SedatChess

[Sedat Canbaz]

Dear Sedat, I have no credit for SF PB. I only added the book options and modified the Polybook code to optimize it, but nothing more. Full credit goes to the Stockfish developers. Thanks for your kind words and tour.

Best regards!!

Dear Anton

As usually, you are welcome and nice to see you )

It seems also that your engine creation method is
So successful that played as serious role about over!
Thank you again and for Me was BIG pleasure to test...

On the other hand,
I am still shocked, even under more critical CuteChess GUIs
(e.g CuteChess EXPER), besides played at fast TC 15s+0.5s but
Your SF-PB never lost on time for this I Congratulate you too!
You've done a good job..we need more programmers as you!
Where M.Z's engine (Hypnos) lost several games on time, sad..
Moreover, Hyonos's time loses were even at slower TC: 1m+0.5s
Actually nothing new again..unfortunately same issue appeared
During Duel NNUE vs BOOK e.g again MZ's engine lost on time:
https://sites.google.com/site/computers ... nn-vs-book

Keep up the good work!

Best,
Sedat

[Sedat Canbaz]

By the way,
And once more, I kindly request from all Eng and Gui programmers
Please check more carefully your work... before sharing online!
Really annoying, leads to wrong conclusions...when time loses are
On scene, yes, I'm good in my job (at least many people say so....)
But as you know, sometimes NOT all in my hands! Thanks in advance

And without to not mention this I can not as well,
If you can not be a master, professional in your job,
STOP blaming, banning others innocent...that's all!

Greetings

[Sedat Canbaz]

BREAKING NEWS (31.05.2025):
As usually, I am very pleased to announce that this time as
'Schumacher' is Tanick Ramz (OrgZ) and once more we noticed,
He is not only talented in programing but in book making too!

And as usual,
Other great ones on Top: OPTIMUS, Esculapio, Skynet, Cevdet Dual etc.

Note also that
I picked exactly same book participants of previous XLIII CS:
https://sites.google.com/site/computers ... nnue-cs-43
As main difference now, exc. TC, tour logo ) Btw I hope you like it )
All of public books played by latest SF-PB 250525 (by Anton CES)

By the way,
Reason of playing same books again...simply to see the influences,
Moreover what will be overall draw percentages..so now it's clear:
96% draw-ratio, these numbers are high.so in same time that means,
I have to work harder and smarter..sure to reduce overall draws!
Otherwise, FUN is NOT scene for my side..I assume for many as well..
Ok dear friends, let's Me see, what I can do in the next future tours...

Latest Public Champion: Tunafish - My Congratulations to T.R (OrgZ)!

Code: Select all

Rank Name              Elo    +    - games score oppo. draws 
   1 Tunafish 3.2     3834    4    4   720   52%  3825   96% 
   2 OPTIMUS 2502     3833    4    4   720   51%  3825   95% 
   3 Esculapio        3831    4    4   720   51%  3825   96% 
   4 Skynet 180724    3830    4    4   720   51%  3825   94% 
   5 Cevdet Dual      3830    4    4   720   51%  3825   97% 
   6 OpenChess 2      3828    4    4   720   50%  3825   94% 
   7 Solidor Dual     3828    4    4   720   50%  3825   95% 
   8 AdrianOliva 3    3827    4    4   720   50%  3825   98% 
   9 DONfree 090724   3826    4    4   720   50%  3825   94% 
  10 Leschats 260325  3824    4    4   720   50%  3825   98% 
  11 UpAndDown 2      3824    4    4   720   50%  3825   96% 
  12 Bronco 180325    3824    4    4   720   50%  3825   98% 
  13 Dara 141024      3823    4    4   720   50%  3825   97% 
  14 SonAs 7 DraGoN   3823    4    4   720   50%  3825   97% 
  15 Akan 2025 ser    3823    4    4   720   50%  3825   96% 
  16 HCanFree 2       3822    4    4   720   50%  3825   97% 
  17 Green Dragon     3820    4    4   720   49%  3826   96% 
  18 LOBO 230225      3820    4    4   720   49%  3826   95% 
  19 Dogon 270325     3810    3    3   720   47%  3826   92% 

GAMES:
https://mega.nz/file/v8pAFQzI#XcnLUsm83 ... 8nc7DOdR7c


Best,
Sedat

[Sedat Canbaz]

UPDATE

Meanwhile, here are LXIV CS's stats (generated by ChessBase program),
And after checking closely, no any game is found to be lost on time...
This is really a good record, because current data is based on fast
Bullet (15s+0.5s) and played 6840 games !!In short, just imagine how
SF-PB is so much stable... and as same words are going for many others!
And once more, Congrats and BIG thanks to all.. for your work over years!



