Question to the members of the ranking lists..

[lkaufman]

Sure I remember Eric. His ratings like SSDF, CCRL, CEGT etc. are based on engine vs engine games, and therefore underrate the old/weak engines in order not to overrrate the current strong ones. Part of the difference is FIDE vs. USCF, but that's actually only about 50 Elo or so on average. His ratings confirm that the Constellation 3.6 is less than a hundred below SuperCon, and therefore would be in the mid 1900s USCF assuming ratings are about in line with the mid '80s, which I think is roughly true (they inflated until 1995, then deflated some).

so you agree Connie 3.6 v 2700 is ok?
or should i use any of the other Connies?
i own each one of them

Steve

Yes, it's okay. I expect the 2700 will win the match, but I also expect that a 2700 player will win a rook odds match against a USCF 1700 player.

[Don]

so perhaps we can conclude that the difference between an average club player and a top GM in terms of material odds is a R
Steve

as an after thought.. Larry
do you think your rating compression formula of 75% x(diff in Elo) + a constant ..should figure in here in some way?

asking another way..
assuming the use of a 2700 Elo PC engine as a constant (i have to use this engine as i own no others)..what elo in your opinion should i use for the Dedicated computer to as closely as possible mimic the difference between an avg club player and a top GM?

Best Regards
Steve

First it depends on where you are getting the ratings. If they are from an engine-engine rating list then indeed you need to allow for compression. For example Novag constellation 3.6 you quote at 1650, but I'm sure that it's USCF rating would be higher than that, because SuperConstellation got an ufficial USCF rating of 2018 (so long ago!) and the two models are probably only a hundred Elo or so apart. So I suppose your 1650 is from an engine list. So probably the test you propose actually is of a stronger unit than the average club player. On the other hand humans are better than computers at using a material advantage, because they put much more emphasis on trading when ahead. So I think your test is a pretty good one for your purpose.

the rating comes from Eric Hallsworth "Selective Search" dedicated computer rating lists and are in ELO not USCF
he still publishes the list every two months
you might remember him when you rated the oldies for ICD back in the early 1990's?
Hallsworth quoted you quite often in those days

he currently(issue 148) shows the following ratings for the Novag Constellation series:
Connie 2 Mhz-1591
Connie 3.6 Mhz-1646
Super Connie-1728
Connie Expert-(AKA Chess Monster..full sized wooden auto-sensory)-1790

Best Regards
Steve

The super connie achieved 2018 in serious tournament play against humans. At that particular point in time the USCF rating pool was somewhat inflated relative to FIDE ratings, but it wasn't ridiculous - maybe 100 ELO at most and probably much less than that.

The number 1728 seems quite a bit too low. However it's possible that the connie got way over rated as people were not yet very skilled at playing computers at the time and the sample of games the USCF required to get an "official" computer rating was still subject to fairly large error.

You mention the ratings are in ELO and not USCF, but the USCF uses the ELO rating system so what you said makes no sense. What you probably meant was FIDE vs USCF.

But it's difficult to see how he got the rating of 1728, it really seems way too low no matter how much you reasonably allow for sample size and rating inflation. So I would say his scale is off by at least 100 ELO and probably more like 200.

hi Don
way back when the Super connie got its OFFICIAL USCF rating ..
USCF ratings were 125-150 higher Then elo ratings
in addition some of the ratings achieved by the USCF Chess Computer ratings agency (now defunct)had several issues as to accuracy
Larry wrote extensively about this back then and he can speak to this far better then i could

in addition to Selective Search ..there is a German dedicated chess computer site which also rates many of the dedicated computers independently from Selective Search
they show very similar rating's for the Conny 3.6 at 1648 ELO
and 1732 for the Super Conny

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

http://www.schach-computer.info/wiki/in ... lation_3.6


what ever the two of you feel is best to achieve the goals of my experiment is OK with me
Regards
Steve

Ok, so maybe the ELO rating really is off by a few hundred ELO.

