Originally posted by BigDoggProblemI haven't / didn't overstep any mark or make any claims. I merely gave my thoughts which I stand by
Which would you say there are more of - strong human players, or engines stronger than human players?
Do you acknowledge that some humans are tempted to cheat by using chess engines in their games?
OK, so st40 overstepped in claiming you are someone you're not. But the frustration of playing engine users makes some people [understandably] paranoid.
Originally posted by greenpawn34In your blogs, you're always bringing up the importance of developing pieces; so it's interesting to see you have three undeveloped pieces. I wonder what kind of advice you'd give to yourself here.
So now I'm down to 20% first move. My confidence is shattered...thank you all.
I have an ongoing game v Scottish Geek and have no qualms about talking
an ongoing game because I cannot see a way of bring up the reserves
so am going to go Qh3+ Qg3+ Qh3+ etc and etc.. (I'm Black)
[fen]rnb5/ppp2k1p/3p1p2/8/3Pp1P1/2P2Pq1/P1P5/R2Q1RK1 w - - 0 20[/f ...[text shortened]... but I can confirm the vast majority don't.
Look at the blog, look at my games. Blunders galore.
Originally posted by BigDoggProblemI give up on the analysis. It's clear to me that there is no validity to it but there is no point arguing. I was a statistician and computer scientist and I have written a chess program, but that was some time ago.
Which would you say there are more of - strong human players, or engines stronger than human players?
Do you acknowledge that some humans are tempted to cheat by using chess engines in their games?
OK, so st40 overstepped in claiming you are someone you're not. But the frustration of playing engine users makes some people [understandably] paranoid.
I don't see the relevance of the number of strong human players vs engines. I don't even think that is meaningful. Engines are better - way better - if you run them for a reasonable time on a fast processor - than any human who has ever lived or will ever live. No question about that now. However, if I play Stockfish and only give it one second per move, I can beat it (sometimes - although even then not consistently :-( but bet a super GM could). All irrelevant, Who cares? I play for fun - why else is anyone here? There is no money at stake, no FIDE ratings. Why on Earth would someone want to "cheat"? They gain nothing. Even if they do "cheat", who cares? You are playing the entity you are playing. Maybe it's 10 GMs who decided to have some fun. Maybe it's a dope with an engine. Maybe (in almost every case in my estimation) it's just someone who enjoys playing chess and may be better - maybe much better - than you.
Originally posted by SkippyJoeWhat about the "tempted" question? Any thoughts on that?
I give up on the analysis. It's clear to me that there is no validity to it but there is no point arguing. I was a statistician and computer scientist and I have written a chess program, but that was some time ago.
I don't see the relevance of the number of strong human players vs engines. I don't even think that is meaningful. Engines are better - way ...[text shortened]... on) it's just someone who enjoys playing chess and may be better - maybe much better - than you.
Originally posted by SkippyJoeYet again, your total lack of understanding of what I previously posted is obvious for all to see.
I give up on the analysis. It's clear to me that there is no validity to it but there is no point arguing. I was a statistician and computer scientist and I have written a chess program, but that was some time ago.
I don't see the relevance of the number of strong human players vs engines. I don't even think that is meaningful. Engines are better - way ...[text shortened]... on) it's just someone who enjoys playing chess and may be better - maybe much better - than you.
Where is my 1 second per half-move analysis?
Who said that human players are "better" than engines?
Why is there no validity in comparing a suspected cheats chosen moves with those chosen by an engine in the same positions?
How would you look for suspected engine users if you wanted to?
I showed that if you give an engine 20 minutes or 80 to analyse a game, even with different engines, the results are very similar.
Your argument is a bunch of strawmen & when you can't respond adequately you just run away.
Whilst most people play for fun, there is a thread running about the lack of titled players on this site & offering free membership . What could a FIDE titled player possibly gain from playing anyone on the first couple of pages of highest rateds on this site that they couldn't get from training against an engine?
The higher rated tournaments & annual championships have long been a joke (I know because I analysed some of them back when idiotic engine users were banned) & any serious OTB players would avoid playing on this site like the plague since engine use is tacitly allowed.
Oh, and by the way, several FIDE titled players were banned from here (and other sites too) for blatant engine use several years ago. Why do they do it? Do some googling & send them an email.
Originally posted by ZygalskiYour arguments make no sense. You seem to think that the fact that super GMs have a lower percentage of moves in the top 3 than 30s Houdini picks means that players with a higher percentage must be "cheating". This is nonsense. It just reflects the fact that a 30s Houdini analysis is not very strong and much weaker (but very good) players will have much higher match rates against 30s Houdini that a super GM (or the 3600s per move engines I analyzed).
Yet again, your total lack of understanding of what I previously posted is obvious for all to see.
Where is my 1 second per half-move analysis?
Who said that human players are "better" than engines?
Why is there no validity in comparing a suspected cheats chosen moves with those chosen by an engine in the same positions?
How would [b]you look fo ...[text shortened]... latant engine use several years ago. Why do they do it? Do some googling & send them an email.[/b]
- Joined
- 08 Apr 09
- Moves
- 20189
Originally posted by SkippyJoe1. The analysis that Zygalski and other use to objectify cheating suspicions is NOT a measure of playing strength
Your arguments make no sense. You seem to think that the fact that super GMs have a lower percentage of moves in the top 3 than 30s Houdini picks means that players with a higher percentage must be "cheating". This is nonsense. It just reflects the fact that a 30s Houdini analysis is not very strong and much weaker (but very good) players will have much higher match rates against 30s Houdini that a super GM (or the 3600s per move engines I analyzed).
2. Instead, this analysis measures the statistical correlation between a players' moves and a strong engine's moves. (*)
3. Zygalski and others have shown that these correlations are bounded above, i.e. no human player consistently has a higher matchup than specific values.
4. As a result of 3, it is statistically very unlikely that a human player consistently scores way above these limits. Consequently, a different explanation for these moves must be found. (**)
(*) This correlation is relevant, because higher correlations are expected for people using or supported by engines to decide on moves.
(**) Most likely explanation: these players have been cheating by using a chess engine.
30 sec. per move is more than strong enough for this purpose to eliminate results being skewed by a weak engine. Keep in mind that the value is only in the statistics over a large number of games/moves, not in single moves or single games. It means that 30s Houdini doesn't need to agree with 3600s Houdini on all moves exactly to be useful. Feel free to prove the contrary.