Different playing style Excalibur Igor and Excalibur Ivan

This forum is for general discussions and questions, including Collectors Corner and anything to do with Computer chess.

Moderators: Harvey Williamson, Steve B, Watchman

Forum rules
This textbox is used to restore diagrams posted with the fen tag before the upgrade.
Post Reply
User avatar
paulwise3
Senior Member
Posts: 1505
Joined: Tue Jan 06, 2015 10:56 am
Location: Eindhoven, Netherlands

Different playing style Excalibur Igor and Excalibur Ivan

Post by paulwise3 »

If I am right, Nick already made some elo-tests with both Ivan and Igor.
Yesterday I got the Igor at home, and wanted to do some tests to see how identical they played.
For that purpose I used the Schachcomputer Testreport (https://www.schach-computer.info/wiki/i ... Testreport)
To my big surprise they play different at testposition 4!
Igor finds 1.Nf6! in 5:45 minutes.
Ivan plays 1.Rf2 after 4:45 minutes...!?
So I thought maybe Ivan is forced to take less computing time, and tried it again at infinite level. But even when I press Move after 7 minutes, Ivan still plays 1.Rf2 ?!
So I wondered if I really have the H8-Ivan?
Then I found a game on Nick's site where they play against each other, won by Ivan. I reproduced that game more or less till move 40 (the book openings were identical, after that Igor deviated once, Ivan 3 or 4 times, could be caused by different ponder-timings when executing the move) so I guess I do have the H8-Ivan.

But there is definitely a difference in playing style it seems.
Will come back when I have the result of the complete Schachcomputer Testreport.

Regards, Paul
2024 Special thread: viewtopic.php?f=3&t=12741
2024 Special results and standings: https://schaakcomputers.nl/paul_w/Tourn ... 25_06.html
If I am mistaken, it must be caused by a horizon effect...
Four.nine
Full Member
Posts: 911
Joined: Tue Mar 21, 2017 7:30 pm

Post by Four.nine »

Hello Paulwise,

I have these computers and can try some of these tests.
In fact, I already did, one:

Igor: Nf6! In 6:02 minutes.

But, I did not know what level to use....so I just picked level 66: 3 minutes average. But, the result is very close to yours!
Let me know what level you are using

Regards,

4.9

Oh....test position 4
User avatar
paulwise3
Senior Member
Posts: 1505
Joined: Tue Jan 06, 2015 10:56 am
Location: Eindhoven, Netherlands

Post by paulwise3 »

Hi 4.9,

You are close, but it should be level 92, which is tournament level 40 moves in 2 hours, then 20 in 1 hour. In effect it is the same average, but the outcome could be a little different.
I press New Game before setting up a new test position.

Regards, Paul
.
2024 Special thread: viewtopic.php?f=3&t=12741
2024 Special results and standings: https://schaakcomputers.nl/paul_w/Tourn ... 25_06.html
If I am mistaken, it must be caused by a horizon effect...
Four.nine
Full Member
Posts: 911
Joined: Tue Mar 21, 2017 7:30 pm

Post by Four.nine »

OK, Paulwise....everything you wanted to know about Test position 4....and asked.

“To my big surprise they play different at testposition 4!
Igor finds 1.Nf6! in 5:45 minutes.
Ivan plays 1.Rf2 after 4:45 minutes...!? “

Here’s mine....Level 92:

Igor: Nf6 5:44
IvanT: Rf2 4:45 10 MHz Ivan

Now:

IvanT2: Nf6. 5:43. 12 MHz Ivan
GM: Nf6. 5:44
GM28: Nf6. 5:45 But “realtime” only 2:28 Due to 28 MHz overclocking


First...it is virtually a miracle that the times correlate...exactly. (This now gives me hope for 2020!!!)
Second, there are 2 Ivans...you have the 10 MHz version (which has MODE button on right side)
The 12 MHz Ivan=Igor=GM
The universe is restored...(well in a couple of years, anyway).