I actually participated in the testing of some of the machines, but I don't remember if I was involved with the constellation or not . I operated the Rebel program (one of many operators) when it was a standalone machine at one of the human tournament for the USCF rating agency.

My recollection was that they required a fairly small number of games - for some reason 48 stands out in my mind but it's probably wrong. But if so, 48 games can produce huge error margins. At the time a lot of human players were not well equipped to play computers - they were rather intimidated by them and tended to play pretty badly in comparison to how they would normally play.

By the way, you keep comparing USCF to ELO as if they were describing the same type of thing. One is an organization and the other is a type of rating system that many organization use including the USCF. It's total nonsense to say, "one is in USCF, the other ELO."

[Steve B]

so perhaps we can conclude that the difference between an average club player and a top GM in terms of material odds is a R
Steve

as an after thought.. Larry
do you think your rating compression formula of 75% x(diff in Elo) + a constant ..should figure in here in some way?

asking another way..
assuming the use of a 2700 Elo PC engine as a constant (i have to use this engine as i own no others)..what elo in your opinion should i use for the Dedicated computer to as closely as possible mimic the difference between an avg club player and a top GM?

Best Regards
Steve

First it depends on where you are getting the ratings. If they are from an engine-engine rating list then indeed you need to allow for compression. For example Novag constellation 3.6 you quote at 1650, but I'm sure that it's USCF rating would be higher than that, because SuperConstellation got an ufficial USCF rating of 2018 (so long ago!) and the two models are probably only a hundred Elo or so apart. So I suppose your 1650 is from an engine list. So probably the test you propose actually is of a stronger unit than the average club player. On the other hand humans are better than computers at using a material advantage, because they put much more emphasis on trading when ahead. So I think your test is a pretty good one for your purpose.

the rating comes from Eric Hallsworth "Selective Search" dedicated computer rating lists and are in ELO not USCF
he still publishes the list every two months
you might remember him when you rated the oldies for ICD back in the early 1990's?
Hallsworth quoted you quite often in those days

he currently(issue 148) shows the following ratings for the Novag Constellation series:
Connie 2 Mhz-1591
Connie 3.6 Mhz-1646
Super Connie-1728
Connie Expert-(AKA Chess Monster..full sized wooden auto-sensory)-1790

Best Regards
Steve

The super connie achieved 2018 in serious tournament play against humans. At that particular point in time the USCF rating pool was somewhat inflated relative to FIDE ratings, but it wasn't ridiculous - maybe 100 ELO at most and probably much less than that.

The number 1728 seems quite a bit too low. However it's possible that the connie got way over rated as people were not yet very skilled at playing computers at the time and the sample of games the USCF required to get an "official" computer rating was still subject to fairly large error.

You mention the ratings are in ELO and not USCF, but the USCF uses the ELO rating system so what you said makes no sense. What you probably meant was FIDE vs USCF.

But it's difficult to see how he got the rating of 1728, it really seems way too low no matter how much you reasonably allow for sample size and rating inflation. So I would say his scale is off by at least 100 ELO and probably more like 200.

hi Don
way back when the Super connie got its OFFICIAL USCF rating ..
USCF ratings were 125-150 higher Then elo ratings
in addition some of the ratings achieved by the USCF Chess Computer ratings agency (now defunct)had several issues as to accuracy
Larry wrote extensively about this back then and he can speak to this far better then i could

in addition to Selective Search ..there is a German dedicated chess computer site which also rates many of the dedicated computers independently from Selective Search
they show very similar rating's for the Conny 3.6 at 1648 ELO
and 1732 for the Super Conny

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

http://www.schach-computer.info/wiki/in ... lation_3.6


what ever the two of you feel is best to achieve the goals of my experiment is OK with me
Regards
Steve

Ok, so maybe the ELO rating really is off by a few hundred ELO.

I actually participated in the testing of some of the machines, but I don't remember if I was involved with the constellation or not . I operated the Rebel program (one of many operators) when it was a standalone machine at one of the human tournament for the USCF rating agency.