Greetings

[Sedat Canbaz]

EDIT: LXIV CS's stats = XLIV

[Sedat Canbaz]

Hello There,

I just ran some more new useful strength tests,
Where this time all Top engines are played with
Slower Bullet control: 30s+0.5s but I've realized
To create/test a new main player: SF-PB 221124 SC,
Which managed to qualify for next book championship!

To be more clear,
SF-PB 221124 SC is played via Non-Default Evalfile:
nn-a1ecdf2c9b0c.nnue, Btw, current idea is mine..but
Special thanks to Linrock, Skynet, rest eng teams too,
Otherwise without all of you,
All of these exciting testings would not exist as well

And according to my latest new strength testings,
SF-PB 221124 SC seems to be not so bad, right ?
Managed to overcome SF-PB 250525 (by Anton CES),
Plus defeated with clear margin Obsedian 16.02 too!

What I can say more,
Actually as one of the most important things is that,
SF-PB SC so stable (good in book settings, no time loses)
Less Drawish...plus in same time a superior strong!
In other words, what we can need more ?) FUN ?
We will see...sure all this 'mystery' in next tour !)

Moreover,
SF-PB 221124 SC engine is incredible small in size,
That makes reasonable sense..at least for my side,
Due for each book player I install a diff. copy eng..

And here are the latest played results, enjoy:

SF-PB 221124 SC vs Obsidian 16.02: +39 Elo for SF-PB SC

Code: Select all

1   SF-PB 221124 SC  +65/-9/=426 55.60%  278.0/500
2   Obsidian 16.02   +9/-65/=426 44.40%  222.0/500[

Draw-Ratio: 85%

-------------------------------------------------------
SF-PB 221124 SC vs SF-PB 250525: +3 Elo for SF-PB SC

Code: Select all

1   SF-PB 221124 SC  +23/-19/=458 50.40%  252.0/500
2   SF-PB 250525     +19/-23/=458 49.60%  248.0/500

Draw-Ratio: 91%

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

SF-PB 221124 SC's Draw Itself Test: 91%, not so high!

Code: Select all

1   SF-PB 221124 SC  +18/-15/=367 50.38%  201.5/400
2   SF-PB SC (Copy)  +15/-18/=367 49.63%  198.5/400

Btw, is this all? not of course...for examples,
I ran many other beta tests, but no much free time for all..

GAMES:
https://mega.nz/file/C5QGyTqQ#CWSN4G6Vn ... CR878CZr-Y

Conditions:
2x Epyc 7B12, CuteChess, Ponder OFF, 30s+0.5s, 1 Core, 64 Hash, 4-MEN
All engines used to play via Unique suite: each opening played twice

Have fun,
Sedat

[Anton101]

Hi Mr. Sedat now you can download SF PB 17.1 052025 with the net embedded off: https://t.me/chessenginesac

[Sedat Canbaz]

Hi Mr. Sedat now you can download SF PB 17.1 052025 with the net embedded off: https://t.me/chessenginesac

Many thanks dear Anton,
Now I am busy but later I will download..

Greetings

[Sedat Canbaz]

Hello Friends,

Occasionally I just found other Deeds series mistakes,
viewtopic.php?t=3648&start=80
It is all about for old tour, but better later than never!

And unfortunately he insulted Me again over..
Such as I say a lie...but this is unacceptable!
And not true in Deeds edited/taken screenshots..

Anyhow,
I realized to make it more clear, actually I'm tired to explain,
Inform.. but it seems Deeds has serious troubles to learn over..
Once more I will try... but this is as last time over, because
I have more important things to do...thanks for understanding

And here we go..

