SedatChess

[Sedat Canbaz]

UPDATE 2

Well, let's take a look (just for comparing) regarding Elo differences,
E.g g between original books vs my recent created derivatives ones..
1st of all, be aware that to expect identical performances is madness!
And due to not enough games: I merged many book vers. in one name..
(Sure here I am referring about my recent created Champ based books)
Moreover, all of them are not played under same test conditions...
As we know, any book performance depends on our used conditions,
But if nothing else..it seems that my based books are not too far
From the original ones, at least if comparing Elo differences, right ?
Besides, some my created based ones even may become stronger...
As main reason I do not allow/import the defeats (weakest) lines,
Sure here I refer about during each my derivatives book creations!

Number Ones CS 2018- 2022 (all are based/derivatives books):
https://sites.google.com/site/computers ... 35-stage-2

Code: Select all

Rank Name              Elo    +    - games score oppo. draws
  7 DON 21/22         3772    6    6  1188   51%  3764   91% 
  9 Zeus 2020         3770    6    6  1188   51%  3764   88% 
 18 MANTIS 21/22      3765    6    6  1188   50%  3765   88% 
 19 Cerebellum 2020   3764    6    6  1188   50%  3765   89% 
 32 Vitamin 2020      3750    6    6  1188   48%  3765   86% 
 34 GM Champions      3745    6    6  1188   47%  3765   86% 

BN X (all books are original, exc. GM Champions, it is based one):
https://sites.google.com/site/computers ... -nnue-cs-x

Code: Select all

Rank Name              Elo     +    - games score oppo. draws
 26  DON-B             3749    3    3  2430   52%  3739   91% 
 44  Zeus              3746    5    6  1000   52%  3736   90% 
 48  MANTIS Upd        3746    5    5  1000   51%  3739   90% 
 69  Cerebellum 3M     3743    6    6  1000   51%  3739   92% 
111  Vitamin 290320    3735    6    6   790   49%  3742   90% 
142  GM Champions      3726    6    5   900   48%  3740   89% 

Same PC Elo diff. (based books, played on 2x Epyc 7B12): 53 Elo

Code: Select all

Rank Name           Elo    +    - games score oppo. draws 
   1 Perfect 16    3790   10   11   650   56%  3749   72% 
  12 Rybka III     3737   11   11   650   48%  3753   82% 

Note: For comparing is used close in date book players

Online Elo diff. (original books, played on Playchess): 139 Elo
And both participants played by same CPU: i7-920 @4.00GHz
https://sites.google.com/site/computers ... k-nn-cs-35

Code: Select all

Rank Name                 Elo   +  –  games score oppo. draws
  1 SedatChess, Rybka 3  3368  11  11 2727   69%  3243   45%
198 Rybka3.ctg, Rybka 3  3229  24  24  482   51%  3220   55%

Some Remarks (according to my experience over long past years):
Why is there much Elo influence/differences? for some reasons:
1st of all, Rybka3.ctg requires much more games (for accurate Elo),
You may know, 482 games (per player) is good data, but not enough!
As other very important influence (sure here I mean during past..):
Rybka III ctg book was publicly available in those times that means
Its mostly strong opening lines were blocked or became drawish,
Or even some after period of time, they are started to loosing..
Actually this is nothing new..this happens with every public book..
Where Perfect 16 was not public available, plus played with often
Updates, adding newer/stronger moves..even in those old times:
I still remember that I made move optimizations in every hour etc.
Sure not always (24/7) of course... ) but quite frequent for sure...

