Dear Benno,

Thanks a lot for kind comments !

About Rybka is approx. 3350 Elo and it is unrealisitic high Elo or realistic...

Sorry to say that, but really I am tired to talk over this issue....

Only comments and no actions...!!)) in other words: no more no less !
As far as I know...chess is not comment, but game !!

Once more I'd like to point out:
•SCCT MP Rating almost perfectly matches to GM’s Elo points (in Elo measuring), for details
http://www.talkchess.com/forum/viewtopic.php?t=51059&highlight=sedat

Exception that I already made some serious rating/tournament calculations (GMs vs Computers)
My challenge to GMs is still valid...!!)) but unfortunately there is no such gladiator (GM), to participate in SCCT
A little Note: I promised to give a prize: 10.000 USD (in case if there is no  +500 Elo difference)
Another little note: 20 % (2.000 USD) commission will be for that person who will find me a GM ))!!

1) In case of such event:
-GM has choice to prefer:
-To play against me (Rybka+Perfect 16 book)
or
2) To participate in SCCT Book Tournaments (Stockfish is used mainly for book players)

A little hint:
-It's better to play against me (Rybka+P16), because Rybka is approx. 200 Elo weaker than Stockfish

And last,
I hope one day.... a X GM will contact me and will agree my challenge !)

Best,
Sedat

Hello again Benno !)

On thing more,
SCCT MP Ratings (Ponder ON) is not same, but I can say:
- Very close in Elo strength to CCRL MP Rating (Ponder OFF)

For examples,

CCRL 40/4

Rybka 4.1 64-bit 4CPU Ponder OFF 3198 Elo (adapted to AMD Athlon 64 X2 4600+ 2.40 GHz)
Rybka 4.1 64-bit 1CPU Ponder OFF 3102 Elo (adapted to AMD Athlon 64 X2 4600+ 2.40 GHz)

-------------------------------------------------------------------------------

SCCT 3m+2s

Rybka 4.1 64-bit 3CPU Ponder ON 3350 Elo (adapted to i7 980X 3.33 GHz)
Rybka 4.1 64-bit Ponder OFF 1CPU is expecting to be approx. 3240 Elo (adapted to i7 980X 3.33 GHz)

More Details:
Stockfish bench 3 cores on i7 980X 3.33GHz = 4208 Kn/s
Stockfish bench 2 cores on Authlon 2.40GHz = 1350 Kn/s
Stockfish bench 1 core  on Authlon 2.40GHz = 700 Kn/s

Note that in case of doubling the processor speed (up to 8 cores):
-Mostly of the MP chess engines perform approx. 70 - 80 Elo
- Ponder ON is approx. 20 Elo points stronger than Ponder OFF
- 3 cores are estimated to be approx. 90 Elo stronger than 1 core

And after checking all the mentioned issues...the formula is here
*Note: my used conditions are faster hardware + ponder on
-Hardware speed doubling x 80 Elo = 240 Elo
-Ponder ON is approx. 2O Elo stronger than Ponder OFF

So here is the final calculation about Rybka 4.1 MP 3350 Elo
- estimated approx. 200 -250 Elo (Hardware speed difference)
-  approx. 20 Elo (Ponder ON/OFF difference)
* Total = 3100 Elo

In other words,
Rybka 4.1 1 core Ponder OFF is expecting to be approx. 3100 Elo (if I test it on my AMD 4600 2.40GHz)
But on my i7 980X 3.33GHz, Rybka 4.1 x64 1 core Ponder OFF is expecting to be approx. 3240 Elo

And last:
- CCRL Elo Ratings (like SCCT) are very close to GMs Elo ratings too !!

For more details:
http://www.sedatcanbaz.com/chess/?page_id=809

Hopes helps...

Best,
Sedat

Continuing, another calculations...

1) Standings - Only Wins (No Draw Games):

   # PLAYER                 : RATING    POINTS  PLAYED    (%)
   1 Stockfish 151014 MP    : 3842.3     687.0     766   89.7%
   2 Stockfish 170914 MP    : 3762.9     777.0     870   89.3%
   3 Komodo 8 MP            : 3730.7     821.0     976   84.1%
   4 Houdini 4 MP           : 3657.5     827.0    1093   75.7%
   5 Gull 3 MP              : 3545.6     641.0     921   69.6%
   6 Rybka 4.1 MP           : 3350.0     459.0     653   70.3%
   7 Protector 1.7.0 MP     : 3146.9     268.0     442   60.6%
   8 Chiron 2 MP            : 3141.4     341.0     605   56.4%
   9 Hannibal 1.4b MP       : 3099.9     326.0     621   52.5%
  10 Protector 1.6.0 MP     : 3075.0     109.0     355   30.7%
  11 Texel 1.04 MP          : 3051.1     301.0     618   48.7%
  12 Hiarcs 14 WCSC MP      : 3046.4     286.0     615   46.5%
  13 Senpai 1.0 MP          : 3021.2     269.0     607   44.3%
  14 Spike 1.4 MP           : 2928.3     214.0     627   34.1%
  15 Junior 13.3 MP         : 2922.2     226.0     633   35.7%
  16 Spark 1.0 MP           : 2911.8     193.0     583   33.1%
  17 Shredder 12 MP         : 2895.7     186.0     574   32.4%
  18 Zappa Mexico II MP     : 2857.0     154.0     519   29.7%
  19 Gaviota 1.0 MP         : 2755.8     114.0     459   24.8%
  20 Arasan 17.4 MP         : 2708.9      99.0     488   20.3%
  21 Tornado 5.0 MP         : 2697.3      96.0     476   20.2%
  22 MinkoChess 1.3 MP      : 2689.4      88.0     479   18.4%
  23 Sjeng WC2008 MP        : 2683.5      86.0     482   17.8%
  24 EXchess 7.31b MP       : 2676.3      93.0     517   18.0%
  25 Crafty 24.1 MP         : 2673.2      94.0     531   17.7%

