Ronald, You are just crouching in wait to pounce on and attack anyone that does not perform to your level. Is that a demonstration of your superiority? Did you just happen to read and remember Marco's post or did you decide to search after the fact. But either way I can just imagine the smile on your face when you got the opportunity to attack someone. Out in the world that bullying behavior might get you a bloody nose. But here on the internet you can be pounded into the ground metaphorically with words and just ignore it. You just don't feel the pain do you?
IMO it is absurd to even think about this (Ronalds post) as an attack...
Then you haven't been paying much attention.
Hi Mike.
It's OK, and, BTW, it's Christmas too. The fact is, I'm too busy with projects about teaching chess at 6-11 y.o. children, and I must build an educational GUI with VB, so I read about 10% of the posts about Alpha0.
A lot of people have a lot, a lot much more knowledge than me and I'm happy when someone indicates me an error. Ronald has never been aggressive or negative with me. I think there's some tensions about the argument of AlphaZreo. Maybe we should consider we know too little.
Ronald, You are just crouching in wait to pounce on and attack anyone that does not perform to your level. Is that a demonstration of your superiority? Did you just happen to read and remember Marco's post or did you decide to search after the fact. But either way I can just imagine the smile on your face when you got the opportunity to attack someone. Out in the world that bullying behavior might get you a bloody nose. But here on the internet you can be pounded into the ground metaphorically with words and just ignore it. You just don't feel the pain do you?
Having a bad day, are we? He was kind enough to search for the exact post and kinder enough to provide the link. What part of that is an 'attack' much less 'bullying'?
"Tactics are the bricks and sticks that make up a game, but positional play is the architectural blueprint."
Rodolfo Leoni wrote: I think SF could have a better performance with better settings tough.
This is trivial to test, just play SF at 1min/move against SF at 1min/move with all the paraphernalia that it was missing, if the new one beats the old one by 30 elo, A0 would only have won by 70 elo intead of 100.
I think you're right. But I'm using my limited free time and old PC for correspondence chess. I could try something like that on next spring, IMO. It could be interesting.
Latest SF Dev + Cerebellum + Syzygy vs naked SF8, 1 min/move.
I think that is easily more than 100 Elo stronger than SF 8 by now, but maybe not when SF 8 has 64 cores at the hardware they were using. The Cerebellum and Syzygy are not really fair IMO, although a form of learning should be allowed because that is what they did too.
I'm not a mathematician you'd have to ask Tord and Joona but I prefer to think of Alpha Zero as Aleph-Null but that is just the German form of the mathematical term. Alpha Zero is more natural to English speakers. A form of infinity like ∞, where Google gets its name from. Aleph is used in the mathematical symbol, I suppose because Georg Cantor used that? Not sure.
It's impossible to emulate the same match conditions with my old i7 and only 4 GBs RAM. In theory, latest SF dev is +55 ELOs, Cerebellum could be +25 ELOs, a good hash size could be +10 ELO and tournament TCs could be another +10 ELO. But this is theory. To emulate a parallel match I'd probably have to run it at 1 hour per move or more. and we know LTCs work as an equalizer. So it could be funny to run it at 1 min/move, but useless for a comparison.
I sincerely hope this "new entry" will participate at WCCC, and that could bring some enthusiasm, some fresh air into the competition.
Rodolfo Leoni wrote: and we know LTCs work as an equalizer.
We do? Why would we know any such thing?
I think Larry Kaufman and Kai Laskos posted about it in the past year or year and a half somewhere, but I would have to look up where. Kai made experiments on limits on what Elo can be achieved. More simply, the amount of Elo you get with a doubling of time falls sharply with longer timecontrols. Vas reported 70 Elo in Rybka testing once but that is something of an upper limit, traditionally the number people used to quote was about 50.
It is also what the Google people are talking about in the short clip that Ed posted here, about the "compression of the Elo scale because of the prevalence of draws" That is why the 100 Elo difference is more impressive on 64 threads.
Last edited by Eelco de Groot on Tue Dec 26, 2017 9:32 pm, edited 1 time in total.
Debugging is twice as hard as writing the code in the first
place. Therefore, if you write the code as cleverly as possible, you
are, by definition, not smart enough to debug it.
-- Brian W. Kernighan
Ronald, You are just crouching in wait to pounce on and attack anyone that does not perform to your level. Is that a demonstration of your superiority? Did you just happen to read and remember Marco's post or did you decide to search after the fact. But either way I can just imagine the smile on your face when you got the opportunity to attack someone. Out in the world that bullying behavior might get you a bloody nose. But here on the internet you can be pounded into the ground metaphorically with words and just ignore it. You just don't feel the pain do you?
Having a bad day, are we? He was kind enough to search for the exact post and kinder enough to provide the link. What part of that is an 'attack' much less 'bullying'?
Which is kinder?
Look here. LINK
or
You haven't been paying much attention. LINK
And I'm having a wonderful day, thank you!
If you are on a sidewalk and the covid goes beep beep
Just step aside or you might have a bit of heat
Covid covid runs through the town all day
Can the people ever change their ways
Sherwin the covid's after you
Sherwin if it catches you you're through
Eelco de Groot wrote:I think Larry Kaufman and Kai Laskos posted about it in the past year or year and a half somewhere, but I would have to look up where. Kai made experiments on limits on what Elo can be achieved. More simply, the amount of Elo you get with a doubling of time falls sharply with longer timecontrols. Vas reported 70 Elo in Rybka testing once but that is something of an upper limit, traditionally the number people used to quote was about 50.
It is also what the Google people are talking about in the short clip that Ed posted here, about the "compression of the Elo scale because of the prevalence of draws" That is why the 100 Elo difference is more impressive on 64 threads.
And they claim that A0 gets stronger with more time, so while Stockfish's strength gets stuck and doesn't improve with more time, allowing weaker engines to reach it, A0 takes off, and just beats Stockfish badlier and badlier.
Ovyron wrote:And they claim that A0 gets stronger with more time, so while Stockfish's strength gets stuck and doesn't improve with more time, allowing weaker engines to reach it, A0 takes off, and just beats Stockfish badlier and badlier.
Except that Stockfish does improve with more time and does not allow weaker engines to reach it. Or is there any solid evidence to the contrary?
Note that I am not contesting that AlphaZero improves more than Stockfish with more time. At least the AlphaZero paper suggests that it does. (My explanation for AlphaZero profiting more from having more time is that it needs that time to correct for the inevitable tactical deficiencies of its NN-based evaluation.)
syzygy wrote:......................................................
Note that I am not contesting that AlphaZero improves more than Stockfish with more time. At least the AlphaZero paper suggests that it does. (My explanation for AlphaZero profiting more from having more time is that it needs that time to correct for the inevitable tactical deficiencies of its NN-based evaluation.)
If you're right, we could then presume AlphaZero would heavily lose with blitz TCs.
And there's another doubt: what if AlphaZero played without that 4 hours self-training session?