| |
Main
Date: 31 Oct 2006 18:59:12
From: Dave (from the UK)
Subject: PROOF a beginner has no rating.
|
There have been several arguments on here about determining the rating
of an absolute beginner at chess, who knows only the moves.
The thread 'Improve your chess in 7 days or less' got its name changed
to 'Rip-off of Reinfeld', then into 'beginner rating' etc.
Rather than change the name again, I thought I'd start a *new* thread.
Values of 0, 500 and 1000 have all been mentioned.
See this thread for example.
http://groups.google.co.uk/group/rec.games.chess.misc/browse_frm/thread/e386d52f22f002df/bd779ed9a29f868b?lnk=st&q=rip-off+reinfield&rnum=1&hl=en#bd779ed9a29f868b
Just to be a devils advocate, here is a PROOF, based on arguments like
Heisenberg's uncertainly principal, that the rating of a complete
beginner, who only knows the moves, can NEVER be determined.
1) In order to determine with any useful statistical significance the
ELO rating of a player, you need to establish that with multiple games.
For example, on ICC one gets a provisional rating until one has played
20 games.
2) After playing the number of games necessary to establish the rating
with useful statistical significance, the player is no longer a complete
beginner.
Hence such a rating can NEVER be determined.
Comments?
--
Dave (from the UK)
Please note my email address changes periodically to avoid spam.
It is always of the form: [email protected]
Hitting reply will work for a few months only - later set it manually.
http://witm.sourceforge.net/ (Web based Mathematica front end)
|
|
| |
Date:
From: Martin Brown
Subject: Re: PROOF a beginner has no rating.
|
|
| | |
Date: 02 Nov 2006 10:12:34
From: Kenneth Sloan
Subject: Re: PROOF a beginner has no rating.
|
tin Brown wrote:
> Kenneth Sloan wrote:
>> tin Brown wrote:
>>> One way of determining the original rating of a complete beginner is to
>>> accept that they will improve with each game played and then model it.
>>> Afterwards you can compute the intercept of their rating vs number of
>>> games played at the 0 games played axis to get a fair estimate.
>>>
>>> My instinct is that you would work hard to find someone with an ELO
>>> rating below 500 (who wasn't deliberately playing to lose).
>> If you are restricting the population to "adults of normal
>> intelligence", then I would agree. However, if you include
>> all players, regardless of age or intelligence, I can find at least
>> 10 players with deserved ratings << 0000.
>>
>> and hundreds upon hundreds of players with deserved ratings << 0500
>
> I may have been a little optimistic about the weakest player who knows
> the basic rules (or perhaps more accurately doesn't break any of them
> in a game). Or in beginners chess that no-one has noticed any of them
> being broken...
>
I'm sorry - I wasn't clear. Most of these VERY low rated players do NOT
know the rules! Believe it or not, there are thousands of players
entering rated tournaments who do NOT know all the rules. And, many of
them are children with undeveloped attention spans. And, of course,
SOME of these will grow up to be "not so st" adults. And, some of
them are only there because Mommy dropped them off and said she'll be
back at 5:00.
A motivated adult novice of normal intelligence, who has studied the
game enough to know all the rules of play is light-years ahead of
such seasoned tournament veterans!
"Knowing the rules" is by no means the bottom rung on the ladder.
And...oh yes...I think we all know chess players who "know a lot about
chess - especially the history and culture". They can perhaps even
quote famous games and pontificate on the relative merits of past
Champions. But, they can't actually *play* worth a damn. My point is
that "knowing much more than the rules" doesn't *necessarily* give you
an edge over the board against someone who *only* knows the rules.
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | |
Date: 02 Nov 2006 11:42:28
From: Ange1o DePa1ma
Subject: Re: PROOF a beginner has no rating.
|
"Kenneth Sloan" <[email protected] > wrote
> "Knowing the rules" is by no means the bottom rung on the ladder.
We needn't have a philosophic debate about what it means to "know the
rules." I drew a game I should have won several years ago (against a young
man who is probably of IM strength today) because I did had forgotten the
rule for time forfeits that requires the winner keep a complete scoresheet.
Your original statement, I believe, mentioned a player who knows how the
pieces moved, not someone who knows as much about chess legal trivia as an
international arbiter.
So now that I know you are a director, and have seen 100s and 1000s of
sub-1000 players, do you have any data on, say, 900-rated adults vs.
900-rated kids? I'd bet there's no difference (except maybe the kids are
improving quickly and the adults are stuck).
>
> And...oh yes...I think we all know chess players who "know a lot about
> chess - especially the history and culture". They can perhaps even
> quote famous games and pontificate on the relative merits of past
> Champions. But, they can't actually *play* worth a damn. My point is
> that "knowing much more than the rules" doesn't *necessarily* give you an
> edge over the board against someone who *only* knows the rules.
Sadly, this is what I am becoming except I'm forgetting the history,
culture, and classic games as well.
|
| | | | |
Date: 02 Nov 2006 11:40:43
From: Kenneth Sloan
Subject: Re: PROOF a beginner has no rating.
|
Ange1o DePa1ma wrote:
> "Kenneth Sloan" <[email protected]> wrote
>
>> "Knowing the rules" is by no means the bottom rung on the ladder.
>
> We needn't have a philosophic debate about what it means to "know the
> rules." I drew a game I should have won several years ago (against a young
> man who is probably of IM strength today) because I did had forgotten the
> rule for time forfeits that requires the winner keep a complete scoresheet.
> Your original statement, I believe, mentioned a player who knows how the
> pieces moved, not someone who knows as much about chess legal trivia as an
> international arbiter.
I'm not talking about trivia. I'm talking about en passant, castling,
and of course the classic "how the horsey moves", etc. When I say
"knows the rules", I mean someone who does not need
to think hard, or resort to a cheat sheet, to immediately see all of the
legal moves in a position.
He need not know the proper procedures for claiming a draw due to
repeated positions when the arbiter asks him to seal his next move.
If you haven't watched thousands of sub-1000 rated kids playing
tournament chess, you may not be able to appreciate hoe HIGHLY you
can be rated (that is, how strong a performance you can turn in) without
yet reaching the stage of "knowing the rules".
On the other hand, it's possible for a player to be absoutely BRILLIANT
most of the time - but have gaping holes in his basic competence.
>
> So now that I know you are a director, and have seen 100s and 1000s of
> sub-1000 players, do you have any data on, say, 900-rated adults vs.
> 900-rated kids? I'd bet there's no difference (except maybe the kids are
> improving quickly and the adults are stuck).
Data? No.
Observations. Yes.
0900 rated kids are usually excellent tactical players who like to go
for quick kills. They are always dangerous, because sometimes their
suicide attacks can occasionally succeed.
0900 rated adults simply play 3rd best moves, all game long, with little
imagination and no flair. They can beat players who drop pieces faster
than they do, but no one else.
Given the choice in a "win this game, or die" situation, I'll take the
0900 rated adult as my opponent. All I need to do against him is
develop my pieces into the center and wait for him to give me pieces.
Against the kid, I actually have to stay awake and not fall for a cheapo.
Note that these are comments about players with reasonably STABLE
ratings of 0900. There's also the issue of how fast the player is
improving. Kids improve in fits and starts (very Piaget-like), and
sometimes play in pools that deflate their ratings. So, the 0900 rated
kid is even *more* dangerous. He might be underrated because we
measured him badly, or he might be underrated because he just learned
how to checkmate with only one rook. Yes, indeed, in my experience there
are lots of 1000 rated players who cannot mate with KRk. Over and over
again I've seen sub-1000 players who have memorized the Morphy Opera
House Game, love the Fried Liver Attack, and know the Albin
Counter-Gambit 15 moves deep - but cannot mate with KRk. To draw
against such a player, you simply have to ward off the cheapos. You can
sac a rook to get out of immediate trouble and then wait for them to
flounder in the endgame. You can't *win* that way, but you can draw.
Even the worst player in the Kindergarten section occasionally scores a
draw by sacrificing all of her pieces and then walking into stalemate.
Of course, there is also the small problem of *recognizing* stalemate
(knowing the rules, again). I've seen tens of games where both players
were in check for more than 10 consecutive moves. Unless you've been
there, you really don't know the rich texture of sub-1000 chess.
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | | | |
Date: 02 Nov 2006 11:48:31
From: David Kane
Subject: Re: PROOF a beginner has no rating.
|
"Kenneth Sloan" <[email protected] > wrote in message
news:[email protected]...
>
> Data? No.
>
> Observations. Yes.
>
> 0900 rated kids are usually excellent tactical players who like to go
> for quick kills. They are always dangerous, because sometimes their
> suicide attacks can occasionally succeed.
>
> 0900 rated adults simply play 3rd best moves, all game long, with little
> imagination and no flair. They can beat players who drop pieces faster than
> they do, but no one else.
>
> Given the choice in a "win this game, or die" situation, I'll take the 0900
> rated adult as my opponent. All I need to do against him is develop my pieces
> into the center and wait for him to give me pieces. Against the kid, I
> actually have to stay awake and not fall for a cheapo.
>
> Note that these are comments about players with reasonably STABLE ratings of
> 0900. There's also the issue of how fast the player is improving. Kids
> improve in fits and starts (very Piaget-like), and
> sometimes play in pools that deflate their ratings. So, the 0900 rated kid is
> even *more* dangerous. He might be underrated because we measured him badly,
> or he might be underrated because he just learned
> how to checkmate with only one rook. Yes, indeed, in my experience there are
> lots of 1000 rated players who cannot mate with KRk. Over and over again I've
> seen sub-1000 players who have memorized the Morphy Opera House Game, love the
> Fried Liver Attack, and know the Albin Counter-Gambit 15 moves deep - but
> cannot mate with KRk. To draw against such a player, you simply have to ward
> off the cheapos. You can sac a rook to get out of immediate trouble and then
> wait for them to flounder in the endgame. You can't *win* that way, but you
> can draw.
> Even the worst player in the Kindergarten section occasionally scores a draw
> by sacrificing all of her pieces and then walking into stalemate. Of course,
> there is also the small problem of *recognizing* stalemate (knowing the rules,
> again). I've seen tens of games where both players were in check for more
> than 10 consecutive moves. Unless you've been there, you really don't know
> the rich texture of sub-1000 chess.
>
Of course, watching games like that *is* painful.
But the mistake is in comparing their play to the
play of the adult 1500's at your club . In fact,
quite a lot of those sub-1000s are better players
than their parents, adult friends etc. I frequently
watch kids play with their parents. Guess what -
they miss checks, they leave pieces en prise,
they miss mates in one etc.
The one place where adults do seem to do better
is in "solving problems",i.e. things that could be
considered more "strategy". How do I win from
here? How do I mate with K+Q? But the kids
overcome that deficit by seeing the tactics on the
board better.
|
| |
Date:
From: Martin Brown
Subject: Re: PROOF a beginner has no rating.
|
|
| | |
Date: 02 Nov 2006 10:02:08
From: Kenneth Sloan
Subject: Re: PROOF a beginner has no rating.
|
tin Brown wrote:
>
>> Over the last decade I've been acquainted with at least 30 adults and
>> children rated below 1000. Maybe 50.
>
> I don't doubt it. But I think they are rare in competitive chess
> tournaments (at least on this side of the pond). I'm surprised that the
> US ratings pool is calibrated that low.
come to the USCF National Elementary Championships. There, you will
find a KINDERGARTEN section (which I directed for many, many years)
chock full of players. I'm talking about approximately 100 players
(every year) - all with ratings below 1000. I would estimate that, on
average, there will be 2 players in the field who literally do not
know how ANY of the pieces move. This can be a challenge for the
floor directors.
It also offers a wealth of opportunity to watch players who are properly
rated at 0200, 0150, 0100, and (if only the system would assign such
ratings) -0500.
Over the last decade, I've directed events with more than 1000 players
rated under 1000.
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | |
Date: 02 Nov 2006 09:31:15
From: Ange1o DePa1ma
Subject: Re: PROOF a beginner has no rating.
|
"tin Brown" <
|
| |
Date: 01 Nov 2006 18:04:45
From: Inconnux
Subject: Re: PROOF a beginner has no rating.
|
> And finally my brother, an intelligent adult, at age 38, after his first
> tournament:
>
> http://www.uschess.org/msa/MbrDtlMain.php?12823377
>
> All are rated below 1000. Brother Dan and I have played since he was 5. He
> knows FAR more than simply how the pieces move.
I dont feel so bad now :)
|
| |
Date: 01 Nov 2006 10:37:20
From: wuolong
Subject: Re: PROOF a beginner has no rating.
|
On 31 Oct 2006, Kenneth Sloan wrote:
> One way to assign a rating is to observe the player's results.
>
> When you can't do that, there are still useful facts about the
> player that can improve the quality of your estimate.
As a statistician, I can't help jumping into this interesting
discussion. A player may have a true intrinsic chess strength, "true
rating". When he/she actual plays the 'observed' performance might be
lower or higher than the true strength. Thus the idea of modeling the
rating as a Normally distributed random variable with mean around the
true strength and certain variance. (Elo made the crucial assumption
that the variance is the same for everybody which clearly is not true.
But we will ignore that for now.)
The question is to estimate the true strength, which we can't do without
data. (By definition, a player who has just learned the moves hasn't
played any games yet. As soon as she/he starts to play, she/he does not
qualify as an 'absolute beginner' anymore. As other posters noted, an
adult of average intelligence can learn a lot in a couple of games.
So we can't estimate the true strength. However we can think about what
is our 'prior mean' for this absolute beginner?
