Human game Appleshampogal vs Novag Obsidian

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.
User avatar
fourthirty
Full Member
Posts: 763
Joined: Fri Dec 06, 2013 8:46 pm
Location: San Francisco

Post by fourthirty »

Steve B wrote:Following Jon's Suggetions i gave the Superior(?)Obisidian three middle game positions
The positions were all selected from the games in Spacious Mind's now cult classic thread on rating the chess computers:
http://hiarcs.net/forums/viewtopic.php? ... sc&start=0

each position offered several reasonable move choices
there are no forced wins or winning tactical shots or forced mates
basically just solid positional play

i set the level to FIXED DEPTH 7 PLY(level FD7)
I recorded the move selected...time ..and Eval

POSITION 1
Botvinnik, M. vs Grob, H. , Zurich 1956
[fen]rnb1kb1r/1p3p1p/pq2pp2/2pP4/4P3/2N2N2/PP3PPP/R2QKB1R w KQkq - 0 10[/fen]
WHITE TO MOVE

Move- Qc2
Time-3 Min. 43Sec.
Eval- +.25
Steve/Kat,

I finally had an opportunity to set up the POSITION 1 test position this afternoon on one of my "inferior" Obsidians.

The result was:
Move: Qc2
Time: 3 Min. 19Sec.
Eval: -.25

I repeated this three times with the same result. Not sure why my unit is slightly favoring black's position.

I'm departing on a trip to Europe early tomorrow for work so I won't be able to post the results of the other two positions until next weekend.

I noticed that our starting point in the "Botvinnik vs Grob" game was slightly different than was posted in Spacious Mind's trial. Was this by design? Curious why we didn't start after 6...c5?

Inferiority complex regards,
Greg
User avatar
Steve B
Site Admin
Posts: 10140
Joined: Sun Jul 29, 2007 10:02 am
Location: New York City USofA
Contact:

Post by Steve B »

fourthirty wrote:
Steve B wrote: POSITION 1

Move- Qc2
Time-3 Min. 43Sec.
Eval- +.25
Steve/Kat,

I finally had an opportunity to set up the POSITION 1 test position this afternoon on one of my "inferior" Obsidians.

The result was:
Move: Qc2
Time: 3 Min. 19Sec.
Eval: -.25

I repeated this three times with the same result. Not sure why my unit is slightly favoring black's position.

I'm departing on a trip to Europe early tomorrow for work so I won't be able to post the results of the other two positions until next weekend.

I noticed that our starting point in the "Botvinnik vs Grob" game was slightly different than was posted in Spacious Mind's trial. Was this by design? Curious why we didn't start after 6...c5?

Inferiority complex regards,
Greg
Hi Greg

i did not use any of the starting positions
they are too early in the game and normal developing moves would be chosen ..i wanted mid-game positions ..a bit more complicated ..so we can see if the two versions are actually playing differently

looks like we have a slight difference in time..about 10% faster for the "Inferior"
not sure about the eval?
are you reporting the eval ..after the move.. by pressing the info button?
it would show the score from blacks position at that point which would of course be the same but with +/- reversed

Have A Safe Trip Regards
Steve
BillT
Member
Posts: 132
Joined: Sat Jun 09, 2012 9:31 pm
Location: Norfolk UK

Post by BillT »

Morning All.
Fascinating topic this, and I'd like to share with you my thoughts and info.
I recently picked up an Obsidian,as new - complete v.nice.
In this thread there is a question as to whether there are 2 different versions ("next best" working or not with opening flexibility or not) of the Obsidian, and that removing the "next best" function MAY make a difference to the playing strength (superior/inferior) The test positions could highlight any difference (see my results below).

So here are the details of my shiny new Obsidian.

Model SW-661
Serial no 237627
small black sticker on box and base of computer 13W35
Manual ref Article 1016, 85-661-004-001
"Next best" does NOT function.

Immediately I notice that the Model is SW-661 not the 1016 as shown on Alains on his spacious mind website for example ? is this relevant ??
I assume this could be from when Novag ceased and I guess production was continued by another manufacturer under license??

Anyway I setup the test positions on my "Inferior" Obsidian.

Test 1

Q-C2
3mins 19 secs
+0.25

11% Quicker than Steve & Kats , identical to Greg's

Test 2

B3
2 mins 14secs
+0.04

10% Quicker

Test 3

Q-E7
2 mins 44 secs
+0.04

9% Quicker than Steves, but as good as the same as Kat's ?????



It seems the "inferior" may have had program tweaks or processor speed upgrade to make this average of 10% difference.