My recollection was that they required a fairly small number of games - for some reason 48 stands out in my mind but it's probably wrong. But if so, 48 games can produce huge error margins. At the time a lot of human players were not well equipped to play computers - they were rather intimidated by them and tended to play pretty badly in comparison to how they would normally play.

By the way, you keep comparing USCF to ELO as if they were describing the same type of thing. One is an organization and the other is a type of rating system that many organization use including the USCF. It's total nonsense to say, "one is in USCF, the other ELO."

well when i say ELO v USCF i mean Swedish Elo vs USCF ELO ratings
but now that you mention it ..
it is a bit too loose to describe the two different rating polls that way but i do see it commonly desrcibed that way
it was simply a fact back then that if a computer was rated xxxx by USCF ratings then you needed to subtract 125-150 to get to the equivlant Swedish Elo rating
anyway..it seems that the Conny 3.6Mhz v PC Engine makes for a good matchup so i will go with that

Best Regards
Steve

[Don]

so perhaps we can conclude that the difference between an average club player and a top GM in terms of material odds is a R
Steve

as an after thought.. Larry
do you think your rating compression formula of 75% x(diff in Elo) + a constant ..should figure in here in some way?

asking another way..
assuming the use of a 2700 Elo PC engine as a constant (i have to use this engine as i own no others)..what elo in your opinion should i use for the Dedicated computer to as closely as possible mimic the difference between an avg club player and a top GM?

Best Regards
Steve

First it depends on where you are getting the ratings. If they are from an engine-engine rating list then indeed you need to allow for compression. For example Novag constellation 3.6 you quote at 1650, but I'm sure that it's USCF rating would be higher than that, because SuperConstellation got an ufficial USCF rating of 2018 (so long ago!) and the two models are probably only a hundred Elo or so apart. So I suppose your 1650 is from an engine list. So probably the test you propose actually is of a stronger unit than the average club player. On the other hand humans are better than computers at using a material advantage, because they put much more emphasis on trading when ahead. So I think your test is a pretty good one for your purpose.

the rating comes from Eric Hallsworth "Selective Search" dedicated computer rating lists and are in ELO not USCF
he still publishes the list every two months
you might remember him when you rated the oldies for ICD back in the early 1990's?
Hallsworth quoted you quite often in those days

he currently(issue 148) shows the following ratings for the Novag Constellation series:
Connie 2 Mhz-1591
Connie 3.6 Mhz-1646
Super Connie-1728
Connie Expert-(AKA Chess Monster..full sized wooden auto-sensory)-1790

Best Regards
Steve

The super connie achieved 2018 in serious tournament play against humans. At that particular point in time the USCF rating pool was somewhat inflated relative to FIDE ratings, but it wasn't ridiculous - maybe 100 ELO at most and probably much less than that.

The number 1728 seems quite a bit too low. However it's possible that the connie got way over rated as people were not yet very skilled at playing computers at the time and the sample of games the USCF required to get an "official" computer rating was still subject to fairly large error.

You mention the ratings are in ELO and not USCF, but the USCF uses the ELO rating system so what you said makes no sense. What you probably meant was FIDE vs USCF.

But it's difficult to see how he got the rating of 1728, it really seems way too low no matter how much you reasonably allow for sample size and rating inflation. So I would say his scale is off by at least 100 ELO and probably more like 200.

hi Don
way back when the Super connie got its OFFICIAL USCF rating ..
USCF ratings were 125-150 higher Then elo ratings
in addition some of the ratings achieved by the USCF Chess Computer ratings agency (now defunct)had several issues as to accuracy
Larry wrote extensively about this back then and he can speak to this far better then i could

in addition to Selective Search ..there is a German dedicated chess computer site which also rates many of the dedicated computers independently from Selective Search
they show very similar rating's for the Conny 3.6 at 1648 ELO
and 1732 for the Super Conny

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

http://www.schach-computer.info/wiki/in ... lation_3.6


what ever the two of you feel is best to achieve the goals of my experiment is OK with me
Regards
Steve

Ok, so maybe the ELO rating really is off by a few hundred ELO.