Elo system translates to expected scores when players match up and that
gives us some ideas. The expected score between a 1000 and a 1600
player is 0.03, that is the 1000 player has probability of 0.03 of
winning. So the question is, would you be willing to bet, with odds
3:100 that the absolute beginner will win?
Personally, I think the odds should be lower than that. The expected
score for a 600-800 player against 1600 is about 0.003-0.01. I think
that is much more plausible range for a 'beginner who just know the
rules'.
Cheers,
- wuolong
|
| |
Date:
From: Martin Brown
Subject: Re: PROOF a beginner has no rating.
|
|
| | |
Date: 01 Nov 2006 15:40:31
From: Kenneth Sloan
Subject: Re: PROOF a beginner has no rating.
|
tin Brown wrote:
>
> One way of determining the original rating of a complete beginner is to
> accept that they will improve with each game played and then model it.
> Afterwards you can compute the intercept of their rating vs number of
> games played at the 0 games played axis to get a fair estimate.
>
> My instinct is that you would work hard to find someone with an ELO
> rating below 500 (who wasn't deliberately playing to lose).
>
> Regards,
> tin Brown
>
If you are restricting the population to "adults of normal
intelligence", then I would agree. However, if you include
all players, regardless of age or intelligence, I can find at least
10 players with deserved ratings << 0000.
and hundreds upon hundreds of players with deserved ratings << 0500
The question at the moment is: what does your instinct tell you about
the AVERAGE beginner who knows the rules.
[beware...that might be a trick question...even World Championship
Candidates have demonstrated that they don't know all the rules.
Do *you* know *all* the rules about 0-0-0? Korchnoi didn't!]
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | |
Date: 01 Nov 2006 11:52:26
From: Ange1o DePa1ma
Subject: Re: PROOF a beginner has no rating.
|
"tin Brown" <
|
| | | |
Date: 01 Nov 2006 15:42:55
From: Kenneth Sloan
Subject: Re: PROOF a beginner has no rating.
|
Ange1o DePa1ma wrote:
>
> It is so obvious that a rating of 1000 is not a rank beginner. I'm
> incredulous that some of you insist it is. Is this some sort of "big lie"
> post or what?
>
>
>
That's right. Everyone who disagrees with you is a liar.
You win.
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | | |
Date: 01 Nov 2006 18:08:10
From: Ange1o DePa1ma
Subject: Re: PROOF a beginner has no rating.
|
"Kenneth Sloan" <[email protected] > wrote
> Ange1o DePa1ma wrote:
>> It is so obvious that a rating of 1000 is not a rank beginner. I'm
>> incredulous that some of you insist it is. Is this some sort of "big lie"
>> post or what?
>
> That's right. Everyone who disagrees with you is a liar.
>
> You win.
I did not mean that you were a liar -- how could your original statement be
a lie when you presented it as an opinion?
What I meant was that it's sometimes easier to convince people of something
outrageous than something that's only slightly off.
If you meant that the rating of one who just knew how pieces moved rules
*could* or *might* be 1000 given other adjustments to the rating scale (e.g.
Kasparov @ 15,000) then I misunderstood you. If you meant under the existing
system, with Kasparov at 2800, then you are clearly wrong.
Note, you did not take the "DePalma Challenge." All those individuals were
rated >1000, and all were "experienced" to some degree.
Angelo
|
| |
Date: 01 Nov 2006 06:21:12
From: [email protected]
Subject: Re: PROOF a beginner has no rating.
|
Dave (from the UK) wrote:
> There have been several arguments on here about determining the rating
> of an absolute beginner at chess, who knows only the moves.
>
> The thread 'Improve your chess in 7 days or less' got its name changed
> to 'Rip-off of Reinfeld', then into 'beginner rating' etc.
>
> Rather than change the name again, I thought I'd start a *new* thread.
>
> Values of 0, 500 and 1000 have all been mentioned.
>
> See this thread for example.
>
> http://groups.google.co.uk/group/rec.games.chess.misc/browse_frm/thread/e386d52f22f002df/bd779ed9a29f868b?lnk=st&q=rip-off+reinfield&rnum=1&hl=en#bd779ed9a29f868b
>
> Just to be a devils advocate, here is a PROOF, based on arguments like
> Heisenberg's uncertainly principal, that the rating of a complete
> beginner, who only knows the moves, can NEVER be determined.
>
> 1) In order to determine with any useful statistical significance the
> ELO rating of a player, you need to establish that with multiple games.
> For example, on ICC one gets a provisional rating until one has played
> 20 games.
>
> 2) After playing the number of games necessary to establish the rating
> with useful statistical significance, the player is no longer a complete
> beginner.
>
> Hence such a rating can NEVER be determined.
>
> Comments?
> --
> Dave (from the UK)
>
> Please note my email address changes periodically to avoid spam.
> It is always of the form: [email protected]
> Hitting reply will work for a few months only - later set it manually.
>
> http://witm.sourceforge.net/ (Web based Mathematica front end)
Opinion:
Ratings are efficient only within the pool of players you are rated
against. It makes perfect sense that a player who has played no rating
games goes as unrated. You cannot assign an arbitrary value (0, 500,
100, whatever) to a player who has not played any games against the
rating pool, as this would skew the ratings eventually.
I don't pretend to know the math involved, but this seems painfully
obvious to me. This is why you can play 1000 games on 5 different
servers and get 5 different ratings, probably not all that close to
each other. Also, a player can effectively and consistently play
opponents rated below himself and increase (INFLATE) his rating
slowly. You run into that scenario all the time on ICC and Playchess.
Interesting thread, however.
http://chess-training.blogspot.com
|
| | |
Date: 01 Nov 2006 15:35:34
From: Kenneth Sloan
Subject: Re: PROOF a beginner has no rating.
|
[email protected] wrote:
> Opinion:
>
> Ratings are efficient only within the pool of players you are rated
> against. It makes perfect sense that a player who has played no rating
> games goes as unrated. You cannot assign an arbitrary value (0, 500,
> 100, whatever) to a player who has not played any games against the
> rating pool, as this would skew the ratings eventually.
Not if you do it right!
And...in some (rare) cases, it turns out to be essential to make
*some* estimate of the rating to use for an UNR player. Once you
find this necessary, it's no longer a question of IF you can
assign a rating, but instead HOW MUCH CONFIDENCE you have
in that assigned rating.
I hope you'll agree that a rating of 1500, with a variance of 10^10, is
a reasonably accurate rating (for ALL players!). Not very precise - but
reasonably accurate.
>
> I don't pretend to know the math involved, but this seems painfully
> obvious to me. This is why you can play 1000 games on 5 different
> servers and get 5 different ratings, probably not all that close to
> each other. Also, a player can effectively and consistently play
> opponents rated below himself and increase (INFLATE) his rating
> slowly. You run into that scenario all the time on ICC and Playchess.
>
In a well designed rating system, you will NOT inflate your rating by
consistently playing low rated players.
In fact, we often hear the opposite complaint - that high rated players
see their ratings DEflate when forced to play low-rated competition.
when you hear both complaints, you are reasonably assured that the world
is nicely balanced.
> Interesting thread, however.
>
> http://chess-training.blogspot.com
>
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| |
Date: 31 Oct 2006 12:31:34
From:
Subject: Re: PROOF a beginner has no rating.
|
Dave (from the UK) wrote:
> There have been several arguments on here about determining the rating
> of an absolute beginner at chess, who knows only the moves.
>
> The thread 'Improve your chess in 7 days or less' got its name changed
> to 'Rip-off of Reinfeld', then into 'beginner rating' etc.
>
> Rather than change the name again, I thought I'd start a *new* thread.
>
> Values of 0, 500 and 1000 have all been mentioned.
>
> See this thread for example.
>
> http://groups.google.co.uk/group/rec.games.chess.misc/browse_frm/thread/e386d52f22f002df/bd779ed9a29f868b?lnk=st&q=rip-off+reinfield&rnum=1&hl=en#bd779ed9a29f868b
>
> Just to be a devils advocate, here is a PROOF, based on arguments like
> Heisenberg's uncertainly principal, that the rating of a complete
> beginner, who only knows the moves, can NEVER be determined.
>
> 1) In order to determine with any useful statistical significance the
> ELO rating of a player, you need to establish that with multiple games.
> For example, on ICC one gets a provisional rating until one has played
> 20 games.
>
> 2) After playing the number of games necessary to establish the rating
> with useful statistical significance, the player is no longer a complete
> beginner.
>
> Hence such a rating can NEVER be determined.
>
> Comments?
> --
> Dave (from the UK)
>
> Please note my email address changes periodically to avoid spam.
> It is always of the form: [email protected]
> Hitting reply will work for a few months only - later set it manually.
>
> http://witm.sourceforge.net/ (Web based Mathematica front end)
It's all well and nice to talk about Heisenberg's uncertainty
principle, but this is the real world. The ELO formula requires that
the player must have SOME number applied to it, although it doesn't
matter WHAT that number is. And since we're talking about "ratings",
we're pretty much talking about ELO (although other formulae exist,
they are much less accepted).
Just because there's a magic rating of 2000 for Masters and 2500 for
GMs, they could easily have been any other numbers. But a player's
rating can never be "undefined", unless the ELO formula is replaced by
something else that can handle such a starting point.
jm
|
| | |
Date: 31 Oct 2006 23:49:50
From: Dave (from the UK)
Subject: Re: PROOF a beginner has no rating.
|
[email protected] wrote:
>>Just to be a devils advocate, here is a PROOF, based on arguments like
>>Heisenberg's uncertainly principal, that the rating of a complete
>>beginner, who only knows the moves, can NEVER be determined.
>http://witm.sourceforge.net/ (Web based Mathematica front end)
>
>
> It's all well and nice to talk about Heisenberg's uncertainty
> principle, but this is the real world.
I would say in the "real world" in the context you mean, the
Heisenberg's Uncertainty Principle has little practical relevance. The
effects of it are too small to worry about.
If someone has played 1000 games of chess, the process of measuring
their performance by getting them to play 20 games against opponents of
a known rating, is probably not going to have much effect on their
performance.
I don't have a copy of Professor Elo's paper on the subject, but to
quote from Wikipedia:
http://en.wikipedia.org/wiki/ELO_rating_system
"Élő's central assumption was that the chess performance of each player
in each game is a normally distributed random variable. Although a
player might perform significantly better or worse from one game to the
next, Élő assumed that the mean value of the performances of any given
player changes only slowly over time."
It is *not* true to say the mean performance of a player changes slowly
over time if they have just leaned the moves. Having played 10 games, I
suspect they are significantly better having played only 1 game. As
such, the process of measurement will have a *very* significant effect
on the quantity you are trying to measure.
So unlike Heisenberg's Uncertainty Principle, the effect would be very
pronounced in the real world.
> The ELO formula requires that
> the player must have SOME number applied to it, although it doesn't
> matter WHAT that number is. And since we're talking about "ratings",
> we're pretty much talking about ELO (although other formulae exist,
> they are much less accepted).
I believe the assumptions Elo made are simply not valid in the case of
an absolute beginner. As such, attaching an ELO number to something
where the assumptions are very wrong is not sensible.
> Just because there's a magic rating of 2000 for Masters and 2500 for
> GMs, they could easily have been any other numbers.
Agreed.
> But a player's
> rating can never be "undefined", unless the ELO formula is replaced by
> something else that can handle such a starting point.
I can't see how you can measure it. As such, I can't see how it can
possibly be defined unless the formula is modified to say "By
definition, a beginner has an ELO of 100" or similar.
Since the method has no definition and it can't be measured, I doubt
there is much point attaching a value to it.
--
Dave (from the UK)
Please note my email address changes periodically to avoid spam.
It is always of the form: [email protected]
Hitting reply will work for a few months only - later set it manually.
http://witm.sourceforge.net/ (Web based Mathematica front end)
|
| | | |
Date: 31 Oct 2006 16:27:26
From: David Kane
Subject: Re: PROOF a beginner has no rating.
|
"Dave (from the UK)" <[email protected] > wrote in
message news:[email protected]...
>
> I believe the assumptions Elo made are simply not valid in the case of an
> absolute beginner. As such, attaching an ELO number to something where the
> assumptions are very wrong is not sensible.
>
Of course they aren't. But what would be your
response if I were to claim in Ken Sloan
fashion that an adult of average intelligence
who just knows the rules has a rating of 2700?
The fact that there is uncertainty to ratings doesn't
make the discussion meaningless.
By the way, I think in the USCF system the rating
formulae change at about 8 games. The USCF rating
algorithm assigns beginners a very high K-constant
(80, I think, which can be effectively doubled via
"bonus" points) So the number of games actually
going into a rating is much smaller than the number
which defines a "provisional" rating. Someone can
correct me, but I don't think provisional ratings are
treated differently in the rating algorithm.
|
| | | | |
Date: 01 Nov 2006 03:01:32
From: Dave (from the UK)
Subject: Re: PROOF a beginner has no rating.
|
David Kane wrote:
> "Dave (from the UK)" <[email protected]> wrote in
> message news:[email protected]...
>
>
>>I believe the assumptions Elo made are simply not valid in the case of an
>>absolute beginner. As such, attaching an ELO number to something where the
>>assumptions are very wrong is not sensible.
>>
>
>
> Of course they aren't.
So we agree!
> But what would be your
> response if I were to claim in Ken Sloan
> fashion that an adult of average intelligence
> who just knows the rules has a rating of 2700?
A large number of players who play the game seriously don't have a
rating of 2700, so it's fairly easy to disprove that one. Putting an
upper bound is fairly easy. Anyone who thinks it would be over 1200 is
really got to be mad.
