full Rybka analysis (ie. 5 first lines for each move) that would be copy-pasted here.

Tano-Urayoan Russi Roman    (2009-01-09 14:09:11)
I accept

Using Thibaults conditions.

Thibault de Vassal    (2009-01-10 17:23:24)

Great ! We have a candidate :) Ben, what do you think ?