cubail test

[cubail]

Blitz 1.0min+1sec 4 core frıtz17 gui syzgyz 6 man Intel(R) Xeon(R) CPU E5-2690 0 @ 2.90GHz play 200 games

DESKTOP-Q4G963E, Blitz 1.0min+1.0sec 0

12
1 CorChess 4 PB 141023s ½½ 1.0/2 1.00
2 CorChess 4 PB 141023 ½½ 1.0/2 1.00

1 CorChess 4 PB 141023s ----- OPTIMUS_2311 bin book
2 CorChess 4 PB 141023 ------ everest bin book

test start

[cubail]

Blitz 1.0min+1sec 4 core frıtz17 gui syzgyz 6 man Intel(R) Xeon(R) CPU E5-2690 0 @ 2.90GHz play 221 games

DESKTOP-Q4G963E, Blitz 1.0min+1.0sec 0



1 CorChess 4 PB 141023s +1/=220/-1 50.00% 111.0/222 -4063.00
2 CorChess 4 PB 141023 +1/=220/-1 50.00% 111.0/222 -4063.00


1 CorChess 4 PB 141023s ----- OPTIMUS_2311 bin book
2 CorChess 4 PB 141023 ------ everest bin book

games
https://pixeldrain.com/u/gK4rZPR9
2 vıns 119 draw game small book everest good performans

[Sedat Canbaz]

Blitz 1.0min+1sec 4 core frıtz17 gui syzgyz 6 man Intel(R) Xeon(R) CPU E5-2690 0 @ 2.90GHz play 221 games

DESKTOP-Q4G963E, Blitz 1.0min+1.0sec 0



1 CorChess 4 PB 141023s +1/=220/-1 50.00% 111.0/222 -4063.00
2 CorChess 4 PB 141023 +1/=220/-1 50.00% 111.0/222 -4063.00


1 CorChess 4 PB 141023s ----- OPTIMUS_2311 bin book
2 CorChess 4 PB 141023 ------ everest bin book

games
https://pixeldrain.com/u/gK4rZPR9
2 vıns 119 draw game small book everest good performans

Many thanks dear Cevdet bey

Yes... that's true, like in your test..plus in many others testers...
Via latest SF Based NNUE engines, no much chances to see wins
As we see, almost 100% draws...and in same time that means,
No much chances to see which book is stronger...

For this reason,
I prefer different conditions...as you know,
In my tours fun is most important part...
But I respect your tours as well...

Have a nice weekend )

[cubail]

i totally agree with you that testing with the nnue chess engine reveals the power of real books

[cubail]

Blitz 1.0min+1sec 4 core frıtz17 gui syzgyz 6 man Intel(R) Xeon(R) CPU E5-2690 0 @ 2.90GHz play 84 games

DESKTOP-Q4G963E, Blitz 1.0min+1.0sec no nnue engıne test Cfish 030820 64 POPCNT NUMA


1 Cfish 030820 64 POPCNT NUMA +17 +4/=80/-0 52.38% 44.0/84
2 Cfish 030820 64 POPCNT NUMAa -17 +0/=80/-4 47.62% 40.0/84


1 Cfish 030820 64 POPCNT NUMA ----- cc bin book
2 Cfish 030820 64 POPCNT NUMAa --- OPTIMUS_2311 bin book

games
https://pixeldrain.com/u/8d1RnJrh

same test nnue engine CorChess 4 PB 141023

1 CorChess 4 PB 141023s ----- OPTIMUS_2311 bin book
2 CorChess 4 PB 141023 ------ cc bin book

DESKTOP-Q4G963E, Blitz 1.0min+1.0sec 0

1 CorChess 4 PB 141023s +0/=37/-0 50.00% 18.5/37 342.25
2 CorChess 4 PB 141023 +0/=37/-0 50.00% 18.5/37 342.25



test start wörk 37 games draw

[cubail]

1 CorChess 4 PB 141023s ----- OPTIMUS_2311 bin book
2 CorChess 4 PB 141023 ------ cc bin book
DESKTOP-Q4G963E, Blitz 1.0min+1.0sec 0
Blitz 1.0min+1sec 4 core frıtz17 gui syzgyz 6 man Intel(R) Xeon(R) CPU E5-2690 0 @ 2.90GHz play 100 games

1 CorChess 4 PB 141023s +7 +2/=98/-0 51.00% 51.0/100
2 CorChess 4 PB 141023 -7 +0/=98/-2 49.00% 49.0/100

games
https://pixeldrain.com/u/QGFV7XRU

Blitz 1.0min+1sec 4 core frıtz17 gui syzgyz 6 man Intel(R) Xeon(R) CPU E5-2690 0 @ 2.90GHz play 84 games

DESKTOP-Q4G963E, Blitz 1.0min+1.0sec no nnue engıne test Cfish 030820 64 POPCNT NUMA


1 Cfish 030820 64 POPCNT NUMA +17 +4/=80/-0 52.38% 44.0/84
2 Cfish 030820 64 POPCNT NUMAa -17 +0/=80/-4 47.62% 40.0/84


1 Cfish 030820 64 POPCNT NUMA ----- cc bin book
2 Cfish 030820 64 POPCNT NUMAa --- OPTIMUS_2311 bin book

games
https://pixeldrain.com/u/8d1RnJrh


which chess engine results do you think are correct

[Sedat Canbaz]

which chess engine results do you think are correct

1st of all,
Thanks for the interesting test..
And it's nice to see that you used different test conditions

You asked about which test is correct..

Well, I will try to answer, but before
There is something weird, strange here... (

I am not sure exactly, since I do not have cc book..
But rather I'd say it is related to book settings error ...

To be more clear,
If we check/compare the played opening codes,
We will see clearly that they are not so close to each other...
I mean there is problem with proportion...

For examples:
With Cfish 020820: all games include mainly C67 plus some C65
Where with CorChess 4 PB 141023: all codes include A, B, C, D, E

In other words, if you are sure that you set all correctly, then
Corchess 4 PB seems to be has serious bug (with books settings)
There is no other explanation of that...

One thing more,
I notice also that your Cfish 030820 crashed two times (2 games in line...)

You know, I use this great Cfish engine quite frequent (since 2020)
And played maybe million games on my various tournament hardwares...
And so far Cfish is never crashed or caused on exception to stop game

I suggest to install fresh Windows...

And instead of engines which comes with Width book option..
Use other Top engine, which may solve the right BIN book settings

E.g I suggest te use and try next time SF-POLY 220723
It is one of best, stable, strong close to 3800 Elo !

Because recently I tried/tested many other 3800 Elo engines,
And I had a bad experience...mostly were buggy...
Sure I am referring in case of running serious book tours!

Btw, one of the buggy engines is Stockfish 220823 sl
It did not to play properly the Everest book's openings....
At least on my hardware... I see also on you too,
It is true that it is not same engine, but...

Hope helps... )

Greetings

[Sedat Canbaz]

By the way,
Mr. Cevdet you can try as alternative Raubfish X48c3 SC
If nothing else, it's very stable, strong close 3750 Elo plus
No any issue is recorded playing with wrong book settings!
And just in case, you will get less draws as well...

E.g with other latest Top NNUE chess engines,
Especially with ones which are close in strength:
To see a win is became a dream...!)

Best,
Sedat

[cubail]

https://pixeldrain.com/u/4fQxVMTN
Raubfisch X48c3_and_GTZ27c3_(nn)_(sl)

[Sedat Canbaz]

https://pixeldrain.com/u/4fQxVMTN
Raubfisch X48c3_and_GTZ27c3_(nn)_(sl)

Yes...

Btw, here are Raubfish 'SC' settings: