|Jun-10-07|| ||Timex: Wow, this is weird. Two strong computers play against each other and one gets whitewashed. It reminds me of Kramnik-Fritz where Kramnik lost the match without winning a game.|
Lately, computer software has been becoming better and better. One day though, a human will match up to their strength and win a match against a strong computer in the end, i hope.
|Jun-10-07|| ||ongyj: What I don't understand is the difference in processing power. From chessbase.com, deep fritz is looking at 12 million moves / second whereas deep junior is 25 million moves / second. If the intention was to see which was the better programme, why not let both be equipped with the same processing power?|
|Jun-10-07|| ||Magician of Riga: Each team was allowed to use the best hardware they could acquire by the time of the match. The Deep Junior team lucked out and found a machine with twice the processing power of the Deep Fritz team's machine (16 cores vs 8 cores respectively). Obviously this was a huge advantage.|
|Jun-10-07|| ||chessmoron: Rybka has a 16-core engine as well but was a borrowed one. I hope Rybka kill Deep Junior between the Mexico championship if FIDE/chessbase agrees on it.|
|Jun-10-07|| ||THE pawn: You know well, <chessmoron>, that chessbase will never accept a match because they know Rybka will kill their top engine.|
|Jun-10-07|| ||djmercury: <Each team was allowed to use the best hardware they could acquire by the time of the match. The Deep Junior team lucked out and found a machine with twice the processing power of the Deep Fritz team's machine (16 cores vs 8 cores respectively). Obviously this was a huge advantage.>
And that made this match even more worthless, although some games were rather interesting from the theory point of view.|
|Jun-11-07|| ||ongyj: I'm new to computer chess, but which is the best software written so far,and why is it 'best'?|
|Jun-11-07|| ||Karpova: <ongyj: I'm new to computer chess, but which is the best software written so far,and why is it 'best'?>|
Rybka is the best. Beats every other computer in tests (except for Hydra which it seldomly plays due to known reasons (Freestyle tournaments are the only chance for both to play each other most of the time). Chessbase products still get more attention though...
It's not known (and Raijlich probably won't tell anybody) how Rybka decides on moves and how he calculates the evaluations for positions.
|Jun-11-07|| ||ongyj: <Karpova> Thanks alot. Now at least I know that Rybka has beaten all but Hydea in tests and Raijilich is the programmer of Rybka^Ô^|
|Jun-11-07|| ||spirit: Rybka vs Deep Junior!!? Somewhat loopsided|
|Jun-11-07|| ||spirit: <Timex> <One day though, a human will match up to their strength and win a match against a strong computer in the end, i hope.> Your statement sounds very very wierd, but me think you are right...these super-machines are missing something,one of them will be embarrased soon...|
|Jun-12-07|| ||Karpova: <ongyi> sorry if you expected more detailed information.
Here's a link to an interesting article on Rybka written by none other than our RandomVisitor chessforum :
|Jun-13-07|| ||Knight13: What now, Fritz users! DEEP JUNIOR RULES!|
|Jun-15-07|| ||s4life: <Knight13: What now, Fritz users! DEEP JUNIOR RULES!>|
Makes you wonder how much worse would have been the final score of a hypothetical Kramnik vs Deep Junior, or even worse a Kramnik vs Rybka... Mickey Adams wouldn't feel so bad afterwards methinks.
|Jun-17-07|| ||Voxation: Whats the difference between Fritz and Deep Fritz?|
|Jun-17-07|| ||whatthefat: Deep Fritz runs on multiple processors.|
|Jun-17-07|| ||erli: Deep Junior won because of superior processing power, how is that a win?|
|Jun-17-07|| ||Overgod: Rybka is not impossible to beat. I have achieved many draws and a couple of wins against Rybka 2.2 with a slight handicap (Rybka has 3 min, I have 15 min or something)...Sure you are going to lose a lot of games, but if you sit down and concentrate and are a strong player, you can draw Rybka. The problem with a lot of chess computers is that they take too many risks against rybka...too agressive and rybka punishes...I very rarely get into open positions against rybka. Although I have held my own against it in some very wild, tactical games. I have a few if any of you are interested...I have 1 game, where i believe i was playing at 2900+ level (very complex tactical battle), but in the end it was a draw. I am quite proud of it. Especially considering I went head to head in tactics against Rybka!|
If anybody is interested, my email is email@example.com
I have many games against rybka...
