noobpwnftw wrote: ↑Tue Feb 23, 2021 7:36 pm
1) Rybka had clear improvements over any other chess engine at the time, while FF2 doesn't.
2) To this day, many people still acknowledge Vas's contributions despite all the controversies, while ASilver has never contributed one single game to Stockfish or Leela training.
I agree.
Rybka may or may not have wrongly used GPLed code, but it certainly was much more original, innovative and valuable for computer chess than the Fat Fritz nets are.
noobpwnftw wrote: ↑Tue Feb 23, 2021 7:36 pm
1) Rybka had clear improvements over any other chess engine at the time, while FF2 doesn't.
2) To this day, many people still acknowledge Vas's contributions despite all the controversies, while ASilver has never contributed one single game to Stockfish or Leela training.
I agree.
Rybka may or may not have wrongly used GPLed code, but it certainly was much more original, innovative and valuable for computer chess than the Fat Fritz nets are.
Agree, Rybka was far above any other engine at the time - clearly Vas was doing something different that no one else was doing, there was much originality in his work ..
noobpwnftw wrote: ↑Tue Feb 23, 2021 7:36 pm
1) Rybka had clear improvements over any other chess engine at the time, while FF2 doesn't.
2) To this day, many people still acknowledge Vas's contributions despite all the controversies, while ASilver has never contributed one single game to Stockfish or Leela training.
I agree.
Rybka may or may not have wrongly used GPLed code, but it certainly was much more original, innovative and valuable for computer chess than the Fat Fritz nets are.
Agree, Rybka was far above any other engine at the time - clearly Vas was doing something different that no one else was doing, there was much originality in his work ..
M ANSARI wrote: ↑Tue Feb 23, 2021 8:08 pm
I really don't get the controversy regarding CB and Fat Fritz as everything is very clearly stated.
I have no idea how you can think that. They have been extremely dishonest about FF (1&2) from the start. Any admissions they've made have been forced out of them by the public.
It may be clear nowto us (on talkchess) that it's just a clone, but it was certainly not clear to the public at the start or even now.
Tord Romstad
@tordr
Feb 13 I don't have time to listen to the actual interview, but the introduction is wrong that I just have no words:
Tord Romstad
@tordr Albert Silver "came across a new neural network technology from Japan"? Seriously?
Tord Romstad
@tordr And then there is this, in the list of interview questions: "How did you overcome the moment of starting almost completely from scratch with the Fat Fritz 2.0 development?" How on Earth is "starting almost completely from scratch" an accurate description?
What most people forget that Hydra Chess Project was also canceled because Rybka became so much more powerful that Hydra that ran on a 32-node Intel Xeons with a Xilinx FPGA accelerator card cluster, with a total of 64 gigabytes of RAM. It evaluates about 200,000,000 chess positions per second, roughly the same as the 1997 Deep Blue which defeated Garry Kasparov, but with several times more overall computing power, could not compete with Rybka so the Pal Group stopped funding the project. This is from the Hydra website courtesy of the Wayback Website: HYDRA, under the patronage of PAL GROUP, originally appeared in four versions: Orthus, Chimera, Scylla and then the final HYDRA version-the strongest of them all. Multiple computers each with its own FPGA acting as chess coprocessors, enables HDYRA to search enormous numbers of positions per second, making each processor more than ten times than an unaided computer.
It runs on a 64CL and evaluates about 200,000,000 chess positions per second, roughly the same as the much older Deep Blue, but with several times more overall computing power. Whilst FPGAs generally have a lower performance level than ASIC chips, Moore’s Law allows modern-day FPGAs to run about as fast as the older ASICs used for Deep Blue.
Clearly nobody could match Rybka no matter how much money was used for custom supercomputer hardware back in 2008.
I think for a while, everyone thought that chess engine progress will improve ... but it will improve mostly due to the improvement in hardware running bean counter engines. Rybka changed that thinking and it made clear that the software side of chess engines is where the major gain in ELO would come from. This has been made even more true with the giant leap that Lc0 gave and later on the inclusion of NN to almost any chess engine. I think you could say with a straight face that Rybka with revolutionary system of using millions of rapid games to tune evaluation as well as some Monte Carlo experimentation was a primitive form of NN. The strength of these new NN chess engines at low depth is mind boggling. I would not be surprised if today's top NN chess engine would beat Deep Blue with only an Apple watch as hardware. A single core SF 13 against DB would probably be an incredibly lopsided affair to SF. It is too bad that If you go through games of DB with SF 13 or Lc0 ... there are so many inaccuracies and weak moves that just makes me think that DB is probably 400 or more ELO points weaker. So software turned out to be a much bigger factor in moving engine ELO forward than hardware would ever have done.
towforce wrote: ↑Tue Feb 23, 2021 10:50 pmOne BIG mistake that the ICGA made is that the rules give the tournament director the right, but not the obligation, to ask contenders for source code: it was obvious to me when I entered the 2000 WCCC (as the Crafty operator) - a decade before this controversy - that the tournament director should be OBLIGED to collect source code from each competitor. When the source is modified between rounds, the changed source should also be handed over. There's no earthly reason to not do this!!!
Given the usual level of integrity in the computer chess world, handing over your soure code to the tournament director is not much different from publishing it on the internet.
M ANSARI wrote: ↑Sun Feb 28, 2021 3:22 pm
It is too bad that If you go through games of DB with SF 13 or Lc0 ... there are so many inaccuracies and weak moves that just makes me think that DB is probably 400 or more ELO points weaker. So software turned out to be a much bigger factor in moving engine ELO forward than hardware would ever have done.
In my view there is no evidence that DB was stronger than 2800 Elo, so 400 Elo seems conservative.
From the few games we have, DB seems to be about as strong (or as weak? ) as Kasparov was at the time, and Kasparov could probably have learned to adjust his playing style to it if he had played more games.
) as Kasparov was at the time, and Kasparov could probably have learned to adjust his playing style to it if he had played more games.