Sedat Canbaz

[Sedat Canbaz]

Hello there again )

Well, just I'd like to inform you also that,
The long awaited Duel: DON Vs HCAN is started!

Some Notes,
200 games will be played, sure if score will be equal..
Then, extra 20 tie-break games will be...and so on..
DON's released date: 10.07.2023 / HCAN: 26.05.2023

In other words,
DON is the same vers, which played vs Breaker book
HCAN player is the same public version: HCanFree Upd
As reason why old, Hamdullah Can informed Me that
He is in Military service, so no free time for chess..
So.. I did not request a newer DON from Vasid Chouhan
Thanks for understanding

More Details,
These authors have no fear..fighting as GLADIATORS!
HCAN's author belongs to the best youngest talents!
This time no comments for DON's author, exception
Vasid Chouhan belongs from the past to present!
And both authors proved to be so good in many times!
Sure not via papers, fake polls etc. but with real facts!

What I can say more,
A Duel match between two great Champions!
Really I've no any idea, which book will win?!

But there is one true,
No doubt that two TITANS will be on 'real' Arena !!

More info very soon and please stay tuned !)

Good luck to both Top authors )

Best,
Sedat

[massimilianogoi]

The last your games with it boosted up Grendel and I won 9-0 in a Adrian Oliva match! Grendel has no manual tuning, it just learns from its wins and losses.

When a match DON vs Goi?

[Sedat Canbaz]

The last your games with it boosted up Grendel and I won 9-0 in a Adrian Oliva match! Grendel has no manual tuning, it just learns from its wins and losses.

When a match DON vs Goi?

1st of all,
Congrats...and nice to see Grendel won 9-0

On other handi
Why Goi especially vs DON ? ))
What about Goi vs Hcan ?

And I think it is too early... let's.. see
Because HCAN may win vs DON, right ?)

By the way, I admit also that,
I have no much free time for duels (such as only 2 players..)
E.g if each time Duels..too much double games may appear...

That's why as usual, I prefer :
Running round-robin tours with many players etc.
Plus in this way is better regarding to see real strength..
Just one example, X book may overcome vs other...but
It does not give us much stats as best, strongest book out there!
You know.. more players, sure with more games are better stats!

Greetings

[massimilianogoi]

As you wish But neither round robin multi books are 100% fair: some time ago I've discovered in a tournament of mine that a previous version of DON exploited a weakness on other books to get the first place but the 4th or 5th move had score like -0.10. I was tempted to tune with that trick too but at the end I decided to not do it.

Speaking of the slanders you got on that other forum: I can confirm that the Sedat Canbaz games are legit: I have verified them dozens of times with the lines of my book and they match.

The last your games with it boosted up Grendel and I won 9-0 in a Adrian Oliva match! Grendel has no manual tuning, it just learns from its wins and losses.

When a match DON vs Goi?

1st of all,
Congrats...and nice to see Grendel won 9-0

On other handi
Why Goi especially vs DON ? ))
What about Goi vs Hcan ?

And I think it is too early... let's.. see
Because HCAN may win vs DON, right ?)

By the way, I admit also that,
I have no much free time for duels (such as only 2 players..)
E.g if each time Duels..too much double games may appear...

That's why as usual, I prefer :
Running round-robin tours with many players etc.
Plus in this way is better regarding to see real strength..
Just one example, X book may overcome vs other...but
It does not give us much stats as best, strongest book out there!
You know.. more players, sure with more games are better stats!

Greetings

[Sedat Canbaz]

As you wish But neither round robin multi books are 100% fair: some time ago I've discovered in a tournament of mine that a previous version of DON exploited a weakness on other books to get the first place but the 4th or 5th move had score like -0.10. I was tempted to tune with that trick too but at the end I decided to not do it.

No..I wish to test your books for sure,
And just do not get me wrong please

As I stated before as another but main reason regarding why
I don't plan to run more similar duels, at least for nowadays:
It's too much complicated (for my side) because FritzGUI has
No any option-running concurrent matches, plus I have other
Promises...where I have to run, resume other paused tours!
Thanks for your understanding!

About running round-robin tours,
It all depends...but I expect after 1000-1500 games (per player)
The new played games: to perform close as previous played...
If not so clear...just in case of re-testing the players etc.

That means in same time, and to be more clear:
Up to 100 or 200 or 300 games etc. the Eror margin can be higher..
But via playing many games... then
The situation, performance, ranking.. as usually can be so 'stable' !)

Speaking of the slanders you got on that other forum: I can confirm that the Sedat Canbaz games are legit: I have verified them dozens of times with the lines of my book and they match.