> The fact that there is uncertainty to ratings doesn't
> make the discussion meaningless.
> By the way, I think in the USCF system the rating
> formulae change at about 8 games. The USCF rating
> algorithm assigns beginners a very high K-constant
> (80, I think, which can be effectively doubled via
> "bonus" points)
I don't know how the USCF method works, but I assume this K-constant you
talk of is some measure of the standard deviation (uncertainty) of the
rating. For a new member, this will be higher until they are established.
> So the number of games actually
> going into a rating is much smaller than the number
> which defines a "provisional" rating. Someone can
> correct me, but I don't think provisional ratings are
> treated differently in the rating algorithm.
Anyone who is a member of the USCF would have to be reasonably serious
about the game, so they would have played many times before. I don't
believe someone will learn the rules, then join USCF and play their
first ever game in a rated competition.
As such, it should be possible to determine a rating reasonably quickly,
as they performance is unlikely to change a huge amount from game to game.
In contrast, as absolute beginner who only just learned the rules,
playing his first few games, should learn a lot very quickly. They
should soon learn about the fact knights can fork. So its very likely
the large K-constant you talk of in USCF rating systems is not
sufficiently large. So I don't believe you can rely on the USCF system
in this case. In any case, the USCF system would have been designed for
serious chess players - not those who just learned the rules.
One *possible* way to establish a rating for absolute beginners might be
as follows.
Take 1000 beginners and let them play their *first* ever chess game
against a low rated player (say 1200 for example). Results of any of
their later games are ignored. Each person plays just one game.
Statistically, if there is a rating difference of x between two players,
the probability of wins and losses can be computed. The 1200 players
should beat the beginners in most games, but the percentage might allow
you to determine how much stronger the 1200 players and and so assign an
average rating for those 1000 people after just one game. That would be
VERY doggy, and not very practical to do, but at least it would give you
a result from only one game, so they don't have any chance to improve.
I can't help feeling the concept of a rating for someone who has just
learned the rules is absurd. You can put a lower limit of 0, an upper
limit of 1200, but I don't believe you can say anything else really.
--
Dave (from the UK)
Please note my email address changes periodically to avoid spam.
It is always of the form: [email protected]
Hitting reply will work for a few months only - later set it manually.
http://witm.sourceforge.net/ (Web based Mathematica front end)
|
| | | | | |
Date: 31 Oct 2006 22:31:46
From: Kenneth Sloan
Subject: Re: PROOF a beginner has no rating.
|
Such complicated thinking on such a simple topic.
A rating is nothing more or less than a prediction about future
performance. It's a single position on a scale.
One way to assign a rating is to observe the player's results.
When you can't do that, there are still useful facts about the
player that can improve the quality of your estimate. For example,
USCF uses the player's AGE to provide an initial rating. One
could conceivably use IQ scores, economic class, gender, eye color...
anything that might be correlated with rating. Of course, some
of these data might be more relevant than others.
Many posters in this thread have demonstrated this by using
the information that the player is playing in a USCF rated event - and
then saying "only players of a certain minimum strength ever play
in USCF rated events, so...". This is a perfect example that makes
my point.
Even if you have ZERO information about the player, it is possible
to come up with a "best estimate" of his rating. Of course, the
variance associated with this estimate will be very high, but the
variance is not infinite.
Someone here demonstrated this by saying: "surely it would be stupid
to claim that a new player is rated 2700".
OK, so...2700 is too high. How about 2000? Still too high...
How about -10000 (yes, that would be a perfectly valid rating). That
seems a bit low, no? (actually, it's too high for my dog - but might
not be too high for a chimpanzee).
So, we now have it bracketed to be somewhere in (-100000, 2000).
Narrowing that range will depend on the data you have at hand.
I still think my estimate is not too far off. Find yourself a random
adult of normal intelligence, teach him the moves and enter him in
5 USCF events of 5 rounds each. I claim that a good estimate of his
performance in those events is that the performance will be that
of a USCF 1000 rated player.
Of course, some individuals will do worse than that - but I'm moderately
confident that many individuals will perform much better than that.
And...some players will continue to perform worse than that, even after
they learn the complete history of chess, memorize 5 useless gimmick
openings, etc - learning much more than the moves, but not being able
to translate that into performance.
If your personal estimate is 0800, I won't argue too much. But, if your
personal estimate is 0100, I think it's clear that that's much, much too
low. By the same token, 1500 is clearly much, much too high.
And, of course, there is NO REASON to select 0000 - certainly not
because of any notion that 0000 means "no rating points".
So...call it (0800, 1200). That's "1000, with a HUGE variance".
A practical rating system can, and does, make SOME (but not much) use
of this very fuzzy estimate. And, as a result, the ratings generated
are just a little bit better than they would be if the system did NOT
make use of this estimate.
That's all there is to it!
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | | | | |
Date: 31 Oct 2006 23:06:33
From: David Kane
Subject: Re: PROOF a beginner has no rating.
|
"Kenneth Sloan" <[email protected] > wrote in message
news:[email protected]...
>
> A practical rating system can, and does, make SOME (but not much) use
> of this very fuzzy estimate. And, as a result, the ratings generated
> are just a little bit better than they would be if the system did NOT
> make use of this estimate.
True, the rating system imputes initial ratings
to unrated players. In that case it makes very good sense
to give unrateds a starting rating that jibes with the rating
that that class had historically achieved in their first events.
But that is something entirely different than assigning that
rating to random adults who know only the rules of chess.
99+% of people who know the rules have never played
a single USCF rated game. You cannot hope to reliably
extrapolate the performance of the 99% by looking at the
1%.
|
| | | | | | | |
Date: 01 Nov 2006 11:22:47
From: Kenneth Sloan
Subject: Re: PROOF a beginner has no rating.
|
David Kane wrote:
> "Kenneth Sloan" <[email protected]> wrote in message
> news:[email protected]...
>> A practical rating system can, and does, make SOME (but not much) use
>> of this very fuzzy estimate. And, as a result, the ratings generated
>> are just a little bit better than they would be if the system did NOT
>> make use of this estimate.
>
> True, the rating system imputes initial ratings
> to unrated players. In that case it makes very good sense
> to give unrateds a starting rating that jibes with the rating
> that that class had historically achieved in their first events.
>
> But that is something entirely different than assigning that
> rating to random adults who know only the rules of chess.
> 99+% of people who know the rules have never played
> a single USCF rated game. You cannot hope to reliably
> extrapolate the performance of the 99% by looking at the
> 1%.
>
>
>
You're not a statistician, are you?
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | | | | | | |
Date: 01 Nov 2006 09:32:23
From: David Kane
Subject: Re: PROOF a beginner has no rating.
|
"Kenneth Sloan" <[email protected] > wrote in message
news:[email protected]...
> David Kane wrote:
>> "Kenneth Sloan" <[email protected]> wrote in message
>> news:[email protected]...
>>> A practical rating system can, and does, make SOME (but not much) use
>>> of this very fuzzy estimate. And, as a result, the ratings generated
>>> are just a little bit better than they would be if the system did NOT
>>> make use of this estimate.
>>
>> True, the rating system imputes initial ratings
>> to unrated players. In that case it makes very good sense
>> to give unrateds a starting rating that jibes with the rating
>> that that class had historically achieved in their first events.
>>
>> But that is something entirely different than assigning that
>> rating to random adults who know only the rules of chess.
>> 99+% of people who know the rules have never played
>> a single USCF rated game. You cannot hope to reliably
>> extrapolate the performance of the 99% by looking at the
>> 1%.
>>
>>
>>
>
> You're not a statistician, are you?
>
You are guilty of what I have heard described the
lazy scientist syndrome. It's the belief that the data you
have on your desk must answer your question.
It's much easier to process data in hand than
to find the correct data. Yet still invalid.
|
| | | | | | |
Date: 01 Nov 2006 05:47:28
From: Dave (from the UK)
Subject: Re: PROOF a beginner has no rating.
|
Kenneth Sloan wrote:
> Such complicated thinking on such a simple topic.
Maybe.
> A rating is nothing more or less than a prediction about future
> performance. It's a single position on a scale.
Agreed
> One way to assign a rating is to observe the player's results.
Yes, obviously.
> I still think my estimate is not too far off. Find yourself a random
> adult of normal intelligence, teach him the moves and enter him in
> 5 USCF events of 5 rounds each. I claim that a good estimate of his
> performance in those events is that the performance will be that
> of a USCF 1000 rated player.
That is after 25 games !!! I thought other threads we were talking about
someone who had JUST learned the rules. Hence they would not have the
experience of playing 25 games.
I taught my wifes grandaugther the rules and played a few games with
her. It was obvious she was better significantly better after the third
game than the first.
What was my point that simply by making the measurement of a rating, you
are affecting the rating.
--
Dave (from the UK)
Please note my email address changes periodically to avoid spam.
It is always of the form: [email protected]
Hitting reply will work for a few months only - later set it manually.
http://witm.sourceforge.net/ (Web based Mathematica front end)
|
| | | | | | | |
Date: 01 Nov 2006 00:25:41
From: Kenneth Sloan
Subject: Re: PROOF a beginner has no rating.
|
Dave (from the UK) wrote:
> Kenneth Sloan wrote:
>> Such complicated thinking on such a simple topic.
>
> Maybe.
>
>> A rating is nothing more or less than a prediction about future
>> performance. It's a single position on a scale.
>
> Agreed
>
>> One way to assign a rating is to observe the player's results.
>
> Yes, obviously.
>
>
>> I still think my estimate is not too far off. Find yourself a random
>> adult of normal intelligence, teach him the moves and enter him in
>> 5 USCF events of 5 rounds each. I claim that a good estimate of his
>> performance in those events is that the performance will be that
>> of a USCF 1000 rated player.
>
> That is after 25 games !!! I thought other threads we were talking about
> someone who had JUST learned the rules. Hence they would not have the
> experience of playing 25 games.
No. The *prediction* is made *before* the games are played.
That's what a rating is: a prediction of future performance.
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | | | |
Date: 31 Oct 2006 20:13:03
From: David Kane
Subject: Re: PROOF a beginner has no rating.
|
"Dave (from the UK)" <[email protected] > wrote in
message news:[email protected]...
> David Kane wrote:
>> "Dave (from the UK)" <[email protected]> wrote
>> in message news:[email protected]...
>>
>>
>>>I believe the assumptions Elo made are simply not valid in the case of an
>>>absolute beginner. As such, attaching an ELO number to something where the
>>>assumptions are very wrong is not sensible.
>>>
>>
>>
>> Of course they aren't.
>
> So we agree!
>
>> But what would be your
>> response if I were to claim in Ken Sloan
>> fashion that an adult of average intelligence
>> who just knows the rules has a rating of 2700?
>
> A large number of players who play the game seriously don't have a rating of
> 2700, so it's fairly easy to disprove that one. Putting an upper bound is
> fairly easy. Anyone who thinks it would be over 1200 is really got to be mad.
Similarly, those with experience with players in the
x00's range, also believe that 1000 is mad.
>> The fact that there is uncertainty to ratings doesn't
>> make the discussion meaningless.
>
>
>> By the way, I think in the USCF system the rating
>> formulae change at about 8 games. The USCF rating
>> algorithm assigns beginners a very high K-constant
>> (80, I think, which can be effectively doubled via
>> "bonus" points)
>
> I don't know how the USCF method works, but I assume this K-constant you talk
> of is some measure of the standard deviation (uncertainty) of the rating. For
> a new member, this will be higher until they are established.
>
>> So the number of games actually
>> going into a rating is much smaller than the number
>> which defines a "provisional" rating. Someone can
>> correct me, but I don't think provisional ratings are
>> treated differently in the rating algorithm.
>
> Anyone who is a member of the USCF would have to be reasonably serious about
> the game, so they would have played many times before. I don't believe someone
> will learn the rules, then join USCF and play their first ever game in a rated
> competition.
Not true. For better of for worse, many kids do start playing rated
games knowing very little.
> As such, it should be possible to determine a rating reasonably quickly, as
> they performance is unlikely to change a huge amount from game to game.
>
If your point is that ratings mean more for some than others, then
certainly that is true.
> In contrast, as absolute beginner who only just learned the rules, playing his
> first few games, should learn a lot very quickly. They should soon learn about
> the fact knights can fork. So its very likely the large K-constant you talk of
> in USCF rating systems is not sufficiently large. So I don't believe you can
> rely on the USCF system in this case. In any case, the USCF system would have
> been designed for serious chess players - not those who just learned the
> rules.
The value of the K-constant is how much a rating can change from
a single game. I have no idea whether the USCF method
has been proven optimal for all ratings but it covers all ratings.
I only brought it up because the number of games that a
rating is "provisional" doesn't necessarily have anything to
do with the number of games for a rating to equilibrate
in the rating formulae (~800/K).
> One *possible* way to establish a rating for absolute beginners might be as
> follows.
>
> Take 1000 beginners and let them play their *first* ever chess game against a
> low rated player (say 1200 for example). Results of any of their later games
> are ignored. Each person plays just one game.
>
> Statistically, if there is a rating difference of x between two players, the
> probability of wins and losses can be computed. The 1200 players should beat
> the beginners in most games, but the percentage might allow you to determine
> how much stronger the 1200 players and and so assign an average rating for
> those 1000 people after just one game. That would be VERY doggy, and not very
> practical to do, but at least it would give you a result from only one game,
> so they don't have any chance to improve.
>
> I can't help feeling the concept of a rating for someone who has just learned
> the rules is absurd. You can put a lower limit of 0, an upper limit of 1200,
> but I don't believe you can say anything else really.
>
As mentioned elsewhere, 0 has no special significance in the rating system.