I actually participated in the testing of some of the machines, but I don't remember if I was involved with the constellation or not . I operated the Rebel program (one of many operators) when it was a standalone machine at one of the human tournament for the USCF rating agency.

My recollection was that they required a fairly small number of games - for some reason 48 stands out in my mind but it's probably wrong. But if so, 48 games can produce huge error margins. At the time a lot of human players were not well equipped to play computers - they were rather intimidated by them and tended to play pretty badly in comparison to how they would normally play.

By the way, you keep comparing USCF to ELO as if they were describing the same type of thing. One is an organization and the other is a type of rating system that many organization use including the USCF. It's total nonsense to say, "one is in USCF, the other ELO."

well when i say ELO v USCF i mean Swedish Elo vs USCF ELO ratings
but now that you mention it ..
it is a bit too loose to describe the two different rating polls that way but i do see it commonly desrcibed that way
it was simply a fact back then that if a computer was rated xxxx by USCF ratings then you needed to subtract 125-150 to get to the equivlant Swedish Elo rating
anyway..it seems that the Conny 3.6Mhz v PC Engine makes for a good matchup so i will go with that

Best Regards
Steve

That's much better. Of course this only works if you see Swedish as some kind of gold standard.

One thing is clear, the USCF rating system was pretty broken and was not any better as a standard that the dollar system.

Probably FIDE ratings make the most sense as a standard as it's international. How to swedish ratings compare to FIDE ratings?

[Steve B]

One thing is clear, the USCF rating system was pretty broken and was not any better as a standard that the dollar system.

Probably FIDE ratings make the most sense as a standard as it's international. How to swedish ratings compare to FIDE ratings?

that i am not certain
actually the Swedish ELO collectors often refer to is the SSDF ELO list which at one time rated many dedicated chess computers

so if you compared a dedicated computer as it appeared on the SSDF list v its rating by the USCF rating agency or by Larry in his work with ICD then you would get a difference mentioned above

to this day Manufactures quote their estimated ratings using a higher number ..stating the rating estimate as xxxx USCF
this tells collectors that we need to automatically deduct 100+ pts to get to the SSDF ELO equivalent
not to mention probably another 200 or more points for gross exaggeration

as an example..Novag Released the Citrine 3 years ago with an ESTIMATED USCF rating of 2330
it states this right on the box
this photo is from my collection
you can see the red sticker on the bottom right of the box
not to clear but i guess you can see where it says"
ESTIMATED ELO 2330 USCF



you can also see this claim here on Novag's web site:
http://www.novag.com/Web%202006/USA-CAN ... -top-E.htm



Hallsworth shows the Citrine to be rated 2017 SSDF ELO on his Selective Search lists

part of the difference is the USCF v SSDF conversion factor and the other part is rating exaggeration by the manufacturer

if there were absolutely NO exaggeration by the manufacturer they would advertise the computer with a rating of about 2140 USCF(2017+125) in order to market the computer with the highest possible rating they can
Best Regards
Steve

[Don]

One thing is clear, the USCF rating system was pretty broken and was not any better as a standard that the dollar system.

Probably FIDE ratings make the most sense as a standard as it's international. How to swedish ratings compare to FIDE ratings?

that i am not certain
actually the Swedish ELO collectors often refer to is the SSDF ELO list which at one time rated many dedicated chess computers

so if you compared a dedicated computer as it appeared on the SSDF list v its rating by the USCF rating agency or by Larry in his work with ICD then you would get a difference mentioned above

to this day Manufactures quote their estimated ratings using a higher number ..stating the rating estimate as xxxx USCF
this tells collectors that we need to automatically deduct 100+ pts to get to the SSDF ELO equivalent
not to mention probably another 200 or more points for gross exaggeration

as an example..Novag Released the Citrine 3 years ago with an ESTIMATED USCF rating of 2330
it states this right on the box
this photo is from my collection
you can see the red sticker on the bottom right of the box
not to clear but i guess you can see where it says"
ESTIMATED ELO 2330 USCF



you can also see this claim here on Novag's web site:
http://www.novag.com/Web%202006/USA-CAN ... -top-E.htm



