Sedat Canbaz

[Sedat Canbaz]

Downloaded, thanks. Is this file also available?

Eman exp (dated on 17.07.2023)

I did not make a buck up of previous exp versions
But the good news is that, the played ones available..

E.g here are both exp files, dated on 25.07.2023
https://pixeldrain.com/u/TtQEzaaf

Note: they are same ones which were dated 17.07.2023
Sure I'm referring ... before starting the Duel match...

Greetings

[Sedat Canbaz]

Shocking difference! Could it be that the opening suite influenced the result?

Is this Raptora private? And the experience file?

Definitely intriguing test, I'd like to make it too, but with the same material.

Sure you are free to re-test, re-run...even
Sounds good.. you know.. more better...!)

For example, Unique v50 suite is available:
https://sites.google.com/site/computers ... test-suite

As I stated above, both exp files, dated on 25.07.2023:
https://pixeldrain.com/u/TtQEzaaf

On other hand,
Why not?) If you check carefully my previous played test conditions..
For more details, please check my test organized on 17.07.2023
So, just in case you can generate same exp files..and exception that..
You have to (as very important note):
Follow strictly the same Eng versions and conditions, which I used..

Anyhow,
If you say it's too much complicated etc....then what I suggest:
Just use same eng versions as mine, and the same exp 25.07.2023
Sure do not forget to use same opening suite, time control etc.

After the end of your test..if nothing else,
We will have new data, e.g
Just in case of re-run the test... then we will see the influences
E.g Eman 8.91 will manege to correct its mistakes or not..
Or who knows? maybe even the Elo differences will be much higher?

Greetings

[massimilianogoi]

Awesome! Thanks.

Downloaded, thanks. Is this file also available?

Eman exp (dated on 17.07.2023)

I did not make a buck up of previous exp versions
But the good news is that, the played ones available..

E.g here are both exp files, dated on 25.07.2023
https://pixeldrain.com/u/TtQEzaaf

Note: they are same ones which were dated 17.07.2023
Sure I'm referring ... before starting the Duel match...

Greetings

[massimilianogoi]

I'll make the same test as soon as I can, but without opening suite. I want to find out if it's responsible of this huge gap.

Shocking difference! Could it be that the opening suite influenced the result?

Is this Raptora private? And the experience file?

Definitely intriguing test, I'd like to make it too, but with the same material.

Sure you are free to re-test, re-run...even
Sounds good.. you know.. more better...!)

For example, Unique v50 suite is available:
https://sites.google.com/site/computers ... test-suite

As I stated above, both exp files, dated on 25.07.2023:
https://pixeldrain.com/u/TtQEzaaf

On other hand,
Why not?) If you check carefully my previous played test conditions..
For more details, please check my test organized on 17.07.2023
So, just in case you can generate same exp files..and exception that..
You have to (as very important note):
Follow strictly the same Eng versions and conditions, which I used..

Anyhow,
If you say it's too much complicated etc....then what I suggest:
Just use same eng versions as mine, and the same exp 25.07.2023
Sure do not forget to use same opening suite, time control etc.

After the end of your test..if nothing else,
We will have new data, e.g
Just in case of re-run the test... then we will see the influences
E.g Eman 8.91 will manege to correct its mistakes or not..
Or who knows? maybe even the Elo differences will be much higher?

Greetings

[Sedat Canbaz]

I'll make the same test as soon as I can, but without opening suite. I want to find out if it's responsible of this huge gap.

I see... just in case that means if your test cond. will be not close as mine..
And in same time, your results will be very likely different than mine...

You know,
Any engine's performance is depending on our used conditions!

But anyhow (in case of test as without books, suite etc.):
My current sent both exp files will not play any role...,
Because they are learned on games played via Unique suite

That's why I suggest,
If you will run as without books, then please simply
Test both engines as without any pre-learned exp files

And frankly, your new test idea sounds good too,
Let's see the learning influences as without books )

Greetings

[Sedat Canbaz]

I'll make the same test as soon as I can, but without opening suite. I want to find out if it's responsible of this huge gap.

Meanwhile, according to my other exp / book tournament:
https://sites.google.com/site/computers ... t-ben-cs-1

I mean that both engines,
Each other are already played as without opening books!

And as we see,
Both engines are performed identically, but here the played
RapTora 2.2 is the version, which is compiled by Anton Ces:

Code: Select all

Rank Name                  Elo     +    - games score oppo. draws 
  17 NoPreExp, Eman        3778    3    3   600   49%  3782   97% 
  18 NoPreExp, RapTora     3778    3    3   600   49%  3782   98% 

In otrher words,
Once more we noticed that, any engine's performance depends on
Our used test conditions...