In fact the minimum rating in the USCF system is 100. 1200 (or 1000, or 800)
can be refuted in the same way that 2700 can be refuted - by showing
that players with those ratings can beat average intelligence
adults. (Whether the adults have strictly just learned the rules is
just a detail, unless you are suggesting that player's ratings would
*decline* by playing)
|
| | | | | | |
Date: 01 Nov 2006 05:07:59
From: Dave (from the UK)
Subject: Re: PROOF a beginner has no rating.
|
David Kane wrote:
>>A large number of players who play the game seriously don't have a rating of
>>2700, so it's fairly easy to disprove that one. Putting an upper bound is
>>fairly easy. Anyone who thinks it would be over 1200 is really got to be mad.
>
>
> Similarly, those with experience with players in the
> x00's range, also believe that 1000 is mad.
I don't doubt that. If I had said 1000, there might have been some
argument about it, so 1200 seemed a sensible upper limit.
>>Anyone who is a member of the USCF would have to be reasonably serious about
>>the game, so they would have played many times before. I don't believe someone
>>will learn the rules, then join USCF and play their first ever game in a rated
>>competition.
>
>
> Not true. For better of for worse, many kids do start playing rated
> games knowing very little.
Having just learned the rules? Which is what I thought we were talking
about.
I could see that someone might learn the rules of chess from a book or
the Internet and not having a human opponent, go onto ICC, FICS or
whatever and play their first game. I doubt they would join USCF and
play their first game there.
I guess there is always the parent who wants his/her child to play
chess, so they pay for membership and get them doing it from the
beginning. I used to teach maths to someone who did not really want to
learn mathematics, but his Dad wanted him to. (I'm not a mathematician
BTW).
> As mentioned elsewhere, 0 has no special significance in the rating system.
Sorry, I thought it did, but I see you are right.
> In fact the minimum rating in the USCF system is 100. 1200 (or 1000, or 800)
> can be refuted in the same way that 2700 can be refuted - by showing
> that players with those ratings can beat average intelligence
> adults. (Whether the adults have strictly just learned the rules is
> just a detail, unless you are suggesting that player's ratings would
> *decline* by playing)
I was talking about someone who had JUST LEARNED THE RULES, which is
what I believed the other threads were referring to. In that
circumstance, by playing several games an average (or even somewhat
below average) intelligence human can't fail to learn and so improve.
--
Dave (from the UK)
Please note my email address changes periodically to avoid spam.
It is always of the form: [email protected]
Hitting reply will work for a few months only - later set it manually.
http://witm.sourceforge.net/ (Web based Mathematica front end)
|
| | | | | | | |
Date: 31 Oct 2006 22:42:26
From: David Kane
Subject: Re: PROOF a beginner has no rating.
|
"Dave (from the UK)" <[email protected] > wrote in
message news:[email protected]...
> David Kane wrote:
>
>>>A large number of players who play the game seriously don't have a rating of
>>>2700, so it's fairly easy to disprove that one. Putting an upper bound is
>>>fairly easy. Anyone who thinks it would be over 1200 is really got to be mad.
>>
>>
>> Similarly, those with experience with players in the
>> x00's range, also believe that 1000 is mad.
>
> I don't doubt that. If I had said 1000, there might have been some argument
> about it, so 1200 seemed a sensible upper limit.
>
>>>Anyone who is a member of the USCF would have to be reasonably serious about
>>>the game, so they would have played many times before. I don't believe
>>>someone
>>>will learn the rules, then join USCF and play their first ever game in a
>>>rated
>>>competition.
>>
>>
>> Not true. For better of for worse, many kids do start playing rated
>> games knowing very little.
>
> Having just learned the rules? Which is what I thought we were talking about.
You can always quibble about how close to learning the rules someone is.
But no doubt tournament-playing kids are far closer to that situation than
tournament-playing adults. The mistake Prof. Sloan was making was equating
new adult tournament players with average adult players who have just learned
the rules.
> I could see that someone might learn the rules of chess from a book or the
> Internet and not having a human opponent, go onto ICC, FICS or whatever and
> play their first game. I doubt they would join USCF and play their first game
> there.
>
> I guess there is always the parent who wants his/her child to play chess, so
> they pay for membership and get them doing it from the beginning. I used to
> teach maths to someone who did not really want to learn mathematics, but his
> Dad wanted him to. (I'm not a mathematician BTW).
>
>> As mentioned elsewhere, 0 has no special significance in the rating system.
>
> Sorry, I thought it did, but I see you are right.
>
>> In fact the minimum rating in the USCF system is 100. 1200 (or 1000, or 800)
>> can be refuted in the same way that 2700 can be refuted - by showing
>> that players with those ratings can beat average intelligence
>> adults. (Whether the adults have strictly just learned the rules is
>> just a detail, unless you are suggesting that player's ratings would
>> *decline* by playing)
>
> I was talking about someone who had JUST LEARNED THE RULES, which is what I
> believed the other threads were referring to. In that circumstance, by playing
> several games an average (or even somewhat below average) intelligence human
> can't fail to learn and so improve.
>
I'm not denying that you have a point. But it only comes into play
when we reach the rating that average beginner adults cannot play at.
Then we can obsess about how close to "just learned the
rules" they are.
|
| | | | | | | |
Date: 01 Nov 2006 00:44:43
From: Ange1o DePa1ma
Subject: Re: PROOF a beginner has no rating.
|
The "Heisenberg" theory is cute, but it only goes as far as the "measurement
affects outcome" idea.
Sloan may be right that the scale is arbitrary, and we may employ negative
numbers, fractions, or even imaginary numbers if we like. We can make an
average rating zero instead of 1500.
But in the real world that is an unnecessary complication for a very simple
idea. We normally think of achievement in terms of numbers from zero to some
higher number, for example:
* I own six chess trophies (when I was 11 I had none)
* Federer's world tennis association ranking is based on some positive
number of points earned in tournaments (at some point he had no points, and
no ranking)
* I have several hundred ACBL master points (at one time I had zero)
What is the sense of using negative numbers here? I never had minus-three
chess trophies; I never won minus-two tournaments. Why should my rating be
negative?
"Dave (from the UK)" <[email protected] >
wrote in message news:[email protected]...
> David Kane wrote:
>
>>>A large number of players who play the game seriously don't have a rating
>>>of
>>>2700, so it's fairly easy to disprove that one. Putting an upper bound is
>>>fairly easy. Anyone who thinks it would be over 1200 is really got to be
>>>mad.
>>
>>
>> Similarly, those with experience with players in the
>> x00's range, also believe that 1000 is mad.
>
> I don't doubt that. If I had said 1000, there might have been some
> argument about it, so 1200 seemed a sensible upper limit.
>
>>>Anyone who is a member of the USCF would have to be reasonably serious
>>>about
>>>the game, so they would have played many times before. I don't believe
>>>someone
>>>will learn the rules, then join USCF and play their first ever game in a
>>>rated
>>>competition.
>>
>>
>> Not true. For better of for worse, many kids do start playing rated
>> games knowing very little.
>
> Having just learned the rules? Which is what I thought we were talking
> about.
>
> I could see that someone might learn the rules of chess from a book or the
> Internet and not having a human opponent, go onto ICC, FICS or whatever
> and play their first game. I doubt they would join USCF and play their
> first game there.
>
> I guess there is always the parent who wants his/her child to play chess,
> so they pay for membership and get them doing it from the beginning. I
> used to teach maths to someone who did not really want to learn
> mathematics, but his Dad wanted him to. (I'm not a mathematician BTW).
>
>> As mentioned elsewhere, 0 has no special significance in the rating
>> system.
>
> Sorry, I thought it did, but I see you are right.
>
>> In fact the minimum rating in the USCF system is 100. 1200 (or 1000, or
>> 800)
>> can be refuted in the same way that 2700 can be refuted - by showing
>> that players with those ratings can beat average intelligence
>> adults. (Whether the adults have strictly just learned the rules is
>> just a detail, unless you are suggesting that player's ratings would
>> *decline* by playing)
>
> I was talking about someone who had JUST LEARNED THE RULES, which is what
> I believed the other threads were referring to. In that circumstance, by
> playing several games an average (or even somewhat below average)
> intelligence human can't fail to learn and so improve.
>
>
> --
> Dave (from the UK)
>
> Please note my email address changes periodically to avoid spam.
> It is always of the form: [email protected]
> Hitting reply will work for a few months only - later set it manually.
>
> http://witm.sourceforge.net/ (Web based Mathematica front end)
|
| | | | | | | | |
Date: 01 Nov 2006 15:22:09
From: Dr A. N. Walker
Subject: Re: PROOF a beginner has no rating.
|
In article <[email protected] >,
Ange1o DePa1ma <[email protected] > wrote:
>Sloan may be right that the scale is arbitrary, and we may employ negative
>numbers, fractions,
Many years ago, it occurred to me that we don't always set
exam papers of identical strength, and our students certainly aren't
all the same, so it made sense to think of the students as "playing
against" the exams. Use the Elo system to rate students and exams,
and lo! it no longer matters that [in a system with many options]
not all the exams are of equal difficulty. We were processing the
ks by computer anyway, so it was "easy" to insert a rating step,
and then the class lists could just be produced from the list of
students in rating order. Equally, those colleagues who year on
year produced exams that were too easy or too hard could easily be
indentified.
The trouble was that some weak students scored very badly
even on easy papers; and some excellent ones scored full ks
even on a difficult paper. Colleagues, despite [or perhaps because
of] being mathematicians, had difficulty with the notion of scoring
below 0 or above 100% on a test. In vain did we point out that what
mattered was not the individual "games" but the overall performance.
It was as though they thought that Kasparov was literally invincible
simply because he was the best player around.
In such contexts, the scale is far from arbitary; people
have views about what 0 and 100 mean [even if they are wrong]. In
the end, there was a majority vote to scrap the system [after nearly
30 years and several simple explanatory documents] on the grounds
that "no-one understood it". It was replaced by a statistical
process that indeed no-one understands, but that was OK because
it was statistics and so "proper" mathematics.
> or even imaginary numbers if we like.
That too was proposed, for "quality" ks -- bonuses for
doing complete questions instead of building up part ks, or for
thinking of a solution that was better than the "official" one.
The idea was zapped, largely because we switched from Algol, which
did understand complex numbers, as the programming language for our
exam ks, to C/Unix, which didn't.
--
Andy Walker, School of MathSci., Univ. of Nott'm, UK.
[email protected]
|
| | | | | | | | |
Date: 01 Nov 2006 00:20:12
From: Kenneth Sloan
Subject: Re: PROOF a beginner has no rating.
|
Ange1o DePa1ma wrote:
> The "Heisenberg" theory is cute, but it only goes as far as the "measurement
> affects outcome" idea.
>
> Sloan may be right that the scale is arbitrary, and we may employ negative
> numbers, fractions, or even imaginary numbers if we like. We can make an
> average rating zero instead of 1500.
Wherever did you get that the average rating was 1500?
Have you been talking to Dubeck?
>
> But in the real world that is an unnecessary complication for a very simple
> idea. We normally think of achievement in terms of numbers from zero to some
> higher number,
How you "think about it" has little relationship with what it is!
for example:
>
> * I own six chess trophies (when I was 11 I had none)
> * Federer's world tennis association ranking is based on some positive
> number of points earned in tournaments (at some point he had no points, and
> no ranking)
> * I have several hundred ACBL master points (at one time I had zero)
>
> What is the sense of using negative numbers here? I never had minus-three
> chess trophies; I never won minus-two tournaments. Why should my rating be
> negative?
Ratings are not master points. You don't start at zero and accumulate them
Your rating might be -0100 because you lose all your games to players
rated 0300...who lose all their games to players rated 0700 (but always
beat you). They, in turn, lose to players rated 1100...and so on up the
scale.
The results of games actually played, by the players who actually play
them, determine the RANGE of actual ratings. If the World Champion is
rated 2900, and the worst player in the world is 4000 points worse than
the World Champion, then we either need to add points to EVERYONE's
rating so that the World Champion is over 4000 (and the worst player
no longer has a negative rating)...OR...we can leave the ratings of
strong players where they have been historically and assign negative
ratings to the new wave of truly terrible players now playing chess.
USCF does not assign negative ratings (even though some players earned
them) - but only because the politicians think that innumerate Americans
can't handle the complication of the '-' sign. Alas, they are probably
right. One need only read this thread to see that innumeracy reigns.
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | | | | | | | |
Date: 01 Nov 2006 11:34:22
From: Ange1o DePa1ma
Subject: Re: PROOF a beginner has no rating.
|
"Kenneth Sloan" <[email protected] > wrote
>> Sloan may be right that the scale is arbitrary, and we may employ
>> negative numbers, fractions, or even imaginary numbers if we like. We can
>> make an average rating zero instead of 1500.
>
> Wherever did you get that the average rating was 1500?
> Have you been talking to Dubeck?
Jesus, Ken, I was just picking a number out of my hat. 1200. Whatever.
> Ratings are not master points. You don't start at zero and accumulate
> them
> Your rating might be -0100 because you lose all your games to players
> rated 0300...who lose all their games to players rated 0700 (but always
> beat you). They, in turn, lose to players rated 1100...and so on up the
> scale.
>
> The results of games actually played, by the players who actually play
> them, determine the RANGE of actual ratings. If the World Champion is
> rated 2900, and the worst player in the world is 4000 points worse than
> the World Champion, then we either need to add points to EVERYONE's rating
> so that the World Champion is over 4000 (and the worst player
> no longer has a negative rating)...OR...we can leave the ratings of
> strong players where they have been historically and assign negative
> ratings to the new wave of truly terrible players now playing chess.
