Fringe- Special Engine and Book Tests

[Cengiz-Baba]

Hello, my old friend Sedat,

For my part, I like 2-engine tests more and I play 100 games in total with 50 games per engine. A mutual result with 50 different openings gives me the right results in the first place. However, in a situation where there is more time, I will definitely listen to you and play duels with more matches. I respect your knowledge about the tournament and I am your follower.

When I made a small comparison according to the results in 100 matches like in my first example above, the winning engines in these matches always helped me to find the more powerful engine compared to other winning engines. Actually you can understand what I'm trying to do with this UHO, the engines are playing 0.0 positions with the classic opening books and because they are clone engines in their memory, over 95% draws are achieved. It becomes impossible to see the engine change difference there. Even in these other tournaments the result can change constantly. One engine can finish 5/10 in this tournament and suddenly 1/10 in the next tournament.

That's why I think we need engines that have a good calculation of the depth of certain advantages leading to a definitive result. UHO is very useful for me to test this.

Best regards my dear friend Sedat,
Fringe.

[Sedat Canbaz]

Hello, my old friend Sedat,

For my part, I like 2-engine tests more and I play 100 games in total with 50 games per engine. A mutual result with 50 different openings gives me the right results in the first place. However, in a situation where there is more time, I will definitely listen to you and play duels with more matches. I respect your knowledge about the tournament and I am your follower.

When I made a small comparison according to the results in 100 matches like in my first example above, the winning engines in these matches always helped me to find the more powerful engine compared to other winning engines. Actually you can understand what I'm trying to do with this UHO, the engines are playing 0.0 positions with the classic opening books and because they are clone engines in their memory, over 95% draws are achieved. It becomes impossible to see the engine change difference there. Even in these other tournaments the result can change constantly. One engine can finish 5/10 in this tournament and suddenly 1/10 in the next tournament.

That's why I think we need engines that have a good calculation of the depth of certain advantages leading to a definitive result. UHO is very useful for me to test this.

Best regards my dear friend Sedat,
Fringe.

Hello Dear Fringe )

It's ok.. no problem )
And you can run at any cond. as you wish...

Simply here I'm trying to help you via sharing my ecperience with you
Just for Computer Chess Progress... )

Btw, for anyone missed,
I have tested many many opening suites, to see what is going on...
And I strongly suggest to check carefully about played results:
https://sites.google.com/site/computers ... testings-2

If still not clear, it is not so bad idea to check 1st testing as well:
https://sites.google.com/site/computers ... testings-1

Btw, I'm World's 1st ever rating creator by UHO openings.. not real author )))

Ok.. for right now that's all from Me )

Greetings

[Cengiz-Baba]

My dear friend Sedat,

Seriously, you're perfect. That's two tests I thought I could do and two tests I couldn't do. I understand what you want to say very clearly. Then testers should test according to many test books and provide more data. I don't like low drawratio, that's why I like UHO so much. I have to explore your site a bit more and see many things because there is a lot of valuable work.

Best regards my friend,
Fringe.

[Cengiz-Baba]

Test book: UHO 2024 16 moves.
Time:1+1
Threads:1
Hash:64

Score of Cool Iris 12.0_D3072_avx2 vs CorChess 4 16.1-3072-avx2: 24 - 27 - 49 [0.485]
... Cool Iris 12.0_D3072_avx2 playing White: 24 - 0 - 26 [0.740] 50
... Cool Iris 12.0_D3072_avx2 playing Black: 0 - 27 - 23 [0.230] 50
... White vs Black: 51 - 0 - 49 [0.755] 100
Elo difference: -10.4 +/- 48.9, LOS: 33.7 %, DrawRatio: 49.0 %
SPRT: llr 0 (0.0%), lbound -inf, ubound inf
100 of 100 games finished.

Thanks to the engine owners.
Fringe.

[Sedat Canbaz]

My dear friend Sedat,

Seriously, you're perfect. That's two tests I thought I could do and two tests I couldn't do. I understand what you want to say very clearly. Then testers should test according to many test books and provide more data. I don't like low drawratio, that's why I like UHO so much. I have to explore your site a bit more and see many things because there is a lot of valuable work.

