- Joined
- 16 Feb 07
- Moves
- 27653
05 Mar 12
Originally posted by watchyourbackrankIts worth remembering that a couple of games is not significant - that's why Zygalski does 20 games. There are plenty of GM games that have 90+% matchup (top three choices). Heck, one of my old OTB games from the 1980's was over 90% when I analyzed it. Its consistently being over 90% that's suspicious.
jesus christ.
either there's something wrong with the analysis method
or else (almost) everyone is playing like chess programs
06 Mar 12
Originally posted by ZygalskiAre you grading on the curve system or the regular system and are you using
Thanks for agreeing to allow the analysis to be published.
If you want ply-by-ply analysis rather than just the cumulative game results then please let me know.
This analysis was done using the same methods that RHP Game Mods used to use before they quit their roles.
Many hundreds of benchmark games have been analysed to see just how engine-like the ve ...[text shortened]... p 2 Match: 645/767 ( 84.1% )
Top 3 Match: 704/767 ( 91.8% )
Top 4 Match: 728/767 ( 94.9% )
the 2rd, 3rd, and 4th best move or just the best move in your decision?
Originally posted by ZygalskiGiven these statistics, RJHinds, I think it would be interesting to know if you are prepared to swear, in the name of Jesus Christ, that you have never used an engine during any competitive game on the RHP site.
Thanks for agreeing to allow the analysis to be published.
If you want ply-by-ply analysis rather than just the cumulative game results then please let me know.
This analysis was done using the same methods that RHP Game Mods used to use before they quit their roles.
Many hundreds of benchmark games have been analysed to see just how engine-like the ve ...[text shortened]... p 2 Match: 645/767 ( 84.1% )
Top 3 Match: 704/767 ( 91.8% )
Top 4 Match: 728/767 ( 94.9% )
.
06 Mar 12
1 edit
Originally posted by ZygalskiInterestingly your final 20 game matchups are almost identical to what i got yet some of the individual games differ quite significantly. Take PA82 results for instance, you got -
Thanks for agreeing to allow the analysis to be published.
If you want ply-by-ply analysis rather than just the cumulative game results then please let me know.
This analysis was done using the same methods that RHP Game Mods used to use before they quit their roles.
Many hundreds of benchmark games have been analysed to see just how engine-like the ve p 2 Match: 645/767 ( 84.1% )
Top 3 Match: 704/767 ( 91.8% )
Top 4 Match: 728/767 ( 94.9% )
{ Top 1 Match: 10/23 ( 43.5% )
{ Top 2 Match: 17/23 ( 73.9% )
{ Top 3 Match: 19/23 ( 82.6% )
{ Top 4 Match: 20/23 ( 87.0% )
Which is all legit, and i got -
{ Top 1 Match: 11/22 ( 50% )
{ Top 2 Match: 19/22 ( 86.4% )
{ Top 3 Match: 22/22 ( 100% )
{ Top 4 Match: 22/22 ( 100% )
That's quite a difference.
- Joined
- 16 Feb 07
- Moves
- 27653
06 Mar 12
Originally posted by Proper KnobWhich engine did you use? Zygalski used Houdini (its mentioned in his post with all the matchups). There is significant variation in results with different engines. Again, that's part of the reason 20 games is necessary.
Interestingly your final 20 game matchups are almost identical to what i got yet some of the individual games differ quite significantly. Take PA82 results for instance, you got ...
06 Mar 12
Originally posted by Proper KnobEven though the % changes look big, it's only a couple of moves here & there.
Interestingly your final 20 game matchups are almost identical to what i got yet some of the individual games differ quite significantly. Take PA82 results for instance, you got -
{ Top 1 Match: 10/23 ( 43.5% )
{ Top 2 Match: 17/23 ( 73.9% )
{ Top 3 Match: 19/23 ( 82.6% )
{ Top 4 Match: 20/23 ( 87.0% )
Which is all legit, and i got -
{ Top 1 ...[text shortened]... % )
{ Top 3 Match: 22/22 ( 100% )
{ Top 4 Match: 22/22 ( 100% )
That's quite a difference.
