Collector's Corner : Cosmic / Maestro Stalemate Bug (Steve?)

[Steve B]

hi Jon
nice recollections indeed!

i will take the liberty of adding some of my own scattered thoughts to your comments

Sensor chess...
had the dubious distinction of requiring the HEAVIEST Adaptor ever released for sale
i think the adaptor weighed even more then the computer itself
always liked the leds under the playing surface..first and i think perhaps the only computer to ever try that

Elegance...
it truly is an Elegant looking computer
and you are spot on regarding its opening book
funkiest opening book in any Fidelity
offering 1.P-F4's and 1..P-F5's not to mention the more then occasional 1.G2 and 1..G6
I usually find myself force feeding it a CB9-CB16 book module only to get it to play lines i am more familiar with!

Mark V....
aaah..a true masterpiece
the most feature rich chess computer ever released for sale
you can tell it was designed by a chess player(Levy) because it leaves no option wanting
personally however i always found it a bit awkward playing a complete game just using the LCD chess board so i always play it with the sensor board attached...but thats only me

Sensory 9...
personally never one of my favorite Fidelity's but a must have for any serious collection of course
i did appreciate all of the modules that were released for it but it just never grabbed me cosmetically

for some reason the Novag Savants leap to mind as i write this(i did warn you these would be random thoughts)

first touch sensory with lcd pieces and it would be considered a technical marvel even today if any of them were still working
i remember Cameron lamenting to me how awful he felt the day his Savant Royale(or was it his Savant II?)stopped working
if i recall correctly he described going to the tip to throw it away but just couldn't bring himself to do it!

finally i feel the strange and compelling need to mention here the Super 9..another classic collector's item then even as we speak is getting bid up to the sky on Ebay:

http://cgi.ebay.com/ws/eBayISAPI.dll?Vi ... &rd=1&rd=1

and of course its blood brother the Super 9 ..
Deluxe Regards
Steve

[JonP01]

Hi Steve,

I am sure I would have put a Savant on my list had I actually been able witness one playing chess. I remember one of Cameron's but perhaps it was the faulty one - I can't remember. The display was extremely dim and the pieces almost impossible to distinguish. I'm not sure if it was working or not.

Anyway, back to the subject of the topic and I am currently still trying to figure out if there is indeed any difference between the selective and brute force searches in this Maestro. My initial testing indicated there was no difference at all. Now I not quite so sure. The problem is that so long as the machine is actualy turned on, it is continually analysing the current position, even if it is the opening position. So I have to keep removing the batteries, re-inserting them and quickly setting the levels, search parameter and the position to see if there really is any difference. I am also trying to find some good composed mates without cooks to see if the brute force search gets them right.

This is turning out to be one of the toughest "analytical" tests I have ever performed on a chess machine. All morning gone and none the wiser...


Tough One Regards

Jonathan

[Robert Weck]

Hi Jonathan,

Hi Steve,
The Sensory 9 was the first machine I played against that was stronger than me, so that will always impress me whichever way you cut it

The same here!

I had a Super System III at this time, which was easy to beat for me, so i was looking for a new challenge.

About 1982, my friend Bernhard had a Sensory 9 and i said, if it defeats me at 5min/Blitz, then i will buy one!

So i had to sell my SSIII, to be able to afford the S9...


still happy owner of this S9 regards,
Robert

[JonP01]

Hi Robert,

Yes Sensory 9 was a breakthrough as far as I was concerned. Long live the O'Kelly Sicilian!

@Steve,

I spent the whole day trying to determine if there is any difference between the Saitek Maestro brute force and selective searches. In all cases the bahaviour seems identical - same search depths for a given amount of time, same range of moves selected, etc.

I did get one strange result though. A composed mate in 5 study took 1 hour and 47 minutes in brute force mode (completed the 15th move of depth 9). But in selective mode it took 2 hours (completed the 16th move of depth 9). But I et when I try the same problem today those results will reverse themselves - that is the sort of inconsistency I have found in my testing.

I don't quite understand the above though. In the end though, for practical purposes, I've yet to see a difference in either the moves it plays or playing strength. I'd love to know what this option actually does to the program, but it definitely does not make it brute force (it failed a couple of composed mates in 3, since it got to 7 ply and still missed them).

I'm even starting to wonder (half joking here, but only half) if this machine underclocks the processor when the batteries run low, especially as it has such a huge appetite for them. It might explain the slower response times the longer I carried out the tests.


None the wiser Regards

Jonathan

[JonP01]

Hi Steve,

Maybe I could ask you to do another quick test on your Siatek Cosmic if possible please? Clear the board and put a White King on D1, White Queen on E1 and Black King on D5.

White to move, 30 secs per move. Let the Cosmic play white and see if it knows how to mate (don't help it one little bit - please keep the black King in the centre where possible).

Please let me know if it knows how to complete the mate

[Steve B]

Hi Steve,

Maybe I could ask you to do another quick test on your Siatek Cosmic if possible please? Clear the board and put a White King on D1, White Queen on E1 and Black King on D5.

White to move, 30 secs per move. Let the Cosmic play white and see if it knows how to mate (don't help it one little bit - please keep the black King in the centre where possible).

Please let me know if it knows how to complete the mate

hi Jon
i can confirm that Cosmic was able to mate

it did not mate in a very efficient manner though boxing in the King with the Queen but rather in a very circuitous route which took about 25-30 moves

it did announce a Mate in two when it was close to delivering the mate

Cosmic Rulez Regards
Steve

[JonP01]

Hi Steve,

Thank you for your patience

Now I am wondering if the Maestro is indeed the same program as the Cosmic. I just had my cat play black against Maestro in the same position and my cat managed to draw by the 50 move rule.