|Jun-17-07|| ||argishti: Seriously, how is this a match. when we already knew who was gonna win this thing?|
|Jun-17-07|| ||Tomlinsky: <erli: Deep Junior won because of superior processing power, how is that a win?> |
Personally, I think that the answer simply runs something like as follows:
From a business standpoint Chessbase has equal interest in healthy sales of both products Fritz and Junior in their various fruity flavours. Setting up a 'challenge' between two of your own products could be a risky business as you can potentially damage sales of one while not necessarily promoting the other in justification.
Now, running on equally powered hardware one could reasonably expect Junior when playing Fritz to 'prove' itself to be the superior product, as far as this contest is concerned anyway, so what do we do? Easy. We run the likely 'inferior' product on inferior hardware so as to leave the result completely in the air to anyone that investigates further than the advertising that can be produced from the match and used in the near future.
Thus, the product that was likely to have won anyway wins but in such a way that our 'loser' product manages to keep its integrity with an excuse clause already built into the conditions of our promotional match pairing of said products.
As organiser and promoter of the contest Chessbase had free reign to set any and all conditions, wether they be equal to both parties or weighted in favour of either, before it was ever even mentioned to the public. It was a product promotion scheme with a good deal of thought behind what were the required objectives from the point of view of Chessbase as a company. All of the what's, if's and why's have a reason.
So, the answer to your question <erli>, feel free to call me a cynic, is that other than giving some electrons an outing for a few days it isn't a win. Unless you have shares in Chessbase or sell chess software for a living of course.
|Jun-18-07|| ||Karpova: <Tomlinsky>
Deep Fritz 10 seems to be better at classical time controls...
Deep Fritz 10 on 5th place (2886)
Deep Junior 10 on place 13 (2825)
Deep Fritz 10 scored against Deep Junior 10:
+17 =23 -10
No Deep Fritz 10 yet (just Deep Fritz 9)
|Jun-18-07|| ||acirce: <As organiser and promoter of the contest Chessbase had free reign to set any and all conditions, wether they be equal to both parties or weighted in favour of either, before it was ever even mentioned to the public.>|
How was Chessbase the organiser? I highly doubt that they had any such "free reign" anyway, as far as I know it was simply up to the respective teams themselves to get the best hardware as usual in these engine-engine contests. We don't have to assume a conspiracy just because Chessbase sells the programs.
Also, it appears people give the difference in hardware in this match way too much significance. It wasn't of the magnitude to have a big impact. Fritz is clearly the better program and, I think, would have won a longer match even with this slight handicap -- a 6-game match is obviously much too short to be statistically relevant.
|Jun-18-07|| ||Tomlinsky: <acirce/Karpova> Arrange the advertising of your product without at least consultating your major global supplier of said product and see what happens. My point was, and still is, that a few questions have been left hanging (that's why people are asking and answering them) and that there are reasons. They arne't oversights or mistakes.|
So, maybe Fritz would win in a longer match with longer time controls and more games and similar hardware. But this wasn't a longer match with longer time controls. Why is that? Junior had hardware with twice as many processors and able to calculate over twice as many positions per second (according to Chessbase). Why is that? So we won't know really until they do will we? Why is that? ;)
|Aug-05-07|| ||ongyj: Just being curious here, are there any tournaments with intentions to find out "the best" chess softwares around? If there's one coming, what would be the correct participating conditions?|
For now, I can think of 2. 1st is obviously, the processing power. 2nd is the sharing of opening books. With these being equal, we can then say "for sure" whether programme A aces over programme B or vice versa...But I'm sure there would be teams of programmers who can't bear to see their software lose, so I guess it may be hard to get every famous software to compete on equal grounds...
|May-26-09|| ||WhiteRook48: what now for Deep Junior users? Rybka rules!|