For reference heres what is printed on the IC which is visible on opening up the unit. its a very small square IC (not sure if its the CPU or ROM ) I suspect its the CPU.... and you'll need a jewellers loupe to read it .....
014A12FPV
NOVAG
10073
0084

Interested to know what others think of all this......
The plot thickens
regards
Bill

Ps
Greg...I noticed that the Obsidian shows the score for the side to move....so as soon as it makes the move Q-c2 it will then show -0.25 score
User avatar
Steve B
Site Admin
Posts: 10140
Joined: Sun Jul 29, 2007 10:02 am
Location: New York City USofA
Contact:

Post by Steve B »

BillT wrote: Anyway I setup the test positions on my "Inferior" Obsidian.

Test 1

Q-C2
3mins 19 secs
+0.25

11% Quicker than Steve & Kats , identical to Greg's

Test 2

B3
2 mins 14secs
+0.04

10% Quicker

Test 3

Q-E7
2 mins 44 secs
+0.04

9% Quicker than Steves, but as good as the same as Kat's ?????



It seems the "inferior" may have had program tweaks or processor speed upgrade to make this average of 10% difference.


Interested to know what others think of all this......
The plot thickens
regards
Bill

Informative post Bill

well it seems that the Inferior's are actually faster by about 10%
in a fixed ply game this would make no difference but under tournament time controls or Avg.Time per move levels it means the Inferiors would search deeper ..10% deeper
which would make it somewhat stronger then the Superiors
the difference in strength would be more pronounced in games with shorter time controls such as in "Active" Chess(30 Sec./Avg or 30 minutes entire game)
Its beginning to look like the Inferiors are really Superior after all
now isnt that something!


we need to rename the two versions moving away from any pejorative undertones

Obsidian A=Next best and more varied book
Obsidian B=10% faster ..No Next best.. less varied book

Role Reversal Regards
Steve
User avatar
Monsieur Plastique
Senior Member
Posts: 1014
Joined: Thu Jul 03, 2008 9:53 am
Location: On top of a hill in eastern Australia

Post by Monsieur Plastique »

So it seems we have consistency in that the ones with non-working Next Best and inferior book are actually later units with much higher serial ranges and slightly faster. Conversely the older ones (possibly only the 12XXXX serial range?) have a working Next Best and are slower. But they definitely have a significantly better, if not still flawed opening book.

That 10% could actually be hardware (but not clock speed) rather than program related and could even be consistent with my hypothesis that there was a change in chip hardware for the later production ones and that hardware change has possibly caused the issue (I'd be willing to bet something along similar lines could be partially responsible for some of the issues in the last Morsch dedicated machines that did not exist when he originally wrote the programs for Saitek back in the 90s).

I realise that there is likely no official documentation to support any hardware "upgrades" in the higher serial units, but then again regular readers would be aware of the inconsistencies I've found in the Novag specs on the boxes and manuals of the machines made over the last 5 years or so. Novag might have sourced a new chip that is the same speed but has faster internal memory access for example. That could account for the speed difference.

That said, I'm not really sure 10% is going to make any significant difference in playing strength anyway. You usually need something more like around 30% or more to really "feel" a difference.

For my own part I would take the slower older one with the better book over the newer faster one. From the games I have seen the older one play this last week, the difference in the openings book is more then adequate compensation for any shortcomings versus the final units.

One day someone will be brave enough to perform surgery on both types to discover the truth :wink: Mind you, I do not recommend doing so out of curiosity - only if the machines are ill to begin with.
Chess is like painting the Mona Lisa whilst walking through a minefield.
User avatar
Steve B
Site Admin
Posts: 10140
Joined: Sun Jul 29, 2007 10:02 am
Location: New York City USofA
Contact:

Post by Steve B »

Monsieur Plastique wrote:

That said, I'm not really sure 10% is going to make any significant difference in playing strength anyway. You usually need something more like around 30% or more to really "feel" a difference.
as i mentioned early on in this thread
it made no sense for Novag to cripple the Obsidian without getting something in return
at least we have discovered that something was enhanced when Novag Crippled the Next Best and opening books
the machine was made faster

now "significant" is up to interpretation
in 40/2 games i imagine it would produce about 7-10 Elo
but in shorter time controls the difference would be greater
maybe another 5-10 Elo
is that Significant?
for some i think it is
personally i have never given much weight to elo differences of 50 Elo or less ..and as a collector
ratings were never very important to me..
regardless of the strength..if they built them ...

I would Come Regards
Steve
User avatar
fourthirty
Full Member
Posts: 763
Joined: Fri Dec 06, 2013 8:46 pm
Location: San Francisco

Post by fourthirty »

Steve B wrote:
fourthirty wrote: The result was:
Move: Qc2
Time: 3 Min. 19Sec.
Eval: -.25
Greg
are you reporting the eval ..after the move.. by pressing the info button?
it would show the score from blacks position at that point which would of course be the same but with +/- reversed
Steve
Steve / Bill,

Yes, I was reporting the eval ..after the move.. by pressing the info button. I did not realize that the Obsidian shows the score for the side to move.

Therefore, the correct result is:
Eval: +.25

Will attempt the other positions this weekend on my two inferiors, or Obsidian B units as they are now called.

