The biggest plastic opening book

[Reinfeld]

Montreux by a mile:

Claimed book size (half-moves):

Mephisto Montreux 380,000
Berlin Pro 180,000
Novag Star Diamond 123,000
Saitek RISC 2500 100,000
Mephisto Master 50,000
Designer 2325/2265 28,000
Mephisto Milano 25,000
Novag Citrine 24,000
Super Connie 20,000
Novag Obsidian 8,900
GK 2100 6,000
Explorer Pro 6,000
2250 XL 6,000

I know - I'm calling the Citrine plastic, but remember, it's a strange beast, and the wood is close to plastic - no modules. I left out a few Mephisto variants at the lower end.

When you get into big wood (TASC R40), the numbers climb out of sight. But it's interesting that the Montreux has a bigger book than both versions of TASC R30.

I assume processing power plays a larger role than the books, but I can't help noticing Montreux's enormous number. Perhaps that accounts for the slight difference in rating with the RISC 2500 (considering the identical processing power).

Another random thought: I suspect that the bigger book plays a bigger role on lower-end machines. Perhaps this helps the Fidelity models punch above their weight. Ditto for old Super Connie.

Would you have guessed that the Mephisto Milano has a bigger book than the Citrine? I didn't. Another reason to put the Citrine in the plastic category. (How do you justify building a wooden Star Diamond and cutting its book by 80 percent? The fancy machine ought to be the one with the extras.)

I was fiddling around with my recently acquired Milano tonight. Tried to play the Benko Gambit against it (1. d4 Nf6 2. c4 c5 3. b5). Milano had only one response: 3...a4. Certainly one of many, but it was the only move in the machine's database. No provision for accepting the gambit, the main theoretical battleground. Selection and omission - fascinating.

- R.

[Steve B]

Informative post
actually its nice to see book sizes all in one place..will be a good reference guide..Thanks

of course i imagine the plastic Phoenix Resurrection II modules with the Noomen books or the Shredder and Hiarcs engine books are larger but interestingly i cant find any information as to the actual book sizes for these engines as used in the Res II modules

Open Book Regards
Steve

[Peter Grayson]

Montreux by a mile:

I was fiddling around with my recently acquired Milano tonight. Tried to play the Benko Gambit against it (1. d4 Nf6 2. c4 c5 3. b5). Milano had only one response: 3...a4. Certainly one of many, but it was the only move in the machine's database. No provision for accepting the gambit, the main theoretical battleground. Selection and omission - fascinating.

- R.

One of the features I particularly liked with the Richard Lang series from Mephisto was that the user could punch in their own opening book to take preference over the provided book but if there was a continuation, it would also drop back into the provided book when the user line was exhausted. Very useful for practicing certain variations. I guess the extra features added to the cost of the unit.

PeterG

[IvenGO]

Kaplan's computers with 100.000 half-moves book has been missed in the list. Also RS2250XL has 20.000

About Polgar's/Milano's - the example with Benko gambit line shows that computer's book was made according to unit's playing style (btw I found myself 3...a4 as a best reply).

But also there are some bugs in it {thats more sad}:

Kings gambit (human plays white): 1.e4, e5 2.f4, Qh4+?? 3.g3, Qe7 4.Nc3

and from this position I won 4-0 (2 games/5 min & 2 games/30 min) no matter of black's reply.

[Reinfeld]

Kaplan's computers with 100.000 half-moves book has been missed in the list. Also RS2250XL has 20.000

I had looked here for the claimed book on the 2250 XL, but the entry says the number is unverified:

http://www.schach-computer.info/wiki/in ... ion_2250XL

You're right about neglecting Kaplan. Numbers for the key plastic:


Radio Shack 2150 - 100,000
Radio Shack 2150L - 17,000

[Reinfeld]

Perhaps that accounts for the slight difference in rating with the RISC 2500 (considering the identical processing power).

Well, then there's this from SSDF:

Mephisto Montreux (ARM 14 MHz 512K) - 18.5
RISC 2500 (ARM2 14 MHz 128K) - 21.5


- R.

[Larry]

Regarding the huge opening book of the Montreux, is that 380,000
half moves verified by Mephisto, or is it just someone's guess that has
been passed around until it was accepted?
I'm asking this because I remember a very knowledgeable poster on this
forum explaining the difference between the risc2500 and the Montreux.
Here are his words from June 11, 2010:
-------------------------------------------------------------------------------
"As a curiosity, the software for RISC2500 is 128kb and used the full 128kb EPROM, while the software for Montreux is 150kb and the remaining 106kb of the 256kb EPROM is blank/empty. So the Montreux software is only 22kb bigger than RISC2500 which most likely accounts for the larger opening book."
------------------------------------------------------------------------------
The guy used to post under the name of Kostea. Note the last sentence,
in which he says the Montreux software is only 22kb bigger than the
risc2500 software. Is 22kb enough space to cram in 280,000 half moves?
It seems to me that if Mephisto wanted to make the Montreux stronger
than the earlier risc2500, it would make way more sense to overclock
and/or include 2mb of ram in all units at the factory, instead of the
paltry 128kb that the Montreux's had. Monstrous opening book sizes
will do little to help playing strength if the valuation function once the
book runs out can't take advantage of a position deemed better by a
professional chess player.
Larry
PS Here's the thread I'm referring to:
http://www.hiarcs.net/forums/viewtopic. ... highlight=

[Reinfeld]

Larry -

The quick answer: I took the 380,000 figure from the wiki-elo entry. Is that a reliable source? I don't know.