2) Standings - Included All (Wins and Draw Games):


   # PLAYER                 : RATING    POINTS  PLAYED    (%)
   1 Stockfish 151014 MP    : 3547.4     954.0    1300   73.4%
   2 Komodo 8 MP            : 3518.1    1172.0    1678   69.8%
   3 Stockfish 170914 MP    : 3518.0    1054.5    1425   74.0%
   4 Houdini 4 MP           : 3491.9    1119.0    1677   66.7%
   5 Gull 3 MP              : 3434.2     964.0    1567   61.5%
   6 Rybka 4.1 MP           : 3350.0     681.0    1097   62.1%
   7 Chiron 2 MP            : 3250.7     542.0    1007   53.8%
   8 Protector 1.7.0 MP     : 3246.3     448.5     803   55.9%
   9 Hannibal 1.4b MP       : 3232.3     519.5    1008   51.5%
  10 Protector 1.6.0 MP     : 3215.1     277.5     692   40.1%
  11 Texel 1.04 MP          : 3213.5     495.0    1006   49.2%
  12 Hiarcs 14 WCSC MP      : 3202.5     482.0    1007   47.9%
  13 Senpai 1.0 MP          : 3191.9     468.5    1006   46.6%
  14 Junior 13.3 MP         : 3142.6     387.5     956   40.5%
  15 Spike 1.4 MP           : 3139.4     404.0    1007   40.1%
  16 Spark 1.0 MP           : 3135.7     380.5     958   39.7%
  17 Shredder 12 MP         : 3133.6     378.0     958   39.5%
  18 Zappa Mexico II MP     : 3103.2     312.0     835   37.4%
  19 Gaviota 1.0 MP         : 3056.1     234.0     699   33.5%
  20 Tornado 5.0 MP         : 3022.1     207.5     699   29.7%
  21 Arasan 17.4 MP         : 3018.2     204.5     699   29.3%
  22 MinkoChess 1.3 MP      : 3009.2     198.0     699   28.3%
  23 Sjeng WC2008 MP        : 3004.6     194.5     699   27.8%
  24 EXchess 7.31b MP       : 2990.5     184.0     699   26.3%
  25 Crafty 24.1 MP         : 2981.9     178.0     699   25.5%


More Details:


   # PLAYER                 : RATING    POINTS  PLAYED    (%)
   1 Stockfish 151014 MP    : 3842.3     687.0     766   89.7%
   3 Komodo 8 MP            : 3730.7     821.0     976   84.1%

As we see (in case of calculation without draws),
The Elo difference between SF and K8 is more than 100 Elo
But in reality is that true ?? ))

So in case of preferring critical weak openings, which are ending mainly with wins (less draws),
Then I should point out once more:
-Imagine Michael Schumacher to be on 1990 model car (i.e LADA ))
- A such testing will be fun, but only fun....no more no less...!!))

In other words,
- Such results should NOT be take very seriously, just saying...!

Instead of draw discussions, I think:
We should concentrate on Winning Percentage of the played lines…

Otherwise...the Error bar margins will be too high,
For example, oftenly we will see 200-300 Elo weaker engine to be the winner
And then I am afraid in that:
- Not 3.000 not 5.000 even 10.000 games per player will be not enough

And I strongly believe in that:
In case of using Perfect 2015 book, it is not needed running thousands of games per player
1000 games per player is quite enough data for determining about the Elo strength of X Engine

And my recommendation is that (about determining X engine Elo strength),
We should prefer short openings, but as far as possible the strongest ones for both sides

About appearing draws in my used openings...as a book author I can say:
- I am not responsible and the reason ... I optimize as far as possible the strongest lines
And I don't claim that my used openings are perfect...
Still some of them need  serious optimizations, but I promise to do my best before final release...

One thing more,
Maybe I am missing some very strong engine openings, which are not included in my books...
So can you inform me please which are those openings ?
And if they are really very strong...then I can include some of them....in my next book release....

BTW, a little note:
Maybe chess engine programers can look more seriously over this issue,
It sounds good...if they work over their engines for newer opening ideas...to see less draws !
Who knows...maybe after several years, some of the top engines will have own opening codes...
In this way, we will start testing the engines without books....wow...it sounds super !)
And then...really I will be very happy...at least I will work less ))!!

Hopes helps....

Best,
Sedat