Hallsworth shows the Citrine to be rated 2017 SSDF ELO on his Selective Search lists

part of the difference is the USCF v SSDF conversion factor and the other part is rating exaggeration by the manufacturer

if there were absolutely NO exaggeration by the manufacturer they would advertise the computer with a rating of about 2140 USCF(2017+125) in order to market the computer with the highest possible rating they can
Best Regards
Steve

There is hardly any product advertised that has an honest label, so this is no surprise. Have you tried to buy healthy food in a supermarket in the USA? Most of the health claims are outright lies.

Once in a great while someone calls them on it and you will hear about it the news - but that is rare.

[Steve B]

If you ever have any desire to sponsor any other computer-human match of any type, just message me.

how about this for an idea??
we have played several matches here pitting dedicated computers vs the forum
these games can be found in the "Tournament Forum"

1 game match
my Conny 3.6 v a real GM
QR odds..GM has white
TC 3 minutes per move
moves to be posted here
GM..YOU


we would have to be on the honor system of course
i certainly trust you to not use any engine help and to time yourself
i could send you the computer after the game ends and you could check each move that was played
with pondering OFF i doubt very highly that the computer would make different moves (even though there might be some built in random factor the computer uses)
i would pay round trip postage for shipping the computer back and forth

appearance fee for you..win lose or draw .. to be discussed by email
game to begin once you receive the appearance fee

moves to be posted here (i will set up a thread for it)at least once per day but i would not claim a win by time forfeit so not strictly enforced
more of a guideline really

if you are game ..please let me know
Best Regards
Steve

[Don]

Why not have a controlled test using one of the chess servers? You could choose the human players in advance and agree upon the rules, etc and you could operate the machines manually.

I've always felt that a concerted effort to rate programs against humans (both older programs and new ones) should be possible with decent samples of games.

The beauty of this is that you could play modern programs, with ponder off and very fast levels - such as 1 second per move for instance. This would enable you to not have to find 2800 humans to play matches.

There are also utilites which will slow down your computer in a controlled way, so you could estimate the strength of modern PC based programs on older hardware. It can be calibrated by aksing someone who actually has a target machine to run a test for you. What is the oldest machine that Rybka will run on?

If you ever have any desire to sponsor any other computer-human match of any type, just message me.

how about this for an idea??
we have played several matches here pitting dedicated computers vs the forum
these games can be found in the "Tournament Forum"

1 game match
my Conny 3.6 v a real GM
QR odds..GM has white
TC 3 minutes per move
moves to be posted here
GM..YOU


we would have to be on the honor system of course
i certainly trust you to not use any engine help and to time yourself
i could send you the computer after the game ends and you could check each move that was played
with pondering OFF i doubt very highly that the computer would make different moves (even though there might be some built in random factor the computer uses)
i would pay round trip postage for shipping the computer back and forth

appearance fee for you..win lose or draw .. to be discussed by email
game to begin once you receive the appearance fee

moves to be posted here (i will set up a thread for it)at least once per day but i would not claim a win by time forfeit so not strictly enforced
more of a guideline really

if you are game ..please let me know
Best Regards
Steve

[Steve B]

Why not have a controlled test using one of the chess servers? You could choose the human players in advance and agree upon the rules, etc and you could operate the machines manually.

I've always felt that a concerted effort to rate programs against humans (both older programs and new ones) should be possible with decent samples of games.

The beauty of this is that you could play modern programs, with ponder off and very fast levels - such as 1 second per move for instance. This would enable you to not have to find 2800 humans to play matches.

There are also utilites which will slow down your computer in a controlled way, so you could estimate the strength of modern PC based programs on older hardware. It can be calibrated by aksing someone who actually has a target machine to run a test for you. What is the oldest machine that Rybka will run on?