Best regards my friend,
Fringe.

Dear Fringe,

Yes..agreed, via UHO openings..
No doıubt that the draw ratio is unbelievable very low!
And during via UHO testings, it was fun for my side as well,
But just in case (via UHO)... simply I was trying to tell you that,
It is required thousands of games...sure for accurate standings!
But even though...still no any guarantee about real strength...

I mean, even if running thousands, let's say 10.000 games (per player)
We may see 100 Elo difference values ...but who cares ? personally I do not..
Do you know why ? no any Top author will use these lines in serious tours!
Only in fun testings..no more no less...

Btw, several months ago,
I created and tested a book based on latest SPCC games,
And very likely it's based on UHO openings and performed
As last place...even I canceled testing SPCC book in the active tour:
https://sites.google.com/site/computers ... k-nn-cs-23

Because many opponents gained quite easy points..
And this is not so much fair! in other words,
My target is test those ones, which come with less holes...

And thank you for your interest...
You make Me feel happy... )

Best,
Sedat

[Cengiz-Baba]

My dear friend Sedat,

Seriously, you're perfect. That's two tests I thought I could do and two tests I couldn't do. I understand what you want to say very clearly. Then testers should test according to many test books and provide more data. I don't like low drawratio, that's why I like UHO so much. I have to explore your site a bit more and see many things because there is a lot of valuable work.

Best regards my friend,
Fringe.

Dear Fringe,

Yes..agreed, via UHO openings..
No doıubt that the draw ratio is unbelievable very low!
And during via UHO testings, it was fun for my side as well,
But just in case (via UHO)... simply I was trying to tell you that,
It is required thousands of games...sure for accurate standings!
But even though...still no any guarantee about real strength...

I mean, even if running thousands, let's say 10.000 games (per player)
We may see 100 Elo difference values ...but who cares ? personally I do not..
Do you know why ? no any Top author will use these lines in serious tours!
Only in fun testings..no more no less...

Btw, several months ago,
I created and tested a book based on latest SPCC games,
And very likely it's based on UHO openings and performed
As last place...even I canceled testing SPCC book in the active tour:
https://sites.google.com/site/computers ... k-nn-cs-23

Because many opponents gained quite easy points..
And this is not so much fair! in other words,
My target is test those ones, which come with less holes...

And thank you for your interest...
You make Me feel happy... )

Best,
Sedat

My dear friend Sedat,

Your great ideas and fun attitude are motivating. I understand you very well in everything. I will also try the opening book with lower rates instead of the book with UHO 0.85-1.04 rates. It is very pleasant to exchange information with you. You are very good at being a road map. I am grateful for your contribution to my tests and to me.

With respect and heartfelt love,
Fringe.

[Sedat Canbaz]

My dear friend Sedat,

Seriously, you're perfect. That's two tests I thought I could do and two tests I couldn't do. I understand what you want to say very clearly. Then testers should test according to many test books and provide more data. I don't like low drawratio, that's why I like UHO so much. I have to explore your site a bit more and see many things because there is a lot of valuable work.

Best regards my friend,
Fringe.

Dear Fringe,

Yes..agreed, via UHO openings..
No doıubt that the draw ratio is unbelievable very low!
And during via UHO testings, it was fun for my side as well,
But just in case (via UHO)... simply I was trying to tell you that,
It is required thousands of games...sure for accurate standings!
But even though...still no any guarantee about real strength...

I mean, even if running thousands, let's say 10.000 games (per player)
We may see 100 Elo difference values ...but who cares ? personally I do not..
Do you know why ? no any Top author will use these lines in serious tours!
Only in fun testings..no more no less...

Btw, several months ago,
I created and tested a book based on latest SPCC games,
And very likely it's based on UHO openings and performed
As last place...even I canceled testing SPCC book in the active tour:
https://sites.google.com/site/computers ... k-nn-cs-23

Because many opponents gained quite easy points..
And this is not so much fair! in other words,
My target is test those ones, which come with less holes...