Thank you very much dear Massimiliano Goi )

Greetings )

[Sedat Canbaz]

Hello Friends,

The Duel (DON vs HCAN) ended, for more details, standings:
https://sites.google.com/site/computers ... on-vs-hcan

Games:
https://mega.nz/file/6lwSGY5Q#AjqMoAAlX ... SwcQKOmnlw

Best,
Sedat

[massimilianogoi]

Just one thing:

Reason, too complicated (for my side) because FritzGUI has
No any option-running concurrent matches

You CAN make concurrent tournaments, I did it last time, but it's kinda tricky: if you plan to use the same engines you must copy them and rename them. Then you use them with more Fritz instances. I did go up to four Fritz tournaments running at the same time. Tricky but possible. And you can pause the tournaments whenever you want, unlike CuteChess.

You opened my mind few times giving me ideas, I want to give back the favour.

Hello Friends,

The Duel (DON vs HCAN) ended, for more details, standings:
https://sites.google.com/site/computers ... on-vs-hcan

Games:
https://mega.nz/file/6lwSGY5Q#AjqMoAAlX ... SwcQKOmnlw

Best,
Sedat

[Sedat Canbaz]

Just one tjing:

You CAN make concurrent tournaments, I did it last time, but it's kinda tricky: if you plan to use the same engines you must copy them and rename them. Then you use them with more Fritz instances. I did go up to four Fritz tournaments running at the same time. Tricky but possible. And you can pause the tournaments whenever you want, unlike CuteChess.

Yes I am aware of concurrent tournaments...
And since many years I run in similar ways

Btw, just in case,
I used to play separately many portable Fritz GUIs

Anyhow, if we are going to compare vs Fritz GUI:
CuteChess GUI does better job in Bullet, Blitz..
E.g it takes not so long time to complete all my tours,
Sure to gain: you need a machine with many cores!
E.g in my case, 2x EPYC 128c/256t plays a BIG role!
And any of my testing ends in shorter time than
Fritz GUIs...as other influence and reason:
As usual, I use to play via 64 concurrent matches!

Otherwise, just imagine:
if we open separately (in same time) 64 Fritz GUIs ?)

That's why I say: too complicated...)

But this is also true that,
If I had a machine with limited CPUs, then very likely
I'd prefer mainly to use Arena, Fritz, Shredder GUIs..
Because the mentioned GUIs offer: pausing, resuming...
Besides, as other reason (with limited CPU machine),
No much chances to run many concurrent matches...

Greetings

[Sedat Canbaz]

You opened my mind few times giving me ideas, I want to give back the favour.

Of course..and I think that this is best way..
And nothing wrong to learn from each other!

Btw, since we talk about minds..
Let Me say please some great sayings:

The mind is like an umbrella. Its most useful when open!

After raining,
Using an umbrella is not useful at all!

Small minds discuss people.
Average minds discuss events.
Great minds discuss ideas.

[Sedat Canbaz]

Hello Friends )

Well, here is a new MP test: Ponder ON vs Ponder OFF,
Which is played under Arena GUIs (in Auto-232 mode):

Code: Select all

SCCT - Ponder ON vs Ponder OFF: No Winner !!
              
1   Ponder OFF  +0/-0/=94 50.00%   47.0/94  
2   Ponder ON   +0/-0/=94 50.00%   47.0/94  

GAMES:
https://mega.nz/file/uxxD2YTY#o_w211Elk ... nt7X2iKsvc

More Details / Conditions:
2x Epyc 7B12, HT+LP OFF, Perfect 2023, 2m+1s, 1024 MB Hash, 4-MEN
All the matches are played by RapTora 2.2 AC (4 Cores + NNUE ON)
And as usual, for each player is installed other copy engine folder

And to be more clear:
This test played in the name of planning new ratings 3700+
In short, just I was wondering what will be the influences..
Sure for better conclusion, we need more games, but anyhow
I realized to stop after 94 games, as main reason: no any win!
Plus as we see: Ponder ON or Ponder OFF: no any influence!

So for this reason,
Probably Super League tour will be payed as Ponder OFF too!
Yes and I admit since NNUE on scene: everything is changed!

Btw, some time ago I ran similar testings (sure as NON-NNUE),
Where previous results ended clearly in favor for Ponder ON:
https://sites.google.com/site/computers ... ain-rating
https://sites.google.com/site/computers ... 022-ponder

As Final Words,
I hope that all of the current played results to be useful...
Sure we need more of them...but that's all from Me...)

Best,
Sedat