In other words,, in those times:
Simply I was referring about SpecTral_1.0 vs No Book (Cfish),
And as we see here, Spectral_1.0's Exp is gained 40-50 Elo:

Code: Select all

Rank Name              Elo    +    - games  score oppo. draws 
  50 SpecTral_1.0     3635    8    8  1280   49%  3641   89% 
  67 No Book          3588    8    8  1040   42%  3641   73% 

Btw, here is my original posting (belonging to 06.12.2024):
https://sites.google.com/site/computers ... bec-1-news

Spectral_1.0 (despite on its new tiny learning exp file) is gained close to 50 Elo
Better than No Book (Cfish..) e.g normally via Balsa..both are close in strength..
Here just imagine e.g for moment: Spectral_1.0's Exp size is about 1 (one) MB
And as a small point more, here the opponents are not so weak..in other words,
No doubt that so far best learning record ever seen is belonging to Spectral_1.0
And to be more clear, here I refer about tiny exp files,which started from scratch..
To be sure, you can check also the other learning engines: Eman, Brainlearn etc.
What I told before? Anton is Anton, not via papers but via facts, sure I admit too
Exp files are strong but not on same level as the strongest books..at least so far
The latest new results indicate this...but you know, there is nothing impossible!!
But till to this date, what I say? just because it's very hard...it does not mean to
Give up..try again..and sooner or later, you will unlock the door to your success!
And as friendly advice: do it now..because sometimes 'later' becomes 'NEVER'

As final words, and once more (for people as Deeds, Sarona etc):
Why so much grudge, arrogance, hatred, selfishness, jealousy?
Where is goodwill, tolerance, common sense, liking, respect?


Best,
Sedat

[Sedat Canbaz]

Hello there,

As usually, I am very pleased to announce also that,
Anther new book tour is started but as official CS!
Where again via same Public players as firstly..but
Note that all books are played by SF-PB 221124 SC,
Plus as slower Bullet time control 30s+0.5s is used,
Despite on almost 2x times slower than recent 15s+0.5s,
Dogon 270325's 30s+0.5s produced draws are about: 91%
Where Dogon270325's 15s+0.5s produced draws were 92%

In other words, via using NON default NN Evalfile,
It's very clear that SF-PB 221124 SC is less Drawish!
Plus as we noticed recently..SF-PB SC is not weaker!

By the way,
Here are the 1st LIVE Gauntlet results:

Code: Select all

   Rank Name                        Elo     +/-   Games   Score   Draw 
   0 Dogon 270325                  -22      10     360   46.8%   91.4% 
   1 Skynet 180724                  70      64      20   60.0%   80.0% 
   2 Tunafish 3.2                   53      56      20   57.5%   85.0% 
   3 Esculapio                      53      56      20   57.5%   85.0% 
   4 DONfree 090724                 53      56      20   57.5%   85.0% 
   5 SonAs 7 DraGoN                 35      46      20   55.0%   90.0% 
   6 Leschats 260325                35      46      20   55.0%   90.0% 
   7 Dara 141024                    35      46      20   55.0%   90.0% 
   8 Cevdet Dual                    35      46      20   55.0%   90.0% 
   9 LOBO 230225                    17      33      20   52.5%   95.0% 
  10 HCanFree 2                     17      33      20   52.5%   95.0% 
  11 Bronco 180325                  17      33      20   52.5%   95.0% 
  12 UpAndDown 2                     0      69      20   50.0%   80.0% 
  13 Solidor Dual                    0       0      20   50.0%  100.0% 
  14 OPTIMUS 2502                    0       0      20   50.0%  100.0% 
  15 Green Dragon                    0       0      20   50.0%  100.0% 
  16 Akan 2025 ser                   0      48      20   50.0%   90.0% 
  17 AdrianOliva 3                   0       0      20   50.0%  100.0% 
  18 OpenChess 2                   -17      33      20   47.5%   95.0% 

360 of 360 games finished.

Note also that,
Soon as possible, many more entries are coming...
Plus as alternative, many Top engines 3700+ Elo,
Which will be played by various opening books:

And here are some of them:
Alexandria 8.0, 23_24
Berserk 100325, BulletSBT
Caissa 1.22, Orange
CFish 260623, OlisFish201123
Clover 8.1, Pampa 3
Fritz 20, Geralt 051024
Honey 060721, SF_SE
Horsie 1.1, CaterpillaR v1
Integral 7, Cubail 081224
K.Dragon 3.3, Stealth 1.1
Lizard 11.2, Cevdet 131024
Neon-Tetra, Donmez Dual
PlentyChess 5.021, Balaklava
Obsidian 16.02, SappHire
Raubfish X48c3 SC, Revenge
RubiChess 170824, Castigo
Starzix 6.1, HuckleBerry 11
Viridithas 17, Oran

And frankly,
I am much excited, since a new tournament idea,
Never organized before...in this way, there will be
More wins..that means more FUN on scene as well!

Moreover, via latest new ongoing book CS:
Book/Eng authors will have chances to check,
To improve their weaknesses etc..because
You will never know about until you try!

And please stay tuned, more updates soon..

Greetings )