MCGE Classic London Book v Champion Book

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
Yarc
Senior Member
Posts: 1051
Joined: Sat Apr 11, 2009 8:13 am
Location: United Kingdom

MCGE Classic London Book v Champion Book

Post by Yarc »

I have just completed another ten game match between the StarDiamond(SD) and the Millennium ChessGenius Exclusive (MCGE) at 40m/2h. This time I set the MCGE to use the Classic London Book and got what I think is a surprising result.

In the ten game against the SD using the Champion book the result was MCGE +5 =5 (75%)
Playing White: +4 =1
Playing Black: +1 =4

ELO: 2364

Against the Magellan at the same time control and same book:
MCGE +7 =1 -2 (75%)
Playing White: +3 -2
Playing Black: +4 =1

ELO: 2397

When using the Classic London Book against the SD
MCGE +8 =1 -1 (85%) A much larger difference than I expected!
Playing White: +5
Playing Black: +3 =1 -1

This gives a rating difference of around 301!

ELO: 2474

This is an extraordinary result for the MCGE and I think it must have had the luck of the chosen openings giving it this much higher rating. I was expecting a figure closer to 2400 at this time control which would still have been a little higher than with the Champion book. Could this be just noise and proving many more games needed or is using the Classic book that much better against computer opponents?

Regards
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 »

Actually we went through this exercise when the MCGP was released
it was thought that perhaps the two different books would show a significant rating difference
there is a thread about this here somewhere....

IIRC..as it turned out..there was no significant difference when a larger number of games were played against a variety of opponents

my guess is that's what is happening with the MCGE

Mix It Up Regards
Steve
User avatar
Yarc
Senior Member
Posts: 1051
Joined: Sat Apr 11, 2009 8:13 am
Location: United Kingdom

Post by Yarc »

Steve B wrote:Actually we went through this exercise when the MCGP was released
it was thought that perhaps the two different books would show a significant rating difference
there is a thread about this here somewhere....

IIRC..as it turned out..there was no significant difference when a larger number of games were played against a variety of opponents

my guess is that's what is happening with the MCGE

Mix It Up Regards
Steve
Hi Steve,

Yes I actually made those tests with the MCGP against my SD and posted the results here. I started to do the same with the Magellan, but never got round to completing it.

Yes, I would think against a large number of opponents a more accurate figure would be achieved. Unfortunately, I only have two other programs strong enough to offer any form of resistance against the MCGE. I don't think 1800 to 1900 rated machines would offer enough.

Maybe a match with HIARCS set to 2400 might be worth a shot?

Regards Ray
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 »

Yarc wrote:
Maybe a match with HIARCS set to 2400 might be worth a shot?
Hiarcs does appear in the Rev I and Rev II dedicated computers ,but
I guess you mean playing the Hiarcs PC engine and somehow setting it to play at a weaker level?
sorry I cant opine on that sort of thing..not knowledgable enough about PC chess engines to know if the rating of 2400 is reliable when setting it to play at a weaker level

perhaps others can say if that would be a good test

Outta My League Regards
Steve
User avatar
Yarc
Senior Member
Posts: 1051
Joined: Sat Apr 11, 2009 8:13 am
Location: United Kingdom

Post by Yarc »

Steve B wrote: Hiarcs does appear in the Rev I and Rev II dedicated computers ,but
I guess you mean playing the Hiarcs PC engine and somehow setting it to play at a weaker level?
sorry I cant opine on that sort of thing..not knowledgable enough to know if the rating of 2400 is reliable when setting it to play at a weaker level

perhaps others can say if that would be a good test

Outta My League Regards
Steve
Unfortunatetly I don't own either a Rev I or II as otherwise would be trying that now.

I also have no experience of using the PC Hiarcs ELO setting. I do know that it is calibrated to the speed of the PC it's running on but that is all.

I normally prefer playing one dedicated against another, but that is the problem with having only one very strong machine like the MCGE. The other option is to lower it's speed to 50Mhz and pit it against my Saitek Cosmos which is rated about 2020, or maybe the Talking Chess Academy at 1900 or maybe at a push the Grandmaster Excalibur at 1800+

Regards
Ray
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 »

Yarc wrote:
I normally prefer playing one dedicated against another
yes I do too
there seems to be a new trend where dedicated computers are being matched up against old and new PC engines
not my sort of thing but I guess others find it interesting
Yarc wrote:
The other option is to lower it's speed to 50Mhz and pit it against my Saitek Cosmos which is rated about 2020, or maybe the Talking Chess Academy at 1900 or maybe at a push the Grandmaster Excalibur at 1800+


I don't see why you cant at least try that
if the results between the two books are still as striking as you find them to be at this point ..then we know you might be on to something very interesting

Book On Regards
Steve
User avatar
Yarc
Senior Member
Posts: 1051
Joined: Sat Apr 11, 2009 8:13 am
Location: United Kingdom

Post by Yarc »

Steve B wrote: I don't see why you cant at least try that
if the results between the two books are still as striking as you find them to be at this point ..then we know you might be on to something very interesting

Book On Regards
Steve
The only thing about reducing the speed of the MCGE is that it will then behave more like its predecessors. Having already played the MCGP against the SD, I might see a similar set of results. Perhaps then, the MCGE at 50Mhz against the Talking Chess Academy and Excalibur Grandmaster would be more interesting?