And thank you for your interest...
You make Me feel happy... )

Best,
Sedat

My dear friend Sedat,

Your great ideas and fun attitude are motivating. I understand you very well in everything. I will also try the opening book with lower rates instead of the book with UHO 0.85-1.04 rates. It is very pleasant to exchange information with you. You are very good at being a road map. I am grateful for your contribution to my tests and to me.

With respect and heartfelt love,
Fringe.

Thank you..you are very kind..

Wishing you good luck...

Best,
Sedat

[Cengiz-Baba]

Test book: UHO 2024 16 moves.
Time:1+1
Threads:1
Hash:64

Score of SF PB 070324-avx2 vs Brainlearn27-x86-64-avx2: 24 - 22 - 54 [0.510]
... SF PB 070324-avx2 playing White: 22 - 0 - 28 [0.720] 50
... SF PB 070324-avx2 playing Black: 2 - 22 - 26 [0.300] 50
... White vs Black: 44 - 2 - 54 [0.710] 100
Elo difference: 6.9 +/- 46.4, LOS: 61.6 %, DrawRatio: 54.0 %
SPRT: llr 0 (0.0%), lbound -inf, ubound inf
100 of 100 games finished.

Thanks to the engine owners.
Fringe.

[Cengiz-Baba]

Test book: UHO 2024 16 moves.
Time:1+1
Threads:1
Hash:64

Score of Cool Iris 12.0_avx2 vs TacticaL Gear 2_avx2: 19 - 32 - 49 [0.435]
... Cool Iris 12.0_avx2 playing White: 19 - 0 - 31 [0.690] 50
... Cool Iris 12.0_avx2 playing Black: 0 - 32 - 18 [0.180] 50
... White vs Black: 51 - 0 - 49 [0.755] 100
Elo difference: -45.4 +/- 48.9, LOS: 3.4 %, DrawRatio: 49.0 %
SPRT: llr 0 (0.0%), lbound -inf, ubound inf
100 of 100 games finished.

Thanks to the engine owners.
Fringe.

[Cengiz-Baba]

My dear friend Sedat,

Seriously, you're perfect. That's two tests I thought I could do and two tests I couldn't do. I understand what you want to say very clearly. Then testers should test according to many test books and provide more data. I don't like low drawratio, that's why I like UHO so much. I have to explore your site a bit more and see many things because there is a lot of valuable work.

Best regards my friend,
Fringe.

Dear Fringe,

Yes..agreed, via UHO openings..
No doıubt that the draw ratio is unbelievable very low!
And during via UHO testings, it was fun for my side as well,
But just in case (via UHO)... simply I was trying to tell you that,
It is required thousands of games...sure for accurate standings!
But even though...still no any guarantee about real strength...

I mean, even if running thousands, let's say 10.000 games (per player)
We may see 100 Elo difference values ...but who cares ? personally I do not..
Do you know why ? no any Top author will use these lines in serious tours!
Only in fun testings..no more no less...

Btw, several months ago,
I created and tested a book based on latest SPCC games,
And very likely it's based on UHO openings and performed
As last place...even I canceled testing SPCC book in the active tour:
https://sites.google.com/site/computers ... k-nn-cs-23

Because many opponents gained quite easy points..
And this is not so much fair! in other words,
My target is test those ones, which come with less holes...

And thank you for your interest...
You make Me feel happy... )

Best,
Sedat

My dear friend Sedat,

Your great ideas and fun attitude are motivating. I understand you very well in everything. I will also try the opening book with lower rates instead of the book with UHO 0.85-1.04 rates. It is very pleasant to exchange information with you. You are very good at being a road map. I am grateful for your contribution to my tests and to me.

With respect and heartfelt love,
Fringe.

Thank you..you are very kind..

Wishing you good luck...

Best,
Sedat

Could you send me the opening book suites, please? I couldn't find most of them. Unless it's special.

e-mail: durakem18@itu.edu.tr


Thanks Sedat.
Fringe.