And that's why and as usually I say over Elo points differences:
Online Play is totally different than on same PC matches, e.g
Some online games are lost on time..some online players are
Lost due to wrong engine or hardware configurations or some
Pardon, even I say quite frequent some lost due to hardware
Speed disadvantage...and mainly for these kind of reasons,
Each online played opening line/game is requited/needed to be
Checked manually and carefully..via analyzing to find stronger
Move...Btw, a small important note more:
This is completely wrong, if checking only at engine outputs!
If still not so clear, Which chess engine says the true to us?
There are many positions, come such as advantage/disadvantage..
Where later White or Blacks are winning/ loosing..that's why,
Sure for better performances: it is required many test-replay...e.g
The same won games..to see this opening is really strong/weak..
Otherwise (without serious tests..) many players may suffer...sure
Here I refer for openings, which are played on same CPU as well
One thing more, unfortunately there is also cheating via Online..
At least on Lichess data: a lot of 'fake' high Elo numbers recorded,
Which gained vs same/several or vs much weaker players.. hehe ))
What a hero are these kind of guys...LOL..In other words,
Come to 'real' Arena (SCCT) and let's see how 'good' you are !)
And just in case, do not blame as engine error/s etc.. because
The 'real' Champions never complain...at least mostly of them..
As reason, these Top book authors are too busy on improving,
Working for over newer stronger ideas that never played before!

And I hope also that
All my produced data, experience etc. will be helpful...

Best Wishes,
Sedat

[Sedat Canbaz]

EDIT:

As Online reference/source (about Rybka III and Perfect 16) used:
https://sites.google.com/site/computers ... tober-2009

[Sedat Canbaz]

Hello there,

Just I'd like to inform that I've realized to run another new CS,
Because, in recent both tours, mostly Champ books lines are
Not participated...so not organizing it..will be BIG miss, right?
Sure later... I hope to organize the planning 2023-2024 CS too

And here are the 1st played results of Number Ones CS -2022:

3rd Stage (2022)'s Current Leader: DON - Congrats to Vasid Chouhan!

Code: Select all

Rank Name                          Elo     +/-   Games   Score    Draw 
   1 DON Class                      26      28     108   53.7%   81.5% 
   2 Ganryu                         13      28     109   51.8%   81.7% 
   3 DON NN                         10      29     107   51.4%   80.4% 
   4 1337chess                       6      28     108   50.9%   81.5% 
   5 Cimarron                        6      30     110   50.9%   78.2% 
   6 Cubail                          3      26     106   50.5%   84.0% 
   7 Guzel Kiz                       3      32     107   50.5%   76.6% 
   8 GalactusTurbo                   0      27     108   50.0%   83.3% 
   9 Varied                         -6      23     109   49.1%   87.2% 
  10 DarkGalactus                  -26      25     108   46.3%   85.2% 
  11 Chucaro                       -34      21     114   45.2%   88.6% 

597 of 2750 games finished.

Conditions:
2x Epyc 7B12, CuteChess, 1 Core, Ponder OFF, 1m+1s, 128 MB Hash, 4-MEN
All Number Ones are used to play via 9-times Champion: Cfish 030820 C40

More Details:
This time all books are based on only 2022 NNUE games, exc. DON Class, e.g
It is tuned to prefer DON's Champion series, which are played as NON-NNUE
Be aware of that too please, each book player will consist min. 500 games

That's all for now and wishing good luck to all

Greetings

[Sedat Canbaz]

UPDATE 2

Via latest update, unfortunately DON Class lost the lead...
But still both DON versions are serious Champ candidates!
Sure Ganryu, Varied... are other serious candidates as well

Where for right now, the current Leader is:
Guzel Kiz Congratulations to Hasan Satrani!

Code: Select all

Rank Name                          Elo     +/-   Games   Score    Draw 
   1 Guzel Kiz                      17      18     355   52.4%   74.4% 
   2 Ganryu                         14      15     362   52.1%   83.1% 
   3 DON Class                       8      14     362   51.1%   85.1% 
   4 DON NN                          5      14     358   50.7%   84.1% 
   5 Varied                          3      14     359   50.4%   84.1% 
   6 1337chess                       1      15     357   50.1%   82.4% 
   7 Cubail                         -2      15     359   49.7%   82.7% 
   8 GalactusTurbo                  -6      14     358   49.2%   83.8% 
   9 Cimarron                       -8      15     360   48.9%   82.8% 
  10 Chucaro                       -11      13     356   48.5%   87.4% 
  11 DarkGalactus                  -21      15     358   46.9%   82.7% 

1972 of 2750 games finished.

[Sedat Canbaz]

UPDATE 3

So far, Guzel Kiz's performance is best - Congrats to Mr. Hasan!

On the other hand,
3rd Championship (2022) is still not over...e.g as next:
All of current books will be played by Raubfish NNUE SC,
In this way, we'll have chances to compare the influences,
Playing via 3750 Elo NN engine and via 3600 Elo NON-NN..

In same that means,
Champion title will be determined after 1000 games (per player)

Code: Select all

   1 Guzel Kiz                      22      15     500   53.1%   74.6% 
   2 Ganryu                         17      13     500   52.5%   82.6% 
   3 DON Class                      12      11     500   51.7%   85.8% 
   4 DON NN                          8      12     500   51.2%   84.4% 
   5 Varied                          0      12     500   50.0%   84.0% 
   6 1337chess                      -3      12     500   49.5%   84.6% 
   7 Cubail                         -7      13     500   49.0%   81.2% 
   8 Cimarron                       -8      12     500   48.8%   83.2% 
   9 Chucaro                       -10      11     500   48.5%   87.0% 
  10 GalactusTurbo                 -11      13     500   48.4%   82.8% 
  11 DarkGalactus                  -19      12     500   47.3%   83.4% 

2450 of 2750 games finished.

Note also that
After checking more carefully Guzel Kiz's data, I've noticed that
Minus plus half of played Guzel Kiz's games are based on NON-NN
So exc. DON Class, Guzel Kiz includes many NON-NNUE games too
But this is also true, all others are tuned to prefer 2022 NN games

Best,
Sedat

[Sedat Canbaz]

UPDATE 4

And here NN tour results, where Ganryu wins..
My Congratulations to Top Playchess Player!

Code: Select all

 #  PLAYER         POINTS(%) 
 1  Ganryu         254.0/500
 2  DON NN         253.5/500
 3  1337chess      251.0/500  
 4  Cubail         251.0/500 
 5  GalactusTurbo  250.5/500
 6  Varied         250.0/500  
 7  Cimarron       250.0/500  
 8  DON Class      248.5/500
 9  Guzel Kiz      248.0/500
10  Chucaro        247.5/500
11  Dark Galactus  246.0/500

Note: All Champs are played by Raubfish NNUE SC

Be aware too,
Overall's both tours results will be shared soon as possible...

Greetings

[Sedat Canbaz]

Breaking News (06.09.2024):
Via this CS: Ganryu Ganryu Ganryu said the last word again !! Yes and once
More Ganryu's opening lines prove to be one of the best e.g in B/N VI too:
https://sites.google.com/site/computers ... ct-bn-cs-6
2nd rank: Guzel Kız, 3rd + 4th ranks are belonging to both DON books, sure
I'm not going to share all their successful contests again, because I may miss,
But you know, they managed to be as Number Ones many times!! Actually as
Same words are going for all other Top players as well! Ok dear chess friends,
It's time to concentrate over next new planning CS 2023 - 2024, be aware too,
There is a possibility to be played two separate CS, e.g Champ books, which
To be based on only 2023 games, where the other CS to be based on 2024...

3nd Stage (2022) Champion: Ganryu- Congratulations to Playchess Player!