I get the point. I'm not sure if it's mathematically relevant if you simply
define a know-nothing rating as zero. I'm not sure what the system gains by
going into negative numbers because as soon as you score 1/2 a point out of
100 games against players rated ELO 100, you have a positive rating. It may
be closer to zero than to one, but it is positive.
> USCF does not assign negative ratings (even though some players earned
> them) - but only because the politicians think that innumerate Americans
> can't handle the complication of the '-' sign. Alas, they are probably
> right. One need only read this thread to see that innumeracy reigns.
Americans' stupidity re math notwithstanding, there is no practical reason
for a negative rating.
|
| | | | | | | | | | |
Date: 01 Nov 2006 11:19:26
From: Kenneth Sloan
Subject: Re: PROOF a beginner has no rating.
|
Ange1o DePa1ma wrote:
> "Kenneth Sloan" <[email protected]> wrote
>
>>> Sloan may be right that the scale is arbitrary, and we may employ
>>> negative numbers, fractions, or even imaginary numbers if we like. We can
>>> make an average rating zero instead of 1500.
>> Wherever did you get that the average rating was 1500?
>> Have you been talking to Dubeck?
>
> Jesus, Ken, I was just picking a number out of my hat. 1200. Whatever.
>
The point is: the "average rating" is not an interesting concept when
specifying the rating system. It depends more on the demographics (who
chooses to play in your system) than on anything else.
The reason I complained is that there *are* people out there who believe
that the rating system is designed to produce a specific "average
rating" (and surprisingly many of these people think that the magic
number is 1500).
>
> I get the point. I'm not sure if it's mathematically relevant if you simply
> define a know-nothing rating as zero.
It may be 'relevant' - but if so, it's flat out wrong!
In any Elo-based system, the only "correct" rating for a player who
truly knows nothing is: -infinity.
>I'm not sure what the system gains by
> going into negative numbers because as soon as you score 1/2 a point out of
> 100 games against players rated ELO 100, you have a positive rating. It may
> be closer to zero than to one, but it is positive.
that's not correct. If you score 0.5-99.5 against 0100 opposition, your
proper Elo rating is most definitely NOT positive. Do the math.
But, at least it's no longer -infinity!
Remember - what matters is the rating DIFFERENCE between you and all
those 0100 players who are beating up on you. What rating DIFFERENCE
predicts a score of 99.5-0.5 (look it up!). Your new rating after going
0.5-99.5 against 0100 competition is 0100-thatDifference. Which, I
assure you, is a negative number.
Why, exactly, do you think that your rating would be greater than 0000
after that performance? Because USCF has an Absolute Floor of 0100?
That's an administrative twiddle (mostly meant to pacify innumerate folk
who believe that ratings must be positive).
>
>
>> USCF does not assign negative ratings (even though some players earned
>> them) - but only because the politicians think that innumerate Americans
>> can't handle the complication of the '-' sign. Alas, they are probably
>> right. One need only read this thread to see that innumeracy reigns.
>
> Americans' stupidity re math notwithstanding, there is no practical reason
> for a negative rating.
Do you know the real reason that the USCF minimum rating is 0100, and
not 0000?
You wouldn't believe me if I told you - it's too silly for words.
In any event, there is indeed a "practical reason for a negative
rating". That reason is: the negative rating more accuratly predicts
the future performance of that player (e.g., that player will score 0.5
out of 100 against 0100 competition).
USCF has decided that this valid reason is counterbalanced by other
reasons (that's how you run a real system - you end up compromising).
But...you'd really laugh if you heard the *real* reason that the minimum
USCF rating is 0100.
HINT: it's closely related to the reason that the minimum USCF rating
used to be (many years ago) 1000.
>
>
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | | | | | | | | | |
Date: 01 Nov 2006 17:58:24
From: Ange1o DePa1ma
Subject: Re: PROOF a beginner has no rating.
|
"Kenneth Sloan" <[email protected] > wrote
> USCF has decided that this valid reason is counterbalanced by other
> reasons (that's how you run a real system - you end up compromising).
> But...you'd really laugh if you heard the *real* reason that the minimum
> USCF rating is 0100.
>
> HINT: it's closely related to the reason that the minimum USCF rating used
> to be (many years ago) 1000.
Because there are 100 years in a century?
|
| | | | | | | | | | | | |
Date: 01 Nov 2006 19:50:19
From: Kenneth Sloan
Subject: Re: PROOF a beginner has no rating.
|
Ange1o DePa1ma wrote:
> "Kenneth Sloan" <[email protected]> wrote
>
>> USCF has decided that this valid reason is counterbalanced by other
>> reasons (that's how you run a real system - you end up compromising).
>> But...you'd really laugh if you heard the *real* reason that the minimum
>> USCF rating is 0100.
>>
>> HINT: it's closely related to the reason that the minimum USCF rating used
>> to be (many years ago) 1000.
>
> Because there are 100 years in a century?
>
>
No. In that case, the minimum rating would be 0999, of course.
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | | | | | | | | | |
Date: 01 Nov 2006 09:53:31
From: David Kane
Subject: Re: PROOF a beginner has no rating.
|
"Kenneth Sloan" <[email protected] > wrote in message
news:[email protected]...
>
> In any Elo-based system, the only "correct" rating for a player who truly
> knows nothing is: -infinity.
This is only true if "knows nothing" equates to "never wins" which you
haven't proven. In fact, if we rated a tournament filled only
with unrated "know-nothings", they'd end up with the rating arbitrarily
assigned to their class by the rating system. No system that I know
of uses -infinity.
|
| | | | | | | | | | | | |
Date: 01 Nov 2006 23:26:48
From: alexmagnus
Subject: Re: PROOF a beginner has no rating.
|
David Kane Wrote:
> "Kenneth Sloan" [email protected] wrote in message
> news:[email protected]...
> -
>
> In any Elo-based system, the only "correct" rating for a player wh
> truly
> knows nothing is: -infinity.-
>
> This is only true if "knows nothing" equates to "never wins" which you
> haven't proven. In fact, if we rated a tournament filled only
> with unrated "know-nothings", they'd end up with the ratin
> arbitrarily
> assigned to their class by the rating system. No system that I know
> of uses -infinity.
Don't knw how the Elo-based ratings, but the BCF-rating may becom
negative (I once saw a player with BCF -3, which corresponds to Elo 57
if we use the usual formula)
--
alexmagnus
|
| | | | | | | | | | | | |
Date: 01 Nov 2006 15:28:53
From: Kenneth Sloan
Subject: Re: PROOF a beginner has no rating.
|
David Kane wrote:
> "Kenneth Sloan" <[email protected]> wrote in message
> news:[email protected]...
>
>> In any Elo-based system, the only "correct" rating for a player who truly
>> knows nothing is: -infinity.
>
> This is only true if "knows nothing" equates to "never wins" which you
> haven't proven. In fact, if we rated a tournament filled only
> with unrated "know-nothings", they'd end up with the rating arbitrarily
> assigned to their class by the rating system. No system that I know
> of uses -infinity.
uses -infinity as *what*? As the initial rating for a first-time
player? Well, of course not - THAT WAS MY POINT!
As the lowest possible rating? USCF does. In *calculating* ratings,
any value (even negative values) are permitted. Only at the end is
an administrative floor imposed.
At one time, the programming assumed a minimum of 0000 - but did not
really impose it (actually, it assumed a minimum of 0001, because 0000
was reserved to mean "UNR". That is, until it was noticed that ratings
for some players were fast approaching 0000, and it was only a matter of
time until a computed rating was actually negative. That's what sparked
the discussion of the Absolute Floor.
Participants in that discussion may recall that I argued for a Floor
of 0400. The end result was a floor at 0100. The official reason
given: "players are embarassed to have double-digit ratings".
The lowest rating ever published (if memory serves) was 0040. There
are now hundreds of players who are being propped up by the floor at
0100 (properly so, in my opinion - for reasons beyond the scope of this
thread), but the basic Elo computation
takes absolutely NO NOTICE of ZERO, for any purpose whatsoever (don't
believe me? read Elo!). There is nothing at all in the current
USCF rating system that prevents a rating from going negative (and,..as
negative as you please, which is a pretty good definition of -infinity).
The ONLY reason you don't see negative ratings reported is the
politically determined Absolute Floor at 0100.
Exercise for statistics journeymen: look at the distribution of USCF
ratings and estimate the lowest "true" USCF rating as of today.
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | | | | | | | | | | | |
Date: 01 Nov 2006 15:12:26
From: David Kane
Subject: Re: PROOF a beginner has no rating.
|
"Kenneth Sloan" <[email protected] > wrote in message
news:[email protected]...
> David Kane wrote:
>> "Kenneth Sloan" <[email protected]> wrote in message
>> news:[email protected]...
>>
>>> In any Elo-based system, the only "correct" rating for a player who truly
>>> knows nothing is: -infinity.
>>
>> This is only true if "knows nothing" equates to "never wins" which you
>> haven't proven. In fact, if we rated a tournament filled only
>> with unrated "know-nothings", they'd end up with the rating arbitrarily
>> assigned to their class by the rating system. No system that I know
>> of uses -infinity.
>
> uses -infinity as *what*? As the initial rating for a first-time player?
> Well, of course not - THAT WAS MY POINT!
Uhh. Actually you've made a number of points and claims.
Including the false one at the top of this post. I agree that
-infinity is the correct rating for someone who loses all their
games. You haven't shown that someone who "knows nothing"
is in that category. If two players who "know
nothing" play each other, someone will still win.
> As the lowest possible rating? USCF does. In *calculating* ratings,
> any value (even negative values) are permitted. Only at the end is
> an administrative floor imposed.
>
> At one time, the programming assumed a minimum of 0000 - but did not
> really impose it (actually, it assumed a minimum of 0001, because 0000 was
> reserved to mean "UNR". That is, until it was noticed that ratings
> for some players were fast approaching 0000, and it was only a matter of time
> until a computed rating was actually negative. That's what sparked
> the discussion of the Absolute Floor.
>
> Participants in that discussion may recall that I argued for a Floor
> of 0400. The end result was a floor at 0100. The official reason given:
> "players are embarassed to have double-digit ratings".
FYI, the rating system in use here (NWSRS) does use
a floor of 400. This leads to ratings generally higher
than the USCF. 23% of the people in the database
are at the floor. This sounds bad, but in practice the vast majority
of them are inactive. (Played a tournament, didn't like it, quit)
> The lowest rating ever published (if memory serves) was 0040. There
> are now hundreds of players who are being propped up by the floor at
> 0100 (properly so, in my opinion - for reasons beyond the scope of this
> thread), but the basic Elo computation
> takes absolutely NO NOTICE of ZERO, for any purpose whatsoever (don't
> believe me? read Elo!). There is nothing at all in the current
> USCF rating system that prevents a rating from going negative (and,..as
> negative as you please, which is a pretty good definition of -infinity).
> The ONLY reason you don't see negative ratings reported is the
> politically determined Absolute Floor at 0100.
>
> Exercise for statistics journeymen: look at the distribution of USCF ratings
> and estimate the lowest "true" USCF rating as of today.
>
|
| | | | | | | | | | | | | | |
Date: 01 Nov 2006 19:49:33
From: Kenneth Sloan
Subject: Re: PROOF a beginner has no rating.
|
David Kane wrote:
> "Kenneth Sloan" <[email protected]> wrote in message
> news:[email protected]...
>> David Kane wrote:
>>> "Kenneth Sloan" <[email protected]> wrote in message
>>> news:[email protected]...
>>>
>>>> In any Elo-based system, the only "correct" rating for a player who truly
>>>> knows nothing is: -infinity.
>>> This is only true if "knows nothing" equates to "never wins" which you
>>> haven't proven. In fact, if we rated a tournament filled only
>>> with unrated "know-nothings", they'd end up with the rating arbitrarily
>>> assigned to their class by the rating system. No system that I know
>>> of uses -infinity.
>> uses -infinity as *what*? As the initial rating for a first-time player?
>> Well, of course not - THAT WAS MY POINT!
>
> Uhh. Actually you've made a number of points and claims.
> Including the false one at the top of this post. I agree that
> -infinity is the correct rating for someone who loses all their
> games. You haven't shown that someone who "knows nothing"
> is in that category. If two players who "know
> nothing" play each other, someone will still win.
Two people do not a rating pool make.
If two people "know nothing" (literally), then it may be that one of
them may win (they may, of course, draw - a much more likely event:
drawn by adjudication of the arbiter), but even if one of them wins, the
*correct* rating for BOTH of these players is -infinity (given that they
will surely lose all of their games against anyone who actually knows
how to play the game). The nice thing about -infinity as a rating for
both of them is that -infinity + 400 (the usual, back-of-the-envelope
calculation) = -infinity!
If the two players who "know nothing" play a long series of games,
I trust that they will split the available points - so their ratings
should be identical. How they rate against the rest of the world
depends, of course, on the results of games played against the rest
of the world.
If they truly "know nothing", and play a significant number of games
against USCF 0600+ players, I have great confidence that the correct
rating will be -infinity. (Knowing nothing, I expect most games to end
when their flag drops on move 1.)
But, enough of -infinity. How about -0400? Is it too difficult for you
to imagine a player whose appropriate rating on the FIDE scale (notice I
didn't say FIDE rating) should be -0400? That means they score 1-3
against -0200 competition. The -0200 folk, of course, score 1-3 against
0000 competetiton...and so on, up the ladder to someone who scores 1-3
against Kasparov.
