Hi All,
Many or most of you may know that Mac users have relatively limited options for chess engines. Fortunately Stockfish is available for Mac (I also bought a Mac copy of Deep Shredder 12 a while back). My questions follow:
1. Does anybody know if Stockfish plays better/worse on PC vs. Mac (with equivalent settings)? Is there any reason to think that chess engines running through Mac OS might be inferior? Superior? --> For instance would Stockfish for Mac out play Stockfish for PC or vice versa?
If there's a discernible large difference, I might switch back to PC next time I buy a computer (approx September to December 2015).
2. Are the ratings posted on the lists verified against humans from time to time? For instance, would Stockfish (Elo= 3311) score 93% against Magnus Carlsen (Elo = 2863) at 40/40 time controls? (This would be dictated by the 448 point rating difference: ExpectedScoreA = 1/ (1 + 10^((RatingA-RatingB)/400)) --> i.e. ExpectedScoreStockfish= 1 / (1 + 10^(448/400)) --> 92.95 approx. )
Cheers,
Jordan
Mac vs. PC vs. humans some noob questions
- Kirill Kryukov
- Site Admin
- Posts: 7399
- Joined: Sun Dec 18, 2005 9:58 am
- Sign-up code: 0
- Location: Mishima, Japan
- Contact:
Re: Mac vs. PC vs. humans some noob questions
No idea, but I'd be surprized if there is any difference between same engine on PC vs Mac, on equal hardware. However you may consider that you can probably buy a more powerful PC for the same money, compared to Mac. Stockfish on PC will probably outplay Stockfish on equally priced Mac.JWILD wrote:1. Does anybody know if Stockfish plays better/worse on PC vs. Mac (with equivalent settings)? Is there any reason to think that chess engines running through Mac OS might be inferior? Superior? --> For instance would Stockfish for Mac out play Stockfish for PC or vice versa?
If there's a discernible large difference, I might switch back to PC next time I buy a computer (approx September to December 2015).
No. Although there were attempts to do that in the past (and CCRL ratings are linked to human ratings through a long chain of calibrations), I think current ratings are not comparable with human ratings. For top engines the error could be in hundreds of points now, comparing to human ratings.JWILD wrote:2. Are the ratings posted on the lists verified against humans from time to time?
KCEC | EGTB Online | 3x3 Chess | 3x4 Chess | 4x4 Chess | Longest Checkmates | EGTB Test Suite | Opening Sampler | EGTB Bounty | NULP
Re: Mac vs. PC vs. humans some noob questions
Thanks, Kirill. This is good to know and makes sense. I am leaning toward getting a PC, partially because of the better-spec-per-dollar (as well as more chess engines available, and other non-chess related factors).Kirill Kryukov wrote:No idea, but I'd be surprized if there is any difference between same engine on PC vs Mac, on equal hardware. However you may consider that you can probably buy a more powerful PC for the same money, compared to Mac. Stockfish on PC will probably outplay Stockfish on equally priced Mac.JWILD wrote:1. Does anybody know if Stockfish plays better/worse on PC vs. Mac (with equivalent settings)? Is there any reason to think that chess engines running through Mac OS might be inferior? Superior? --> For instance would Stockfish for Mac out play Stockfish for PC or vice versa?
If there's a discernible large difference, I might switch back to PC next time I buy a computer (approx September to December 2015).
Interesting. It probably makes sense to keep the CCRL ratings related to human ratings. I'm glad their is calibration with this.Kirill Kryukov wrote:[quote="JWILD']No. Although there were attempts to do that in the past (and CCRL ratings are linked to human ratings through a long chain of calibrations), I think current ratings are not comparable with human ratings. For top engines the error could be in hundreds of points now, comparing to human ratings.JWILD wrote:2. Are the ratings posted on the lists verified against humans from time to time?
Also:
In which direction would you say the error for top engines is? Carlsen is currently rated FIDE 2863. Stockfish is currently rated CCRL 3314, so theres a 451 point difference.
Do you thinks this 451 over or under estimates the real performance difference?
In other words:
A: Are top engines even stronger in comparison to top humans than their CCRL rating would suggest?
Or
B: Perhaps the world's top humans could put up a better fight against top engines than suggested by a 450 point rating difference?
I think you meant "A," but just wanted to check.
- Kirill Kryukov
- Site Admin
- Posts: 7399
- Joined: Sun Dec 18, 2005 9:58 am
- Sign-up code: 0
- Location: Mishima, Japan
- Contact:
Re: Mac vs. PC vs. humans some noob questions
No, I did not mean 'A'. I honestly have no idea. Your guess is as good as any.JWILD wrote:In which direction would you say the error for top engines is? Carlsen is currently rated FIDE 2863. Stockfish is currently rated CCRL 3314, so theres a 451 point difference.
Do you thinks this 451 over or under estimates the real performance difference?
In other words:
A: Are top engines even stronger in comparison to top humans than their CCRL rating would suggest?
Or
B: Perhaps the world's top humans could put up a better fight against top engines than suggested by a 450 point rating difference?
I think you meant "A," but just wanted to check.
It would be nice if someone could organize some human-computer games under controlled conditions, especially with strong human players. However it's probably very unrewarding for human player, so it could be difficult.
KCEC | EGTB Online | 3x3 Chess | 3x4 Chess | 4x4 Chess | Longest Checkmates | EGTB Test Suite | Opening Sampler | EGTB Bounty | NULP