The Montreux manual (via the great Zanchetta) offers little help:

"Uses true 32-bit RISC processor, with 128Kb of high-speed static RAM."

[This appears to refer to the initial version. There are 512 kb and 2 mb versions of Montreux, per wiki-elo...but the manual doesn't address the SIZE of the opening book...]

[Reinfeld]

And again, considering the SSDF records of Montreux vs RISC 2500, its hard to see what's going on:

RISC 2500 (128 KB) - 21.5
Montreux (512 KB) - 18.5

That's not many games for a good sample, but the "smaller" machine with the (allegedly) smaller book appears to beat the bigger one, which was (according to lore) built to beat its predecessor. So what does that mean?

[Steve B]

Larry -

The quick answer: I took the 380,000 figure from the wiki-elo entry. Is that a reliable source? I don't know.

The Montreux manual (via the great Zanchetta) offers little help:

"Uses true 32-bit RISC processor, with 128Kb of high-speed static RAM."

[This appears to refer to the initial version. There are 512 kb and 2 mb versions of Montreux, per wiki-elo...but the manual doesn't address the SIZE of the opening book...]

Hi Guys
having personally and painstakingly gone through each and every line and move of the Montreux book ..i can say for a fact certain there are in fact 380k half moves in its book..actually 380,023 to be exact


Seriously though..
the box for the Montreux also does not mention the book size
turning to the literature of the day ..a 1995 Saitek/Mephisto catalogue does show the Montreux with a 380k book size
It is interesting as De Konings King program which is in the Montreux and Risc 2500 is of course also in the Tasc R units and the R30 only has a 200k book size

Interestingly...when i research the Montreux in Selective Search i come across this from Selective Search Issue 61(Supplement) Dec.95-Jan. 96 page 5
"80,000 position opening book"
could the 380k be a typo??


discovering this discrepancy i looked up the Montreux in the CCR's
sadly the CCR's ended in 1996 so they only briefly discuss the Montreux and did not mention the book-size

in skimming through all of the literature i did come across another computer which had a 380,000 book size
the legendary Saitek Turbo Star 2600..right there in the Catalogue..380k
one small detail though..the Computer was

Never Actually Released For Sale Regards
Steve

[Steve B]

And again, considering the SSDF records of Montreux vs RISC 2500, its hard to see what's going on:

RISC 2500 (128 KB) - 21.5
Montreux (512 KB) - 18.5

That's not many games for a good sample, but the "smaller" machine with the (allegedly) smaller book appears to beat the bigger one, which was (according to lore) built to beat its predecessor. So what does that mean?

i guess it means that the book size doesn't matter all that much when playing two computers against each other that have basically the same engine and hardware
basically the Risc and Montreux have the same King engines although released 3 years apart
Selective Search shows the ratings for the two to be very close with the difference due largely to the hash table size...

Montreux(128k)-2210
Risc 2500(128k)-2191
Risc 2500(512k)-2231

also of note
the Montreux was released with only 128k ram and it was not factory upgradeable as was the Risc 2500
versions of the Montreux with more then 128k are around(and in fact tested by SSDF we can see) but these would be computers modified by the owners

Book Knowledge alone does not make Johnny Smart regards
Steve

[Larry]

however in skimming through the catalogues i did come across another computer which had a 380,000 book size
the legendary Saitek Turbo Star 2600..right there in the Catalogue..380k
one small detail though..the Computer was

Just clicked a few links, including this one:
http://www.schach-computer.info/wiki/in ... ostar_2600

This machine and the Megathon 2400 never made it to market possibly
because the market was becoming flooded around that time. Instead they
built the TS2600 into a module... the RISC 1 of Ed Schroder, with a much
smaller opening book. The risc1 processor and clock speed seem to match
with the risc2500 and montreux. It's all history now, but rest assured
TS2600's, had they been released in the familiar washboard design of
the Montreux, would be keenly sought after by now.

Larry

[chessguru]

Hi,

Larry -

The quick answer: I took the 380,000 figure from the wiki-elo entry. Is that a reliable source? I don't know.

hmmm, I think you can trust her.

But, I think you are in doubt. OK, what says the manufacturer about all this? Let's take a look at the product catalog from the year 1995.





We find: Mephisto Montreux.





Here's an overview:




Do not doubt. Just ask whatever you want to know.

Best regards,
Micha