>
>> As the lowest possible rating? USCF does. In *calculating* ratings,
>> any value (even negative values) are permitted. Only at the end is
>> an administrative floor imposed.
>>
>> At one time, the programming assumed a minimum of 0000 - but did not
>> really impose it (actually, it assumed a minimum of 0001, because 0000 was
>> reserved to mean "UNR". That is, until it was noticed that ratings
>> for some players were fast approaching 0000, and it was only a matter of time
>> until a computed rating was actually negative. That's what sparked
>> the discussion of the Absolute Floor.
>>
>> Participants in that discussion may recall that I argued for a Floor
>> of 0400. The end result was a floor at 0100. The official reason given:
>> "players are embarassed to have double-digit ratings".
>
> FYI, the rating system in use here (NWSRS) does use
> a floor of 400. This leads to ratings generally higher
> than the USCF. 23% of the people in the database
> are at the floor. This sounds bad, but in practice the vast majority
> of them are inactive. (Played a tournament, didn't like it, quit)
I'd be interested in statistics (ratings distributions, etc.) on this
rating system. I'd be *especially* interested in a set of ratings for
players who are also rated in other systems (USCF, FIDE, anything!).
You say "ratings generally higher than the USCF. I'm not sure what you
mean by that. Can you please explain?
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | | | | | | | | | | | | | |
Date: 01 Nov 2006 22:44:12
From: David Kane
Subject: Re: PROOF a beginner has no rating.
|
"Kenneth Sloan" <[email protected] > wrote in message
news:[email protected]...
> David Kane wrote:
>> "Kenneth Sloan" <[email protected]> wrote in message
>> news:[email protected]...
>>> David Kane wrote:
>>>> "Kenneth Sloan" <[email protected]> wrote in message
>>>> news:[email protected]...
>>>>
>>>>> In any Elo-based system, the only "correct" rating for a player who truly
>>>>> knows nothing is: -infinity.
>>>> This is only true if "knows nothing" equates to "never wins" which you
>>>> haven't proven. In fact, if we rated a tournament filled only
>>>> with unrated "know-nothings", they'd end up with the rating arbitrarily
>>>> assigned to their class by the rating system. No system that I know
>>>> of uses -infinity.
>>> uses -infinity as *what*? As the initial rating for a first-time player?
>>> Well, of course not - THAT WAS MY POINT!
>>
>> Uhh. Actually you've made a number of points and claims.
>> Including the false one at the top of this post. I agree that
>> -infinity is the correct rating for someone who loses all their
>> games. You haven't shown that someone who "knows nothing"
>> is in that category. If two players who "know
>> nothing" play each other, someone will still win.
>
> Two people do not a rating pool make.
I never said that they do.
> If two people "know nothing" (literally), then it may be that one of them may
> win (they may, of course, draw - a much more likely event: drawn by
> adjudication of the arbiter), but even if one of them wins, the *correct*
> rating for BOTH of these players is -infinity (given that they
> will surely lose all of their games against anyone who actually knows how to
> play the game).
The nice thing about -infinity as a rating for
> both of them is that -infinity + 400 (the usual, back-of-the-envelope
> calculation) = -infinity!
>
> If the two players who "know nothing" play a long series of games,
> I trust that they will split the available points - so their ratings
> should be identical. How they rate against the rest of the world depends, of
> course, on the results of games played against the rest
> of the world.
>
> If they truly "know nothing", and play a significant number of games
> against USCF 0600+ players, I have great confidence that the correct
> rating will be -infinity. (Knowing nothing, I expect most games to end when
> their flag drops on move 1.)
>
No. Against people who "know almost nothing", they'll win 25%. Against
people who "know a little more than almost nothing, they'll win 10% etc.
All the way up to Topalov.
> But, enough of -infinity. How about -0400? Is it too difficult for you to
> imagine a player whose appropriate rating on the FIDE scale (notice I didn't
> say FIDE rating) should be -0400? That means they score 1-3 against -0200
> competition. The -0200 folk, of course, score 1-3 against 0000
> competetiton...and so on, up the ladder to someone who scores 1-3 against
> Kasparov.
I'm not arguing against negative ratings. I'm pointing out the incorrectness of
your
statement that the only "correct" rating for a player who truly knows nothing
is: -infinity.
>>
>>> As the lowest possible rating? USCF does. In *calculating* ratings,
>>> any value (even negative values) are permitted. Only at the end is
>>> an administrative floor imposed.
>>>
>>> At one time, the programming assumed a minimum of 0000 - but did not
>>> really impose it (actually, it assumed a minimum of 0001, because 0000 was
>>> reserved to mean "UNR". That is, until it was noticed that ratings
>>> for some players were fast approaching 0000, and it was only a matter of
>>> time
>>> until a computed rating was actually negative. That's what sparked
>>> the discussion of the Absolute Floor.
>>>
>>> Participants in that discussion may recall that I argued for a Floor
>>> of 0400. The end result was a floor at 0100. The official reason given:
>>> "players are embarassed to have double-digit ratings".
>>
>> FYI, the rating system in use here (NWSRS) does use
>> a floor of 400. This leads to ratings generally higher
>> than the USCF. 23% of the people in the database
>> are at the floor. This sounds bad, but in practice the vast majority
>> of them are inactive. (Played a tournament, didn't like it, quit)
>
> I'd be interested in statistics (ratings distributions, etc.) on this
> rating system. I'd be *especially* interested in a set of ratings for players
> who are also rated in other systems (USCF, FIDE, anything!).
http://www.whsca.org/ratings.html
But keep in mind that it is not the norm to dual-rate events.
So the difference between the ratings is not solely caused by the
minor differences in the algorithm (the floor, also some differences in
imputing initial ratings), but because they involve different rating events.
However, there are a fair number who play regularly in both systems.
I did note that there are about 100 players who are at the floor even
though they've played 4 tournaments.
> You say "ratings generally higher than the USCF. I'm not sure what you mean
> by that. Can you please explain?
>
There are 9104 in the database. 1974 have a USCF rating. 784 have a
non-provisional USCF rating. The NWSRS has a practice of
periodically bumping up the NWSRS ratings of those with
higher non-provisional USCF ratings. Currently there are
156 players whose rating have been bumped up in that way.
This suggests that most have higher NWSRS ratings.
|
| | | | | | | | | | | | | | | | |
Date: 02 Nov 2006 09:53:52
From: Kenneth Sloan
Subject: Re: PROOF a beginner has no rating.
|
David Kane wrote:
>
> There are 9104 in the database. 1974 have a USCF rating. 784 have a
> non-provisional USCF rating. The NWSRS has a practice of
> periodically bumping up the NWSRS ratings of those with
> higher non-provisional USCF ratings. Currently there are
> 156 players whose rating have been bumped up in that way.
> This suggests that most have higher NWSRS ratings.
They bump ratings UP to conform to USCF but not DOWN? And you think
it's the higher floor that generates higher ratings?
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | | | | | | | | | | | | | | | |
Date: 02 Nov 2006 10:46:54
From: David Kane
Subject: Re: PROOF a beginner has no rating.
|
"Kenneth Sloan" <[email protected] > wrote in message
news:[email protected]...
> David Kane wrote:
>
>>
>> There are 9104 in the database. 1974 have a USCF rating. 784 have a
>> non-provisional USCF rating. The NWSRS has a practice of
>> periodically bumping up the NWSRS ratings of those with
>> higher non-provisional USCF ratings. Currently there are
>> 156 players whose rating have been bumped up in that way.
>> This suggests that most have higher NWSRS ratings.
>
>
> They bump ratings UP to conform to USCF but not DOWN? And you think
> it's the higher floor that generates higher ratings?
>
As I said, the number who receive points that way is minimal. I think it is
an odd practice, to say the least, but it's inflationary effect is very small.
Some
of those receiving points that way are the older players who have stopped
playing games in the NWSRS, so it has no effect other than symbolic.
The bump occurs once or twice per year and the number of points added is small.
On the other hand, people playing at their floor pump points into
the system every tournament. Also, the initially imputed rating
is in most cases higher than the USCF - it's changed over the years
and the details have never been clear. But for most age groups its
higher than 50*Age.
Also, consider the "bonus point" mechanism for inflating the ratings.
Imagine two identical pools of players, but players in pool A play
twice as frequenctly as those in Pool B. Ratings in Pool A will inflate
faster than those of Pool B. My guess is that our pool is more active
than the USCF Pool.
And ratings of players who play more frequently in the NWSRS will
not have the lag that their USCF ratings have. There have
been comic misjustices in tournaments where USCF ratings were used -
e.g. one boy won the "biggest upset" prize every round even though
in the NWSRS system he had the highest rating and was winning
as expected. But his USCF rating was several years old, and
600 points out of date.
I don't believe trying to align the scales would have any practical
advantages.
|
| | | | | | | | | | | | | | | | | | |
Date: 02 Nov 2006 13:55:36
From: Kenneth Sloan
Subject: Re: PROOF a beginner has no rating.
|
Oh dear...where to start?
David Kane wrote:
> ...
> As I said, the number who receive points that way is minimal. I think it is
> an odd practice, to say the least, but it's inflationary effect is very small.
> Some
> of those receiving points that way are the older players who have stopped
> playing games in the NWSRS, so it has no effect other than symbolic.
> The bump occurs once or twice per year and the number of points added is small.
OK
>
> On the other hand, people playing at their floor pump points into
> the system every tournament. Also, the initially imputed rating
> is in most cases higher than the USCF - it's changed over the years
> and the details have never been clear. But for most age groups its
> higher than 50*Age.
That can be seriously wrong for Seniors...
50*Age is probably too low for anyone aged lower than 20.
>
> Also, consider the "bonus point" mechanism for inflating the ratings.
> Imagine two identical pools of players, but players in pool A play
> twice as frequenctly as those in Pool B. Ratings in Pool A will inflate
> faster than those of Pool B. My guess is that our pool is more active
> than the USCF Pool.
You are assuming that the net effect of play is inflationary. That's
certainly not the usual starting point for most Elo-based systems.
Unadorned Elo is usually a deflating system. If the bonus point system
you are using is more inflationary than necessary to buck the natural
deflationary pressure, perhaps you need to fix it. In the current
USCF system, the bonus points threshold has been adjusted to be
mildly inflationary, by design. We have a political mandate to
re-inflate ratings to the 1997 standard.
>
> And ratings of players who play more frequently in the NWSRS will
> not have the lag that their USCF ratings have.
What lag? USCF events are now rated the same day they are played.
> There have
> been comic misjustices in tournaments where USCF ratings were used -
> e.g. one boy won the "biggest upset" prize every round even though
> in the NWSRS system he had the highest rating and was winning
> as expected. But his USCF rating was several years old, and
> 600 points out of date.
Probably because he was playing "rated" chess under a competing system
which reduced his incentive/opportunity to play in USCF-rated events.
It's a bit twisted to discourage USCF play...and then use USCF ratings.
>
> I don't believe trying to align the scales would have any practical
> advantages.
>
Depends on what you mean by "align the scales". Elo points out a need
for a national rating service to attempt to keep the rating scale stable
over time (so that historical comparisons are at least in the right
ballpark).
The same requirement holds for multiple systems at the same time. If
the scales overlap at all, it's nice if "2000" on one scale roughly
matches "2000" on the other scale.
Note that that is NOT the same as saying that every individual should
have his rating adjusted so it's the same in both systems. In fact,
that is the k of a bad ratings administrator! if a player is active
in both systems, and OTHER means are used to effectively align the
scales, then the player's two ratings should be reasonably close, and
there should be NO NEED to "adjust" one or the other.
Systems which *use* USCF ratings, but which do not submit their events
for USCF rating, are doomed to create these "comic injustices".
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | | | | | | | | | | | | | | | | | |
Date: 02 Nov 2006 13:04:02
From: David Kane
Subject: Re: PROOF a beginner has no rating.
|
"Kenneth Sloan" <[email protected] > wrote in message
news:[email protected]...
> Oh dear...where to start?
>
> David Kane wrote:
>
>> ...
>> As I said, the number who receive points that way is minimal. I think it is
>> an odd practice, to say the least, but it's inflationary effect is very
>> small. Some
>> of those receiving points that way are the older players who have stopped
>> playing games in the NWSRS, so it has no effect other than symbolic.
>> The bump occurs once or twice per year and the number of points added is
>> small.
>
> OK
>>
>> On the other hand, people playing at their floor pump points into
>> the system every tournament. Also, the initially imputed rating
>> is in most cases higher than the USCF - it's changed over the years
>> and the details have never been clear. But for most age groups its
>> higher than 50*Age.
> That can be seriously wrong for Seniors...
> 50*Age is probably too low for anyone aged lower than 20.
I thought that is what the USCF used (for kids). Again I haven't looked
into the details but it is a possible explanation for
different ratings.
>
>>
>> Also, consider the "bonus point" mechanism for inflating the ratings.
>> Imagine two identical pools of players, but players in pool A play
>> twice as frequenctly as those in Pool B. Ratings in Pool A will inflate
>> faster than those of Pool B. My guess is that our pool is more active
>> than the USCF Pool.
>
> You are assuming that the net effect of play is inflationary. That's
> certainly not the usual starting point for most Elo-based systems.
> Unadorned Elo is usually a deflating system. If the bonus point system
> you are using is more inflationary than necessary to buck the natural
> deflationary pressure, perhaps you need to fix it. In the current
> USCF system, the bonus points threshold has been adjusted to be
> mildly inflationary, by design. We have a political mandate to
> re-inflate ratings to the 1997 standard.
You completely miss the point. Simply, the bonus point
mechanism will inflate the ratings differently depend on the
activity level of the players. So it is a possible explanation
for why the scales do not align exactly.