Regards
Ray
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 »

Yarc wrote:
Steve B wrote: I don't see why you cant at least try that
if the results between the two books are still as striking as you find them to be at this point ..then we know you might be on to something very interesting

Book On Regards
Steve
The only thing about reducing the speed of the MCGE is that it will then behave more like its predecessors. Having already played the MCGP against the SD, I might see a similar set of results. Perhaps then, the MCGE at 50Mhz against the Talking Chess Academy and Excalibur Grandmaster would be more interesting?

Regards
Ray
your call Ray
you are doing all of the hard field work
the rest of us will just be ....

Spectators Regards
Steve
User avatar
Yarc
Senior Member
Posts: 1051
Joined: Sat Apr 11, 2009 8:13 am
Location: United Kingdom

Post by Yarc »

Steve B wrote: your call Ray
you are doing all of the hard field work
the rest of us will just be ....

Spectators Regards
Steve
Thanks. It's a very time consuming thing so don't expect results anytime soon, but it might prove interesting as you say. I may hold back until the Lang update comes through at the end of the month as this is likely to improve its rating a little and I will be eager to install it.

Regards
Ray
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 »

Yarc wrote:
Steve B wrote: your call Ray
you are doing all of the hard field work
the rest of us will just be ....

Spectators Regards
Steve
Thanks. It's a very time consuming thing so don't expect results anytime soon, but it might prove interesting as you say. I may hold back until the Lang update comes through at the end of the month as this is likely to improve its rating a little and I will be eager to install it.

Regards
Ray

very good
if your results still show a big difference in rating ...then perhaps in the spring we could play an online match (Lichess) and i'll operate the Rev I or Rev II Hiarcs
Rev I Hiarcs has an SSDF Rating of ...2771(40/2)
Rev II Hiarcs has an SSDF Rating of ...2923(40/2)

http://www.ssdf.bosjo.net/list.htm

Lets See How It Goes Regards
Steve
User avatar
Yarc
Senior Member
Posts: 1051
Joined: Sat Apr 11, 2009 8:13 am
Location: United Kingdom

Post by Yarc »

Steve B wrote: very good
if your results still show a big difference in rating ...then perhaps in the spring we could play an online match (Lichess) and i'll operate the Rev I or Rev II Hiarcs
Rev I Hiarcs has an SSDF Rating of ...2771(40/2)
Rev II Hiarcs has an SSDF Rating of ...2923(40/2)

http://www.ssdf.bosjo.net/list.htm

Lets See How It Goes Regards
Steve
Ok, sounds like a good plan.

I notice that so far on rating lists the MCG and MCGP are listed but not the MCGE. I guess it's too soon, not enough results.

Regards
Ray
lexman
Member
Posts: 121
Joined: Mon Oct 12, 2015 11:35 pm

Post by lexman »

I think Hiarcs at full strength on the Revelation 2 would be a walkover for Hiarcs. I played a short match between Hiarcs on a palm tungsten at 200mhz against chess genius on a tablet at 1.2ghz. Hiarcs won 5-1 with 2 lucky draws for the chessgenius.
Hiarcs 1% on the Rev 2 might be more interesting, its rated at 2369 on the wiki and none of the traditional dedicated comps have acquitted themselves at all well against it.
Of course the big match would be against the wundermachine for the strongest lang classic dedicated of all time title! Personally I don t think the title would change hands, but it would be a good measure of whether the new lang has made any progress.
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 »

lexman wrote: Of course the big match would be against the wundermachine for the strongest lang classic dedicated of all time title! Personally I don t think the title would change hands, but it would be a good measure of whether the new lang has made any progress.
Ah Yes..the legendary WunderMachine
I matched it against two Lang Engines
it defeated the Rev II London 4-1 and
ate the MCG's lunch at 3-0
it even beat the MCG at Knight odds(but lost at Rook odds)

Matches can be found here:
http://hiarcs.net/forums/viewtopic.php?p=74387#74387


Perhaps your right and a better match up Vs. the MCGE would be Wunder Boy

Lets See What The Future Brings Regards
Steve
User avatar
Yarc
Senior Member
Posts: 1051
Joined: Sat Apr 11, 2009 8:13 am
Location: United Kingdom

Post by Yarc »

Steve B wrote:
lexman wrote: Of course the big match would be against the wundermachine for the strongest lang classic dedicated of all time title! Personally I don t think the title would change hands, but it would be a good measure of whether the new lang has made any progress.
Ah Yes..the legendary WunderMachine
I matched it against two Lang Engines
it defeated the Rev II London 4-1 and
ate the MCG's lunch at 3-0
it even beat the MCG at Knight odds(but lost at Rook odds)

Matches can be found here:
http://hiarcs.net/forums/viewtopic.php?p=74387#74387


Perhaps your right and a better match up Vs. the MCGE would be Wunder Boy

Lets See What The Future Brings Regards
Steve
Steve, I know you're not interested in PC engines v dedicated. I don't really like the idea. However, out of curiosity started a game with PC Hiarcs set to 2400 ELO. The first game and probably the last, is showing Hiarcs looking 3 to 4 extra ply ahead and gaining an early advantage. Doesn't seem right somehow, but then what did I expect ? I don't think this is proving anything. I'll continue to the end, but not convinced this was such a good idea.

Ray
User avatar
Yarc
Senior Member
Posts: 1051
Joined: Sat Apr 11, 2009 8:13 am
Location: United Kingdom

Post by Yarc »

I've been forced to abandon the game. Had paused it from PC Hiarcs menu and put my laptop into standby. However, on resuming Hiarc's clock ended up at zero time when it should have been close to 2 hours, and it lost!? Well, I wasn't convinced it was a good idea, and this has helped confirm it!

Keep dedicated for dedicated use!

Regards
Ray
Post Reply