How can Saitek market this machine as "4 out of 5" playing strength (same as Advanced Travel). It's about 600 - 700 points weaker by my reckoning!

[Event "?"]
[Site "?"]
[Date "2007.05.10"]
[Round "?"]
[White "Saitek Maestro"]
[Black "Hilary the Cat"]
[Result "1/2-1/2"]
[SetUp "1"]
[FEN "8/8/8/3k4/8/8/8/3KQ3 w - - 0 0"]
[PlyCount "100"]

1. Qc3 Ke4 2. Kc2 Kd5 3. Kc1 Ke4 4. Kd2 Kd5 5. Qc8 Ke5 6. Qa6 Ke4 7. Qf1 Kd5 8.
Qa6 Ke4 9. Qf6 Kd5 10. Kc3 Ke4 11. Qf2 Kd5 12. Kb4 Ke5 13. Qf3 Kd4 14. Ka5 Ke5
15. Qh1 Kd4 16. Kb4 Ke5 17. Kc5 Kf4 18. Qg2 Ke5 19. Qg6 Kf4 20. Kd6 Ke3 21. Qg2
Kd4 22. Qf3 Kc4 23. Kc6 Kd4 24. Kb7 Ke5 25. Ka6 Kd4 26. Kb6 Ke5 27. Kc6 Kd4 28.
Kd7 Ke5 29. Kc8 Kd4 30. Kc7 Ke5 31. Kd7 Kd4 32. Qa3 Ke4 33. Ke8 Kd5 34. Kd7 Ke4
35. Kd8 Kd5 36. Qb3+ Ke5 37. Ke7 Ke4 38. Qg3 Kd4 39. Kd8 Ke4 40. Kd7 Kd4 41.
Ke7 Ke4 42. Ke8 Kd5 43. Qg2+ Kd4 44. Kf7 Ke5 45. Kg6 Kd4 46. Qg3 Ke4 47. Qf2
Kd3 48. Qb2 Ke4 49. Qa3 Kf4 50. Kh6 Ke4 1/2-1/2

[Steve B]

Hi Steve,

Thank you for your patience

Now I am wondering if the Maestro is indeed the same program as the Cosmic. I just had my cat play black against Maestro in the same position and my cat managed to draw by the 50 move rule.

How can Saitek market this machine as "4 out of 5" playing strength (same as Advanced Travel). It's about 600 - 700 points weaker by my reckoning!

[Event "?"]
[Site "?"]
[Date "2007.05.10"]
[Round "?"]
[White "Saitek Maestro"]
[Black "Hilary the Cat"]
[Result "1/2-1/2"]
[SetUp "1"]
[FEN "8/8/8/3k4/8/8/8/3KQ3 w - - 0 0"]
[PlyCount "100"]

1. Qc3 Ke4 2. Kc2 Kd5 3. Kc1 Ke4 4. Kd2 Kd5 5. Qc8 Ke5 6. Qa6 Ke4 7. Qf1 Kd5 8.
Qa6 Ke4 9. Qf6 Kd5 10. Kc3 Ke4 11. Qf2 Kd5 12. Kb4 Ke5 13. Qf3 Kd4 14. Ka5 Ke5
15. Qh1 Kd4 16. Kb4 Ke5 17. Kc5 Kf4 18. Qg2 Ke5 19. Qg6 Kf4 20. Kd6 Ke3 21. Qg2
Kd4 22. Qf3 Kc4 23. Kc6 Kd4 24. Kb7 Ke5 25. Ka6 Kd4 26. Kb6 Ke5 27. Kc6 Kd4 28.
Kd7 Ke5 29. Kc8 Kd4 30. Kc7 Ke5 31. Kd7 Kd4 32. Qa3 Ke4 33. Ke8 Kd5 34. Kd7 Ke4
35. Kd8 Kd5 36. Qb3+ Ke5 37. Ke7 Ke4 38. Qg3 Kd4 39. Kd8 Ke4 40. Kd7 Kd4 41.
Ke7 Ke4 42. Ke8 Kd5 43. Qg2+ Kd4 44. Kf7 Ke5 45. Kg6 Kd4 46. Qg3 Ke4 47. Qf2
Kd3 48. Qb2 Ke4 49. Qa3 Kf4 50. Kh6 Ke4 1/2-1/2

well Jon
you know me,i always prefer to look at the bright side when it comes to our much maligned little dedicated computers

in your case....
at least the Maestro recognized the draw by the 50 move rule!


otherwise you could still even now be playing the position out ad infinitum

Impressed Regards
Steve

[JonP01]

Well, another glass half full way to look at it is that unless a computer has been given specific instructions on how to mate, it is almost impossible to do it efficiently unless the machine can look ahead far enough into the position to see the actual mate. My guess is Saitek filled the 16K up with other things and had no room left for these things - perhaps on the premise that such positions are fairly rare in actual play (very few of my computer games ever end with King and soon-to-be-promoted pawn versus King). Of course, the glass half empty is that they forgot about these things...

I have made an interesting discovery about this Maestro. It is clocked at 6 Mhz - not the 7 Mhz of the Cosmic. Perhaps Saitek did this on account of the very high battery consumption of this unit with it's (admittedly extremely effective) backlight.

I went to the supermarket yesterday to pick up a 12 pack of "AAA" sized batteries for this thing. The batteries put out so much collective energy in their tiny package that they actually set off the supermarket anti-theft scanner as I passed through it towards the entrance. Before I could produce my docket as proof of payment I was whisked away by the local constabularly and placed in the overnight lockup.

The next morning the Duty Sergeant interviewed me and was wondering if I was attempting to traffic in "AAA" batteries, since these were not "commercial" quantities for "private use". After I explained to him that they were for my Saitek Maestro, he was full of apologies, gave me a hot meal and drove me straight home.

Now that's Service Regards

Jonathan