Personally I don't think bonus points make sense for
this pool, but I don't see it doing significant harm.
>>
>> And ratings of players who play more frequently in the NWSRS will
>> not have the lag that their USCF ratings have.
>
> What lag? USCF events are now rated the same day they are played.
If people play one USCF event per year and 10 NWSRS events per year,
then their NWSRS rating will be up to date; their USCF rating will be
old. Given that players of this age improve, the NWSRS rating will be
higher.
The rating formulae do not factor in the time (though probably could
or should for kids) At these ratings, ratings are determined over ~10 games,
and the formulae makes no distinction whether they were played last week, or
3 years ago. But the predictive value of 3-year-old results is very low when
you are talking about a 9-year-old.
>> There have
>> been comic misjustices in tournaments where USCF ratings were used -
>> e.g. one boy won the "biggest upset" prize every round even though
>> in the NWSRS system he had the highest rating and was winning
>> as expected. But his USCF rating was several years old, and
>> 600 points out of date.
>
> Probably because he was playing "rated" chess under a competing system which
> reduced his incentive/opportunity to play in USCF-rated events.
> It's a bit twisted to discourage USCF play...and then use USCF ratings.
I agree that it would have been more intelligent to use ratings
from the more accurate NWSRS system for that award. But
that would have undercut the organizer's purpose to "sell" USCF ratings.
It's a free country. No one was coerced into playing that event,
and the organizer was allowed to do whatever he wanted.
>>
>> I don't believe trying to align the scales would have any practical
>> advantages.
>
> Depends on what you mean by "align the scales". Elo points out a need
> for a national rating service to attempt to keep the rating scale stable over
> time (so that historical comparisons are at least in the right ballpark).
>
> The same requirement holds for multiple systems at the same time. If the
> scales overlap at all, it's nice if "2000" on one scale roughly matches "2000"
> on the other scale.
>
> Note that that is NOT the same as saying that every individual should have his
> rating adjusted so it's the same in both systems. In fact, that is the k
> of a bad ratings administrator! if a player is active in both systems, and
> OTHER means are used to effectively align the scales, then the player's two
> ratings should be reasonably close, and there should be NO NEED to "adjust"
> one or the other.
>
> Systems which *use* USCF ratings, but which do not submit their events for
> USCF rating, are doomed to create these "comic injustices".
>
People who play in the NWSRS are ranked quite accurately
by it. If people want to know what their rating would be in the
USCF, they could look for USCF events and play in them.
Or they could look at ratings of players who play in both and
try to figure out a correction. You're starting to sound a bit
like those "innumerate" folks you deride. Do I have to remind
you that ratings are not an absolute scale?
BTW, do you also denigrate online ratings, ratings from other countries,
etc that are not under your control? Or is it the particular fact
that this one facilitates an active scholastic chess scene that has
you so upset?
|
| | | | | | | | | | | | | | | | | | | | |
Date: 02 Nov 2006 16:25:45
From: Kenneth Sloan
Subject: Re: PROOF a beginner has no rating.
|
David Kane wrote:
>
> You completely miss the point. Simply, the bonus point
> mechanism will inflate the ratings differently depend on the
> activity level of the players. So it is a possible explanation
> for why the scales do not align exactly.
How does this bonus point system work? It's surely not the same as
the USCF bonus point system. USCF bonus points do not reward
*play* - they reward *good performances*.
>
> Personally I don't think bonus points make sense for
> this pool, but I don't see it doing significant harm.
What is it about the pool that obviates the need for bonus points?
>
>>> And ratings of players who play more frequently in the NWSRS will
>>> not have the lag that their USCF ratings have.
>> What lag? USCF events are now rated the same day they are played.
>
> If people play one USCF event per year and 10 NWSRS events per year,
> then their NWSRS rating will be up to date; their USCF rating will be
> old. Given that players of this age improve, the NWSRS rating will be
> higher.
Theres a solution for that...
>
> The rating formulae do not factor in the time (though probably could
> or should for kids) At these ratings, ratings are determined over ~10 games,
> and the formulae makes no distinction whether they were played last week, or
> 3 years ago. But the predictive value of 3-year-old results is very low when
> you are talking about a 9-year-old.
do you have data to support that?
I mean...it *sounds* right - but not everything that sounds right
pans out.
>
>
>>> There have
>>> been comic misjustices in tournaments where USCF ratings were used -
>>> e.g. one boy won the "biggest upset" prize every round even though
>>> in the NWSRS system he had the highest rating and was winning
>>> as expected. But his USCF rating was several years old, and
>>> 600 points out of date.
>> Probably because he was playing "rated" chess under a competing system which
>> reduced his incentive/opportunity to play in USCF-rated events.
>> It's a bit twisted to discourage USCF play...and then use USCF ratings.
>
> I agree that it would have been more intelligent to use ratings
> from the more accurate NWSRS system for that award. But
> that would have undercut the organizer's purpose to "sell" USCF ratings.
> It's a free country. No one was coerced into playing that event,
> and the organizer was allowed to do whatever he wanted.
>
Perhaps the organizer means to reward players who have AVOIDED
USCF-rated play over the last year? There exist coaches who steer
their kids away from USCF-rated play during the year, so that they will
be radically underrated when they go to Nationals.
>
>>> I don't believe trying to align the scales would have any practical
>>> advantages.
>> Depends on what you mean by "align the scales". Elo points out a need
>> for a national rating service to attempt to keep the rating scale stable over
>> time (so that historical comparisons are at least in the right ballpark).
>>
>> The same requirement holds for multiple systems at the same time. If the
>> scales overlap at all, it's nice if "2000" on one scale roughly matches "2000"
>> on the other scale.
>>
>> Note that that is NOT the same as saying that every individual should have his
>> rating adjusted so it's the same in both systems. In fact, that is the k
>> of a bad ratings administrator! if a player is active in both systems, and
>> OTHER means are used to effectively align the scales, then the player's two
>> ratings should be reasonably close, and there should be NO NEED to "adjust"
>> one or the other.
>>
>> Systems which *use* USCF ratings, but which do not submit their events for
>> USCF rating, are doomed to create these "comic injustices".
>>
>
> People who play in the NWSRS are ranked quite accurately
> by it. If people want to know what their rating would be in the
> USCF, they could look for USCF events and play in them.
> Or they could look at ratings of players who play in both and
> try to figure out a correction. You're starting to sound a bit
> like those "innumerate" folks you deride. Do I have to remind
> you that ratings are not an absolute scale?
>
No, but perhaps I need to remind you to "read Elo". If you are running
an Elo-like system and report ratings on a scale of roughly 1000-3000,
then it is irresponsible (bordering on fraud - which is the usual
motivation in these cases) to *not* try to keep roughly aligned with the
FIDE scale.
If you want to avail yourself of the "the scale is arbitrary" argument,
then perhaps you should choose a different scale? Here, I'm repeating
the same argument when I opposed the transition in USCF from the old
Postal Rating scale to a new one which appears to be 'aligned' with the
USCF Regular OTB scale. The problem, of course, is that it is *not*
"aligned" at all.
> BTW, do you also denigrate online ratings, ratings from other countries,
> etc that are not under your control?
What rating system have i "denigrated"?
>Or is it the particular fact
> that this one facilitates an active scholastic chess scene that has
> you so upset?
>
Doesn't upset me at all.
If it's "for the children" it must be OK.
>
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | | | | | | | | | | | | | | | | | | | |
Date: 08 Nov 2006 04:04:05
From: David Kane
Subject: Re: PROOF a beginner has no rating.
|
"Kenneth Sloan" <[email protected] > wrote in message
news:[email protected]...
> David Kane wrote:
>
>>
>> You completely miss the point. Simply, the bonus point
>> mechanism will inflate the ratings differently depend on the
>> activity level of the players. So it is a possible explanation
>> for why the scales do not align exactly.
>
> How does this bonus point system work? It's surely not the same as
> the USCF bonus point system. USCF bonus points do not reward
> *play* - they reward *good performances*.
>
Here is how the USCF system works, according to the
Glickman document on the USCF website.
The bonus threshold BT = B*sqrt(m), where B = 10 and m= number
of games in a tournament. For m=5, BT=22.4. An increase greater
than this is defined as a "good performance".
The rating change as a result of a tournament is
2*K*(S-E)-BT if K*(S-E) >BT, and K*(S-E) otherwise.
S and E are the score and the expected score respectively.
The K constant a a function of rating is:
400 58.6
800 50.0
1200 39.9
1600 28.3
Consider a pool of accurately-rated non-improving
players with a rating of 400, playing in 5-game tournaments.
Winning 3, 4, or 5 games will result in
bonus points according to the USCF. The numbers (negl. draws):
S P BP E(BP)
3 10/32 6.9 2.16
4 5/32 65.5 10.23
5 1/32 124.1 3.88
In this scenario, bonus points are awarded
50% of the time, and the expected number
of BP that such a player will receive,
per tournament, is 16.2. [Note for a rating
of 800, this number is 11.4]
So, to first order, if we took a pool of 400
players, had them play two tournaments per month,
then in 6 months they'd have average
ratings of 594. If the same pool played one
tournament in 6 months, the average rating
would rise to 416.
The USCF bonus points *do* reward play.
|
| | | | | | | | | | | | | | | | | | | | | | |
Date: 09 Nov 2006 08:57:18
From: Kenneth Sloan
Subject: Re: PROOF a beginner has no rating.
|
David Kane wrote:
>
> Winning 3, 4, or 5 games will result in
> bonus points according to the USCF. The numbers (negl. draws):
> S P BP E(BP)
> 3 10/32 6.9 2.16
> 4 5/32 65.5 10.23
> 5 1/32 124.1 3.88
Don't any of these games ever end in draws?
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | | | | | | | | | | | | | | | | | | | | | |
Date: 09 Nov 2006 08:01:41
From: David Kane
Subject: Re: PROOF a beginner has no rating.
|
"Kenneth Sloan" <[email protected] > wrote in message
news:[email protected]...
> David Kane wrote:
>
>>
>> Winning 3, 4, or 5 games will result in
>> bonus points according to the USCF. The numbers (negl. draws):
>> S P BP E(BP)
>> 3 10/32 6.9 2.16
>> 4 5/32 65.5 10.23
>> 5 1/32 124.1 3.88
>
>
>
> Don't any of these games ever end in draws?
I stated my assumption that I was neglecting
draws. This is a very good approximation at this
level of course. Adding draws does not change
the point being made in any significant way.
I will leave that as homework for you.
Apart from nit-picking, do you have any
substantive comment?
> --
> Kenneth Sloan [email protected]
> Computer and Information Sciences +1-205-932-2213
> University of Alabama at Birmingham FAX +1-205-934-5473
> Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | | | | | | | | | | | | | | | | | | | | | | |
Date: 09 Nov 2006 10:16:15
From: Kenneth Sloan
Subject: Re: nOfMaze.dpo.uab.edu>
|
3199 wrote:
> "Kenneth Sloan" <[email protected]> wrote in message
> news:[email protected]...
>> David Kane wrote:
>>
>>> Winning 3, 4, or 5 games will result in
>>> bonus points according to the USCF. The numbers (negl. draws):
>>> S P BP E(BP)
>>> 3 10/32 6.9 2.16
>>> 4 5/32 65.5 10.23
>>> 5 1/32 124.1 3.88
>>
>>
>> Don't any of these games ever end in draws?
>
> I stated my assumption that I was neglecting
> draws. This is a very good approximation at this
> level of course. Adding draws does not change
> the point being made in any significant way.
> I will leave that as homework for you.
>
> Apart from nit-picking, do you have any
> substantive comment?
On these issues, nit-picking IS substantive comment.
we don't design rating systems for fantasy worlds where everyone is
correctly equally rated and yet all games are decisive, and where no
player every learns anything and improves by playing.
I'll leave that field of endeavor to you - you appear to be good at it.
I will note that you are using the constant '10' in your bonus
computations, which is clearly advertised as an intentional inflationary
factor.
I will also note that bonus points do not long stay with a player who
has not truly improved, so distortion is not a serious issue. The way
to test this, of course, is to take real game data for a year or 5 and
vary the parameters of the rating system until you find settings which
maximize the predictive power of the system. Have you tried that? We did.
It's also nice to constantly monitor the system, watching for either
inflationary or deflationary pressures to drive the ratings away from
their desired positions. Do you do that? We do.
Right now, we have a political mandate to make USCF ratings as close as
possible to the 1997 standard. At present, USCF ratings are mildly
deflated, when compared against that standard. We're using the modified
(10 instead of 16) bonus point threshold to slowly re-inflate the
system. if you have any clever ideas on how to do that in a less
disruptive and distorting fashion, I'm open to suggestion.
Oh...and last...anyone who speaks of rating points as 'rewards' is
immediately suspect in my eyes. Ratings are estimates of future
performance. They are not poker chips.
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | | | | | | | | | | | | | | | | | | | | | | | |
Date: 09 Nov 2006 09:32:40
From: David Kane
Subject: Bonus points etc.
|
"Kenneth Sloan" <[email protected] > wrote in message
news:[email protected]...
> 3199 wrote:
>> "Kenneth Sloan" <[email protected]> wrote in message
>> news:[email protected]...
>>> David Kane wrote:
>>>
>>>> Winning 3, 4, or 5 games will result in
>>>> bonus points according to the USCF. The numbers (negl. draws):
>>>> S P BP E(BP)
>>>> 3 10/32 6.9 2.16
>>>> 4 5/32 65.5 10.23
>>>> 5 1/32 124.1 3.88
>>>
>>>
>>> Don't any of these games ever end in draws?