Regards,

4.9
User avatar
paulwise3
Senior Member
Posts: 1505
Joined: Tue Jan 06, 2015 10:56 am
Location: Eindhoven, Netherlands

Post by paulwise3 »

Hi 4.9,

Wow, I didn't know there are two different H8-Ivans!?
And yes, the picture on Schacomputer.info does not have the MODE button at right, as mine does have.
At first I was a little disappointed to read your conclusion, but on the other hand I should be glad that my Igor and Ivan differ. That also explains why my Ivan deviates a number of times from Nick's game at http://www.spacious-mind.com/html/ivan_ ... rible.html, allthough... the picture there also has MODE on the right side!?!

Thanx very much for your info.

And I will give my Ivan 10 MHz a last chance for finding Nf6! in more time.

About the restoring of the universe... Since the recent Dutch elections result I am not so sure about that...

Universal worries regards,
Paul
2024 Special thread: viewtopic.php?f=3&t=12741
2024 Special results and standings: https://schaakcomputers.nl/paul_w/Tourn ... 25_06.html
If I am mistaken, it must be caused by a horizon effect...
User avatar
paulwise3
Senior Member
Posts: 1505
Joined: Tue Jan 06, 2015 10:56 am
Location: Eindhoven, Netherlands

Post by paulwise3 »

I let Ivan once more try testposition 4.
First attempt: 10 minutes: still 1.Rf2 :-(
Second attempt: 31 minutes: 1.Nf6!

But not finding the right move in 10 minutes leaves me in doubt if it is only the downgrading from 12 to 10 MHz, or that also some (minor) changes have been made in the program code itself...!?

Ok, I will finish the testpositions so we can have a better insight.

Testing regards,
Paul
2024 Special thread: viewtopic.php?f=3&t=12741
2024 Special results and standings: https://schaakcomputers.nl/paul_w/Tourn ... 25_06.html
If I am mistaken, it must be caused by a horizon effect...
Four.nine
Full Member
Posts: 911
Joined: Tue Mar 21, 2017 7:30 pm

Post by Four.nine »

paulwise3 wrote:Hi 4.9,

About the restoring of the universe... Since the recent Dutch elections result I am not so sure about that...

Universal worries regards,
Paul
Oh for f’s sake.....not you Dutch too.
The Dutch?
The land of Cruyff and Neeskens????
Say it aint so.

Hasn’t the world learned anything from us Germans?

BTW...the 31 minutes (to 6 mins.) means a big improvement of the program was made from 10 MHz IvanT to 12 MHz IvanT2.

4.9
User avatar
paulwise3
Senior Member
Posts: 1505
Joined: Tue Jan 06, 2015 10:56 am
Location: Eindhoven, Netherlands

Post by paulwise3 »

Hi 4.9,

It finds the move somewhere between 10 and 31 minutes, have to do some tries to refine it ;-)
But You are right, there must have been a considerable improvement in the program, because just a raise of 2 MHz cannot cause such a difference.

Paul
2024 Special thread: viewtopic.php?f=3&t=12741
2024 Special results and standings: https://schaakcomputers.nl/paul_w/Tourn ... 25_06.html
If I am mistaken, it must be caused by a horizon effect...
User avatar
paulwise3
Senior Member
Posts: 1505
Joined: Tue Jan 06, 2015 10:56 am
Location: Eindhoven, Netherlands

Post by paulwise3 »

I finished the tests for Igor, he scores considerably better then the Ivan 10 MHz:
Igor 12 MHz: 147 points
Ivan 10 MHz: 118 points

Igor solves positions 4, 5 and 21 where Ivan fails.
And solves some within 4 minutes where Ivan uses more then 4 minutes.

Further details will follow regards,
Paul
2024 Special thread: viewtopic.php?f=3&t=12741
2024 Special results and standings: https://schaakcomputers.nl/paul_w/Tourn ... 25_06.html
If I am mistaken, it must be caused by a horizon effect...
User avatar
paulwise3
Senior Member
Posts: 1505
Joined: Tue Jan 06, 2015 10:56 am
Location: Eindhoven, Netherlands

Post by paulwise3 »

Ok, I tried two mate problems.
First is derived from Colditz testposition 9:

[fen]r2r4/1p1R3p/5p1k/b1B1Pp2/p4P2/P7/1P5P/1K4R1 w - - 0 2[/fen]
Igor finds 1.Lf8+! in 0:46
Ivan 10 MHz finds it in 2:26

The second I found at https://kain.tweakblogs.net/blog/8625/w ... -in-4.html:

[fen]6Bk/6rP/8/7p/7K/2Q5/8/8 w - - 0 0[/fen]
Igor finds 1.Kh3! in 1:45
Ivan 10 MHz finds it in 5:47

So together with the previous mentioned result of Testreport's position 4, one gets the idea that the Ivan 10 MHz is a factor 3 slower then the Igor 12 MHz!?
That should indicate that there has been made some major optimization and/or improvement in the later Igor/Grandmaster program.