Greg
User avatar
fourthirty
Full Member
Posts: 763
Joined: Fri Dec 06, 2013 8:46 pm
Location: San Francisco

Post by fourthirty »

BillT wrote: Test 1
Q-C2
3mins 19 secs
+0.25

Test 2
B3
2 mins 14secs
+0.04

Test 3
Q-E7
2 mins 44 secs
+0.04
I have now had an opportunity to test all three positions with both of my "B" Obisidans:

Obsidian B (#1)
Model 1016
83-661-000
Date Code = 07W43 (on bottom of Obsidian not the box)
Serial number 202541
My manual is coded 85-661-004 (which includes the addendum 86-661-003 regarding the Obsidian not having the NEXT BEST function)
manual copyright-2002

Obsidian B (#2)
Model 1016
83-661-000
Date Code = 08W38 (on bottom of Obsidian and on the box)
Serial number 212636
Manual is coded 85-661-004 (does not include the addendum 86-661-003 regarding the Obsidian not having the NEXT BEST function)
manual copyright-2002

My results were identical to Bill's. To clarify, the evaluation of +0.04 in position #3 would be from the perspective of Black, correct?. I receive "-00.04" on the Info screen after executing the black move of Q-E7.

Test 1
Q-C2
3mins 19 secs
+0.25

Test 2
B3
2 mins 14secs
+0.04

Test 3
Q-E7
2 mins 44 secs
+0.04

Regards from,
The Obsidians formerly known as Inferior...
User avatar
Monsieur Plastique
Senior Member
Posts: 1014
Joined: Thu Jul 03, 2008 9:53 am
Location: On top of a hill in eastern Australia

Post by Monsieur Plastique »

Thanks Greg. I think all the contributions have wrapped this one up, though the sobering news appears to be that Novag were making "B" Obsidians for at least 6 years! We have your "B" with a late 2007 build date right up to Bills with a late 2013 build date. I doubt whether there would even have been any production runs after Bill's too.
Chess is like painting the Mona Lisa whilst walking through a minefield.
User avatar
Monsieur Plastique
Senior Member
Posts: 1014
Joined: Thu Jul 03, 2008 9:53 am
Location: On top of a hill in eastern Australia

Post by Monsieur Plastique »

BillT wrote:Immediately I notice that the Model is SW-661 not the 1016 as shown on Alains on his spacious mind website for example ? is this relevant ??
This would be consistent with the buyout of Novag by Solar Wide Industries (hence the SW prefix). The Solar Wide units would be the later units as confirmed by your build date and high serial number.
Chess is like painting the Mona Lisa whilst walking through a minefield.
User avatar
Steve B
Site Admin
Posts: 10140
Joined: Sun Jul 29, 2007 10:02 am
Location: New York City USofA
Contact:

Post by Steve B »

Monsieur Plastique wrote:Thanks Greg. I think all the contributions have wrapped this one up
Yes Obsidian's A and B are now in the record books
new information never before known

still...something is gnawing at my gut
studious readers of this thread will notice
there was a difference in solve times for Positon 3 between the Obsidian A's
my A returned a slower solve time then PoGal's A
3.01 Vs. 2.43
i wonder if there were TWO A versions??
or was that just an Anomaly??

i guess this is something we will never know....

A Puzzlement For The Ages Regards
Steve
User avatar
fourthirty
Full Member
Posts: 763
Joined: Fri Dec 06, 2013 8:46 pm
Location: San Francisco

Post by fourthirty »

Monsieur Plastique wrote:Thanks Greg. I think all the contributions have wrapped this one up, though the sobering news appears to be that Novag were making "B" Obsidians for at least 6 years!
You're welcome Monsieur Plastique. I believe the Obsidian first came on the market in 2005. It appears the "A" version may have only been produced for a year or two, which may make them fairly difficult to find!
trontreez

Post by trontreez »

Yes your quite right! There were indeed two versions of the Novag Obsidian. It's been well known for some time now.

The initial release featured the functionality of the next best key, pre-programmed positions/games and the more varied opening book. It was running on older inferior hardware running @ 2x16Mhz.

The second release featured an upgraded program with a "tighter" opening book on a H8-chip @ 20Mhz. The second release being much stronger (by 50-100 elo) of course!

This makes the weaker, albeit rarer, version somewhat more desirable than the stronger secondary release :roll:
mrcdoc
Member
Posts: 49
Joined: Thu Apr 19, 2012 4:12 am
Location: California, USA
Contact:

Post by mrcdoc »

Yes, I confirm. It exists minimally 2 different models on the hardware side:
- one with a H8/3214 microprocessor and a clock at 2x16MHz;
- and a second with a H8/3614 and a clock at 20MHz.

Maurice
sahana
Member
Posts: 3
Joined: Wed Nov 05, 2014 2:58 am

Post by sahana »

Nice share....
Post Reply