FOR FINAL RANKING:
https://sites.google.com/site/computers ... -35-stage3

GAMES:
https://mega.nz/file/G1Qx2BJD#UxvO8w3_c ... VQ8Ush7gWk


More Details (when the tour was active):
• Two separate tournaments are organized as NN plus as NON-NN as well
E.g in 1st tour: all books are played via 9-times Champion: Cfish 030820
But in the 2nd tour: all books are participated via Raubfish X48c3 NN SC
• In this new Championship: only Number One / Champ books are allowed
Sure I refer from Past to Present, where all of them are based/derivatives,
Created by Me / all books are BIN, based on only SCCT- NN 2022 tours...
And as players picked which performed Best (Champions/Number Ones)
• Maximum book size limit rule: up to 1 (one) MB (uncompressed) allowed
• Main Settings (for all participants): Book Depth 100 / Best Move Disabled
• Any Opening Book Leader (after 1000 games) will be titled as Champion
• The current overall Draw-Ratio (based on 5500 games) is a bit high: 88%
E.g the tour via NNUE engine: 94% draws are produced / NON-NN: 83%
Yes.. there are higher draws (than other organized tours) as main reason:
Mostly of the opening lines are played each other (sure during 2022 year)
And played serious influence to appear more draws.. since mostly were in
Fight each other..that's why and for more fun..1st tour played as NON-NN

Best,
Sedat

[Homayoun]

[Sedat Canbaz]

[Sedat Canbaz]

Dear Chess Friends,

As usually, I am pleased to announce that
One of my greatest book tours is started !!

And here are the 1st results of Number Ones CS (2023):

By the way, for right now we have 2 (two) Top Book Leaders:
Red Dragon+ Unicornio - Congrats to I. Hurnavich, A. Morano!

Code: Select all

Rank Name                          Elo     +/-   Games   Score    Draw 
   1 Red-Dragon                     43      31     138   56.2%   70.3% 
   2 Unicornio                      43      26     138   56.2%   79.0% 
   3 OrionPax                       33      33     136   54.8%   68.4% 
   4 C.B Conqueror                  25      29     137   53.6%   75.2% 
   5 Grendel                        23      27     135   53.3%   78.5% 
   6 Goddess                        20      28     137   52.9%   76.6% 
   7 M.Canseven                     15      28     137   52.2%   76.6% 
   8 Pimienta                       15      25     138   52.2%   81.2% 
   9 DON M.Chopra                   15      30     138   52.2%   72.5% 
  10 HCan-Upd                       10      29     136   51.5%   75.0% 
  11 Sentinel                       10      24     137   51.5%   82.5% 
  12 Cacho-SCCT                      8      29     137   51.1%   74.5% 
  13 Hamish                          5      28     137   50.7%   76.6% 
  14 Optimus                         3      25     137   50.4%   81.8% 
  15 LesChats                        0      25     137   50.0%   81.0% 
  16 HCanFree                        0      25     137   50.0%   81.0% 
  17 Chucaro                         0      25     138   50.0%   81.2% 
  18 Feta                           -3      30     136   49.6%   74.3% 
  19 Everest 3009                   -3      29     136   49.6%   75.7% 
  20 Taram                          -5      28     138   49.3%   76.8% 
  21 Solista                        -5      28     138   49.3%   76.8% 
  22 DON Hercules                   -5      28     137   49.3%   76.6% 
  23 Everest 0902                   -5      29     136   49.3%   75.0% 
  24 Alchemist                     -10      27     139   48.6%   78.4% 
  25 SF-SE                         -10      31     137   48.5%   72.3% 
  26 Bloom                         -13      32     137   48.2%   70.1% 
  27 SARI                          -13      26     136   48.2%   80.1% 
  28 OlisFish                      -13      30     135   48.1%   74.1% 
  29 Skynet-SCCT                   -15      31     138   47.8%   71.0% 
  30 Sambala                       -15      27     137   47.8%   78.1% 
  31 StudentChess                  -18      26     138   47.5%   79.0% 
  32 Barnabas                      -25      30     137   46.4%   73.7% 
  33 Lion                          -49      28     135   43.0%   75.6% 
  34 Wang Hao                      -64      31     138   40.9%   70.3% 
  
  2329 of 17952 games finished.

Conditions:
2x Epyc 7B12, CuteChess, 1 Core, Ponder OFF, 1m+06s, 32 MB Hash, 4-MEN
All Number Ones are used to play via 9-times Champion: Cfish 030820 C40

More info very soon...

Good luck and have fun )

Best,
Sedat