Here's some more analysis of that particular game:
Original results:
{ White: PA82 }
{ Top 1 Match: 10/23 ( 43.5% )
{ Top 2 Match: 17/23 ( 73.9% )
{ Top 3 Match: 19/23 ( 82.6% )
{ Top 4 Match: 20/23 ( 87.0% )
{ Black: RJHinds }
{ Top 1 Match: 18/23 ( 78.3% )
{ Top 2 Match: 21/23 ( 91.3% )
{ Top 3 Match: 21/23 ( 91.3% )
{ Top 4 Match: 22/23 ( 95.7% )
Now an interesting one, which proves that small background fluctuations can affect the RAM.
Ran again under exactly the same settings with the same engine on the same system:
{ White: PA82 }
{ Top 1 Match: 10/23 ( 43.5% )
{ Top 2 Match: 17/23 ( 73.9% )
{ Top 3 Match: 18/23 ( 78.3% )
{ Top 4 Match: 18/23 ( 78.3% )
{ Black: RJHinds }
{ Top 1 Match: 19/23 ( 82.6% )
{ Top 2 Match: 19/23 ( 82.6% )
{ Top 3 Match: 20/23 ( 87.0% )
{ Top 4 Match: 20/23 ( 87.0% )
Same settings again (30Sec per ply, 512MB Hash, Max Depth 20 ply) on the same PC Using Stockfish 1.9.1 JA 4xCPU:
{ White: PA82 }
{ Top 1 Match: 9/23 ( 39.1% )
{ Top 2 Match: 19/23 ( 82.6% )
{ Top 3 Match: 21/23 ( 91.3% )
{ Top 4 Match: 21/23 ( 91.3% )
{ Black: RJHinds }
{ Top 1 Match: 17/23 ( 73.9% )
{ Top 2 Match: 21/23 ( 91.3% )
{ Top 3 Match: 21/23 ( 91.3% )
{ Top 4 Match: 21/23 ( 91.3% )
Originally posted by watchyourbackrankFor RJHinds, quite a bit, I would suggest, judging by his comments in the Spirituality Forum. Let's see if he responds positively.
what difference would that make?
I have no idea whether he uses an engine or not, but simply to claim the truth in the name of his Saviour should present no problem.
Originally posted by Rank outsiderProper Knob asked me a similiar question in a messege he sent to me and I said, "No". He did not believe me and I doubt you will either. Jesus said,
For RJHinds, quite a bit, I would suggest, judging by his comments in the Spirituality Forum. Let's see if he responds positively.
I have no idea whether he uses an engine or not, but simply to claim the truth in the name of his Saviour should present no problem.
“Again you have heard that it was said to those of old, ‘You shall not swear falsely, but shall perform your oaths to the Lord.’ But I say to you, do not swear at all: neither by heaven, for it is God’s throne; nor by the earth, for it is His footstool; nor by Jerusalem, for it is the city of the great King. Nor shall you swear by your head, because you cannot make one hair white or black. But let your ‘Yes’ be ‘Yes,’ and your ‘No,’ ‘No.’ For whatever is more than these is from the evil one.
(Matthew 5:33-37 NKJV)
- Joined
- 06 Aug 07
- Moves
- 8299
06 Mar 12
Originally posted by watchyourbackrankBe careful using databases. I find that they are useful for looking at lines and analyzing them yourself. However, I do not place a lot of faith in the win/loss/draw statistics since some of the games in the data base are relatively weaker players and you can't get a good indication of the strength of a line by those statistics alone. I find NIC the best resource. I haven't used chessbase. NIC is nice since there is background text explaining the reasoning behind lines to some extent. memorizing lines is useless unless you understand why a move is played.
which online database is most up-to-date?