By the way,
Not sure exactly, but maybe later I may test RaptoRa 2.2 compiled by
Eduard Nemeth in exp/book tour...at least we will have more data...

And plus if you run a test...then will be much better..as I stated before,
More better... )

As final words,
RaptoRa is RapTora...managed to win the latest duel..
So...my opinion is same as before...best of the best!
Same words are going for many other engines as well

Greetings

[Sedat Canbaz]

Update

New test, but played under same conditions as before
Note: used same learned exp fies (dated on 25.07.2023)

In short,
The target is to see the influence via previous learned exp...

By the way,
Here are current played results, RapTora is leading again:

Score of Eman 8.91 vs RapTora 2.2: 3 - 7 - 27 [0.446]
... Eman 8.91 playing White: 3 - 2 - 13 [0.528] 18
... Eman 8.91 playing Black: 0 - 5 - 14 [0.368] 19
... White vs Black: 8 - 2 - 27 [0.581] 37
Elo difference: -37.7 +/- 58.1, LOS: 10.3 %, DrawRatio: 73.0 %
37 of 100 games finished.

Best,
Sedat

[Sedat Canbaz]

Update 2

Here are FINAL results, Winner is again RapTora!
My Congratulations to Anton, Eduard, SF team!

Code: Select all

Score of Eman 8.91 vs RapTora 2.2: 12 - 14 - 74 [0.490]
...      Eman 8.91 playing White: 12 - 3 - 35  [0.590] 50
...      Eman 8.91 playing Black: 0 - 11 - 39  [0.390] 50
...      White vs Black: 23 - 3 - 74  [0.600] 100
Elo difference: -6.9 +/- 34.8, LOS: 34.7 %, DrawRatio: 74.0 %
100 of 100 games finished.

GAMES:
https://mega.nz/file/ylxThLwQ#5hSYe1dPf ... VdEUE6CyMM

As Final Conclusion,
Even via latest re-run test.. it seems that RapTora is
Stronger..But according to other SCCT' competitions,
I can say that their performance is almost identical!

And what we noticed also that (via current Duel match),
With more learning games, Eman is performed better...

Plus it is clear 100% that,
The learning influence is playing as important role!
And RapTora 2.2 is doing not bad via Eman exp files!


Best,
Sedat

[Sedat Canbaz]

Hello Friends )

I am pleased to announce that,
RapTora test series in action...and for anyone
Interested, here are the latest played results

1st of all, about both tested versions:
AC = Anton Ces / EN = Eduard Nemeth

What I can say more,
Simply current tests are organized out of curiosity
E.g now we have more data about RapTora 2.2 AC / EN
Anyhow, I think that their points will be more close,
If the testings will contain over than 1500+ games...

Standings via Unique-319 openings: + 7 Elo in favor
For RapTora 2.2 AC, which is compiled by Anton Ces

Code: Select all

1   RapTora 2.2 AC  +43/-30/=565 51.02%  325.5/638
2   RapTora 2.2 EN  +30/-43/=565 48.98%  312.5/638

 Draw-Ratio (based on 638 games) not so high: 88.5%

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

Standings via Unique-50 openings: + 17 Elo in favor
For RapTora 2.2 AC...

Code: Select all

1   RapTora 2.2 AC  +16/-11/=73 52.50%   52.5/100
2   RapTora 2.2 EN  +11/-16/=73 47.50%   47.5/100

   Draw-Ratio (based on 100 games) very low: 73%

GAMES:
https://mega.nz/file/SlAXzBQY#v0wfX-ivy ... PVZ9usKNY8

Conditions / More Details:
2x Epyc 7B12, CuteChess, 1 Core, Ponder OFF, 1m+06s, 128 MB Hash, 4-MEN
Unique suite used and as usual, each of the opening line is played as twice
Note also that all matches are played via GUI's Concurrent option: 58
Concurrent numbers dropped, the reason high room temps, where I live...
Actually Dual processors are cooled via 2x NH-U12S TR4-SP3, but against
Any kind of risks...for these testings, Air Conditioner is used as well )
You may know, Dual EPYC machine is not so cheap, at least where I live!

As Final Words,
I'm doing my best since many years..mainly for Computer Chess's progress!
And I hope that you like my chess activities.. sure they are not perfect, but..
That's all from Me... !)

Have Fun,
Sedat

[Sedat Canbaz]

Hello Friends,

Meanwhile,
In next 2-3 weeks, I am planning to be on Holiday..

So.. please do not worry much...I am all right )

Sure later when I'll have more free time ...
I will try to be active again..at least on forums )

Best,
Sedat