If you ever have any desire to sponsor any other computer-human match of any type, just message me.

how about this for an idea??
we have played several matches here pitting dedicated computers vs the forum
these games can be found in the "Tournament Forum"

1 game match
my Conny 3.6 v a real GM
QR odds..GM has white
TC 3 minutes per move
moves to be posted here
GM..YOU


we would have to be on the honor system of course
i certainly trust you to not use any engine help and to time yourself
i could send you the computer after the game ends and you could check each move that was played
with pondering OFF i doubt very highly that the computer would make different moves (even though there might be some built in random factor the computer uses)
i would pay round trip postage for shipping the computer back and forth

appearance fee for you..win lose or draw .. to be discussed by email
game to begin once you receive the appearance fee

moves to be posted here (i will set up a thread for it)at least once per day but i would not claim a win by time forfeit so not strictly enforced
more of a guideline really

if you are game ..please let me know
Best Regards
Steve

well
i guess i should describe my interest in our shared hobby a bit
i am a collector of dedicated computers and that is my chief area of interest
i have been collecting the oldies since 1977(year first commercially available dedicated computer was released for sale)
i only own one PC Engine and about 500 Dedicated computers
so i am trying to do this in a way that will at least be of interest to me and to provide some chess entertainment for the forum members here who could them follow the game day by day
the other forum games did seem to attract a fairly large amount of interest

i do regret having made the offer in such a public way now that i come to think of it
probably should have contacted Larry privately first

another thought would be to play the game( or games ) vs Komodo
what is the rating at 40/2?

Sigh Regards
Steve

[Don]

Why not have a controlled test using one of the chess servers? You could choose the human players in advance and agree upon the rules, etc and you could operate the machines manually.

I've always felt that a concerted effort to rate programs against humans (both older programs and new ones) should be possible with decent samples of games.

The beauty of this is that you could play modern programs, with ponder off and very fast levels - such as 1 second per move for instance. This would enable you to not have to find 2800 humans to play matches.

There are also utilites which will slow down your computer in a controlled way, so you could estimate the strength of modern PC based programs on older hardware. It can be calibrated by aksing someone who actually has a target machine to run a test for you. What is the oldest machine that Rybka will run on?

If you ever have any desire to sponsor any other computer-human match of any type, just message me.

how about this for an idea??
we have played several matches here pitting dedicated computers vs the forum
these games can be found in the "Tournament Forum"

1 game match
my Conny 3.6 v a real GM
QR odds..GM has white
TC 3 minutes per move
moves to be posted here
GM..YOU


we would have to be on the honor system of course
i certainly trust you to not use any engine help and to time yourself
i could send you the computer after the game ends and you could check each move that was played
with pondering OFF i doubt very highly that the computer would make different moves (even though there might be some built in random factor the computer uses)
i would pay round trip postage for shipping the computer back and forth

appearance fee for you..win lose or draw .. to be discussed by email
game to begin once you receive the appearance fee

moves to be posted here (i will set up a thread for it)at least once per day but i would not claim a win by time forfeit so not strictly enforced
more of a guideline really

if you are game ..please let me know
Best Regards
Steve

well
i guess i should describe my interest in our shared hobby a bit
i am a collector of dedicated computers and that is my chief area of interest
i have been collecting the oldies since 1977(year first commercially available dedicated computer was released for sale)
i only own one PC Engine and about 500 Dedicated computers
so i am trying to do this in a way that will at least be of interest to me and to provide some chess entertainment for the forum members here who could them follow the game day by day
the other forum games did seem to attract a fairly large amount of interest

i do regret having made the offer in such a public way now that i come to think of it
probably should have contacted Larry privately first

another thought would be to play the game( or games ) vs Komodo
what is the rating at 40/2?

Sigh Regards
Steve

That's no problem, I am glad you made it public.

Whether you also try to rate modern programs is not important, what is important is that some kind of methodology is applied to getting ratings for computers you can believe in - based on human only play.

So if you can somehow get these older dedicated machines rated that would also be quite nice.

It's more fun to do it face to face in person of course, but it may not be practical if you want a reasonable variety of players and large enough sample of games to narrow it down to within some reasonable margin.