So I have to take my Mirage off the attick to see if there is a similarity to find...

Surprised regards,
Paul
2024 Special thread: viewtopic.php?f=3&t=12741
2024 Special results and standings: https://schaakcomputers.nl/paul_w/Tourn ... 25_06.html
If I am mistaken, it must be caused by a horizon effect...
User avatar
paulwise3
Senior Member
Posts: 1505
Joined: Tue Jan 06, 2015 10:56 am
Location: Eindhoven, Netherlands

Post by paulwise3 »

paulwise3 wrote: So I have to take my Mirage off the attick to see if there is a similarity to find...
And here is the (probably) final verdict:

Mate Problem 1
Mirage finds 1.Lf8+! in 0:52

Mate Problem 2:
Mirage finds 1.Kh3! in 2:02

So (to my relief) the Mirage must have the same program as Igor and Grandmaster, but running at 10 MHz.
And the 10 MHz Ivan seems to run an earlier H8 program version, less optimized. And therefor could well be a collector's item in itself!

Is there someone who could contact Ron Nelson and ask him about this?

Unraveled secrets regards,
Paul

EDIT: the pieces of my Ivan are not the black and white Excalibur pieces, but look like the usual Krypton pieces in grey and black, so that also adds to the conclusion that the 10 MHz Ivan is an earlier version.
2024 Special thread: viewtopic.php?f=3&t=12741
2024 Special results and standings: https://schaakcomputers.nl/paul_w/Tourn ... 25_06.html
If I am mistaken, it must be caused by a horizon effect...
Four.nine
Full Member
Posts: 911
Joined: Tue Mar 21, 2017 7:30 pm

Post by Four.nine »

paulwise3 wrote:
paulwise3 wrote: So I have to take my Mirage off the attick to see if there is a similarity to find...
And here is the (probably) final verdict:

Mate Problem 1
Mirage finds 1.Lf8+! in 0:52

Mate Problem 2:
Mirage finds 1.Kh3! in 2:02

So (to my relief) the Mirage must have the same program as Igor and Grandmaster, but running at 10 MHz.
And the 10 MHz Ivan seems to run an earlier H8 program version, less optimized. And therefor could well be a collector's item in itself!

Is there someone who could contact Ron Nelson and ask him about this?

Unraveled secrets regards,
Paul

EDIT: the pieces of my Ivan are not the black and white Excalibur pieces, but look like the usual Krypton pieces in grey and black, so that also adds to the conclusion that the 10 MHz Ivan is an earlier version.
Good work!

You now know probably as much as anyone about the Igor/Ivan/GM/Mirage story.

Definitely program Mirage=Igor=GM=IvanT2. They’re all the same thing, marketed in different plastic.

As you state: IvanT (10 MHz) is an earlier (considerably weaker program) H8 version.

Also, congrats on your working Mirage....mine is now in “Stall” mode and I will now use as a door stop.
User avatar
paulwise3
Senior Member
Posts: 1505
Joined: Tue Jan 06, 2015 10:56 am
Location: Eindhoven, Netherlands

Post by paulwise3 »

Hi 4.9,

Sorry to hear about your Mirage.
But there are people around who can repair such a machine, so I hope your doorstop can promote to chesspartner again.

Nice doorstop regards,
Paul
2024 Special thread: viewtopic.php?f=3&t=12741
2024 Special results and standings: https://schaakcomputers.nl/paul_w/Tourn ... 25_06.html
If I am mistaken, it must be caused by a horizon effect...
qortahelots

Blincoe

Post by qortahelots »

Hein also lists the buyers of the first 11 Rev II's. Among them a certain mister S. Blincoe Wink.
Post Reply