>>
>> I stated my assumption that I was neglecting
>> draws. This is a very good approximation at this
>> level of course. Adding draws does not change
>> the point being made in any significant way.
>> I will leave that as homework for you.
>>
>> Apart from nit-picking, do you have any
>> substantive comment?
>
>
> On these issues, nit-picking IS substantive comment.
>
> we don't design rating systems for fantasy worlds where everyone is correctly
> equally rated and yet all games are decisive, and where no player every learns
> anything and improves by playing.
>
> I'll leave that field of endeavor to you - you appear to be good at it.
If I had put in more realistic assumptions, you would have had
even more nits to pick. The example illustrates very clearly that
bonus points reward play.
> I will note that you are using the constant '10' in your bonus computations,
> which is clearly advertised as an intentional inflationary factor.
aka "distortion"
>
> I will also note that bonus points do not long stay with a player who has not
> truly improved, so distortion is not a serious issue.
This is only partially true. At any given point in time, there will
be players with underserved bonus points. But the point is that
"frequency of play" is an important variable in a discussion of
the bonus point mechanism.
When I last looked at this, the WA rating pool had something like
6 times the per capita activity of the USCF. It's certainly a
possibility that frequency of play is very different for the pools.
The way
> to test this, of course, is to take real game data for a year or 5 and vary
> the parameters of the rating system until you find settings which maximize the
> predictive power of the system. Have you tried that? We did.
I can only hope that the "we" to which you refer included some
people who actually understand how the system works - not just
pompous folks with opinions as to how it *should* work.
But you've also admitted to finding settings that achieve a political
mandate - you can't have it both ways.
> It's also nice to constantly monitor the system, watching for either
> inflationary or deflationary pressures to drive the ratings away from their
> desired positions. Do you do that? We do.
>
> Right now, we have a political mandate to make USCF ratings as close as
> possible to the 1997 standard. At present, USCF ratings are mildly deflated,
> when compared against that standard. We're using the modified (10 instead of
> 16) bonus point threshold to slowly re-inflate the system. if you have any
> clever ideas on how to do that in a less disruptive and distorting fashion,
> I'm open to suggestion.
Yet before you said:
"When a player performs "surprisingly well", it is entirely proper to add
points to the system. This does not "distort" the player's rating."
Now you are talking about "a less disruptive and distorting
fashion".
Has learning occurred?
> Oh...and last...anyone who speaks of rating points as 'rewards' is immediately
> suspect in my eyes. Ratings are estimates of future performance. They are
> not poker chips.
>
Please quote where I said that ratings were rewards.
I think you will have to find another reason for
being suspicious in my case.
The USCF is not in the business of "estimating future
performance". It is selling a product. Where does the
mandate come for inflating to the 1997 standard come
from? It's to make the product more ketable to
adult buyers who don't want to see their numbers go
down.
|
| | | | | | | | | | | | | | | | | | | | | | | | | | |
>
>> I will note that you are using the constant '10' in your bonus computations,
>> which is clearly advertised as an intentional inflationary factor.
>
> aka "distortion"
>
Absolutely false. You clearly have a different definition of
"distortion" than I do.
>> I will also note that bonus points do not long stay with a player who has not
>> truly improved, so distortion is not a serious issue.
>
> This is only partially true. At any given point in time, there will
> be players with underserved bonus points.
There are no "deserved" or "undeserved" points. There are good
estimates of future performance and bad estimates of future performance.
Using language like "deserved" is the k of a ratings amateur.
But the point is that
> "frequency of play" is an important variable in a discussion of
> the bonus point mechanism.
>
> When I last looked at this, the WA rating pool had something like
> 6 times the per capita activity of the USCF. It's certainly a
> possibility that frequency of play is very different for the pools.
>
> The way
>> to test this, of course, is to take real game data for a year or 5 and vary
>> the parameters of the rating system until you find settings which maximize the
>> predictive power of the system. Have you tried that? We did.
>
> I can only hope that the "we" to which you refer included some
> people who actually understand how the system works - not just
> pompous folks with opinions as to how it *should* work.
Answer the question: have you simulated your claims? do you have data
to back up your speculations?
>
> But you've also admitted to finding settings that achieve a political
> mandate - you can't have it both ways.
Ah, but I can. The current system achieves the political goals withOUT
distortion. I'm sorry that I am unable to explain this more clearly to you.
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | | | | | | | | | | | | | | | | | | | |
Date: 02 Nov 2006 14:58:00
From: David Kane
Subject: Re: PROOF a beginner has no rating.
|
"Kenneth Sloan" <[email protected] > wrote in message
news:[email protected]...
> David Kane wrote:
>
>>
>> You completely miss the point. Simply, the bonus point
>> mechanism will inflate the ratings differently depend on the
>> activity level of the players. So it is a possible explanation
>> for why the scales do not align exactly.
>
> How does this bonus point system work? It's surely not the same as
> the USCF bonus point system. USCF bonus points do not reward
> *play* - they reward *good performances*.
Sigh. It uses the USCF formulae. But if there are two
identical pools, pool A and pool B, but pool A plays more
frequently than pool B, then pool A will get more bonus points.
Why? Because there will be more performances, and hence
more "good" performances.
>>
>> Personally I don't think bonus points make sense for
>> this pool, but I don't see it doing significant harm.
>
> What is it about the pool that obviates the need for bonus points?
It is a scholastic rating system with no old-timers to protect.
I have seen the following many times: play, get a good result,
get bonus points. Next week, play again, underperform
the new rating. If we looked at both tournaments as one
"performance" there would be no bonus. But with bonus
points applied to each week, players come out ahead of
where they would have been. Not only does this
have the "desired" effect of pumping points into the system
(what the USCF wants to keep grandpa happy) but it also *distorts*
the system for those playing actively. I am not saying it is a huge
or significant distortion, but a needless one.
>>
>>>> And ratings of players who play more frequently in the NWSRS will
>>>> not have the lag that their USCF ratings have.
>>> What lag? USCF events are now rated the same day they are played.
>>
>> If people play one USCF event per year and 10 NWSRS events per year,
>> then their NWSRS rating will be up to date; their USCF rating will be
>> old. Given that players of this age improve, the NWSRS rating will be
>> higher.
>
> Theres a solution for that...
>
>>
>> The rating formulae do not factor in the time (though probably could
>> or should for kids) At these ratings, ratings are determined over ~10 games,
>> and the formulae makes no distinction whether they were played last week, or
>> 3 years ago. But the predictive value of 3-year-old results is very low when
>> you are talking about a 9-year-old.
>
> do you have data to support that?
>
> I mean...it *sounds* right - but not everything that sounds right
> pans out.
I've already given anecdotal evidence.
>>
>>
>>>> There have
>>>> been comic misjustices in tournaments where USCF ratings were used -
>>>> e.g. one boy won the "biggest upset" prize every round even though
>>>> in the NWSRS system he had the highest rating and was winning
>>>> as expected. But his USCF rating was several years old, and
>>>> 600 points out of date.
>>> Probably because he was playing "rated" chess under a competing system which
>>> reduced his incentive/opportunity to play in USCF-rated events.
>>> It's a bit twisted to discourage USCF play...and then use USCF ratings.
>>
>> I agree that it would have been more intelligent to use ratings
>> from the more accurate NWSRS system for that award. But
>> that would have undercut the organizer's purpose to "sell" USCF ratings.
>> It's a free country. No one was coerced into playing that event,
>> and the organizer was allowed to do whatever he wanted.
>>
>
>
> Perhaps the organizer means to reward players who have AVOIDED USCF-rated play
> over the last year? There exist coaches who steer
> their kids away from USCF-rated play during the year, so that they will
> be radically underrated when they go to Nationals.
Not at all. This is a well-meaning director trying to encourage
USCF play.
>>
>>>> I don't believe trying to align the scales would have any practical
>>>> advantages.
>>> Depends on what you mean by "align the scales". Elo points out a need
>>> for a national rating service to attempt to keep the rating scale stable
>>> over time (so that historical comparisons are at least in the right
>>> ballpark).
>>>
>>> The same requirement holds for multiple systems at the same time. If the
>>> scales overlap at all, it's nice if "2000" on one scale roughly matches
>>> "2000" on the other scale.
>>>
>>> Note that that is NOT the same as saying that every individual should have
>>> his rating adjusted so it's the same in both systems. In fact, that is the
>>> k of a bad ratings administrator! if a player is active in both systems,
>>> and OTHER means are used to effectively align the scales, then the player's
>>> two ratings should be reasonably close, and there should be NO NEED to
>>> "adjust" one or the other.
>>>
>>> Systems which *use* USCF ratings, but which do not submit their events for
>>> USCF rating, are doomed to create these "comic injustices".
>>>
>>
>> People who play in the NWSRS are ranked quite accurately
>> by it. If people want to know what their rating would be in the
>> USCF, they could look for USCF events and play in them.
>> Or they could look at ratings of players who play in both and
>> try to figure out a correction. You're starting to sound a bit
>> like those "innumerate" folks you deride. Do I have to remind
>> you that ratings are not an absolute scale?
>>
>
> No, but perhaps I need to remind you to "read Elo". If you are running
> an Elo-like system and report ratings on a scale of roughly 1000-3000, then it
> is irresponsible (bordering on fraud - which is the usual motivation in these
> cases) to *not* try to keep roughly aligned with the
> FIDE scale.
No one has ever pretended that these are USCF
or FIDE ratings. They are clearly ked appropriately
on each and every tournament flyer. The vast majority of readers
of these flyers will never have heard of the USCF or
FIDE, of course. To them, it's just their rating.
Your accusation of fraud is unfounded and desperate. I think
that most would consider the systems "roughly" aligned by the way.
The NWSRS is probably much closer to USCF than most
online ratings, e.g.
> If you want to avail yourself of the "the scale is arbitrary" argument,
> then perhaps you should choose a different scale? Here, I'm repeating the
> same argument when I opposed the transition in USCF from the old Postal Rating
> scale to a new one which appears to be 'aligned' with the
> USCF Regular OTB scale. The problem, of course, is that it is *not*
> "aligned" at all.
>
|
| | | | | | | | | | | | | | | | | | | | | | |
Date: 02 Nov 2006 17:08:40
From: Kenneth Sloan
Subject: Re: PROOF a beginner has no rating.
|
David Kane wrote:
> "Kenneth Sloan" <[email protected]> wrote in message
> news:[email protected]...
>> David Kane wrote:
>>
>>> You completely miss the point. Simply, the bonus point
>>> mechanism will inflate the ratings differently depend on the
>>> activity level of the players. So it is a possible explanation
>>> for why the scales do not align exactly.
>> How does this bonus point system work? It's surely not the same as
>> the USCF bonus point system. USCF bonus points do not reward
>> *play* - they reward *good performances*.
>
> Sigh. It uses the USCF formulae. But if there are two
> identical pools, pool A and pool B, but pool A plays more
> frequently than pool B, then pool A will get more bonus points.
>
> Why? Because there will be more performances, and hence
> more "good" performances.
>
Oh, I see...we've moved on to THIRD order effects.
>
>
>
>>> Personally I don't think bonus points make sense for
>>> this pool, but I don't see it doing significant harm.
>> What is it about the pool that obviates the need for bonus points?
>
> It is a scholastic rating system with no old-timers to protect.
Now that's just silly.
>
> I have seen the following many times: play, get a good result,
> get bonus points. Next week, play again, underperform
> the new rating. If we looked at both tournaments as one
> "performance" there would be no bonus. But with bonus
> points applied to each week, players come out ahead of
> where they would have been. Not only does this
> have the "desired" effect of pumping points into the system
> (what the USCF wants to keep grandpa happy) but it also *distorts*
> the system for those playing actively. I am not saying it is a huge
> or significant distortion, but a needless one.
and, have you measured the entire system to observe the effects with,
and without, the bonus point system?
Did you *like* it when 25% of all players were at their 1xx floors?
>>
>> I mean...it *sounds* right - but not everything that sounds right
>> pans out.
>
> I've already given anecdotal evidence.
there was a time when anecdotal evidence was good enough. No longer.
> Not at all. This is a well-meaning director trying to encourage
> USCF play.
>
So, he's not evil, just incompetent? How does giving "biggest upset"
prizes in this event *encourage* USCF play. sound to me like he's
created a strong DISincentive to play in (other) USCF-rated events.
> No one has ever pretended that these are USCF
> or FIDE ratings. They are clearly ked appropriately
> on each and every tournament flyer. The vast majority of readers
> of these flyers will never have heard of the USCF or
> FIDE, of course. To them, it's just their rating.
>
> Your accusation of fraud is unfounded and desperate. I think
> that most would consider the systems "roughly" aligned by the way.
does anyone consider that a "goal"? does anyone do measurement and
tweaking to achieve the goal?
If USCF ratings are irrelevant, why "bump up" ratings based on USCF ratings?
> The NWSRS is probably much closer to USCF than most
> online ratings, e.g.
damning with faint praise.
>
>
>> If you want to avail yourself of the "the scale is arbitrary" argument,
>> then perhaps you should choose a different scale? Here, I'm repeating the
>> same argument when I opposed the transition in USCF from the old Postal Rating
>> scale to a new one which appears to be 'aligned' with the
>> USCF Regular OTB scale. The problem, of course, is that it is *not*
>> "aligned" at all.
>>
You forgot to reply to this point.
>
>
>
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | | | | | | | | | | | | | | | | | | | | | |
Date: 02 Nov 2006 17:53:06
From: David Kane
Subject: Re: PROOF a beginner has no rating.
|
"Kenneth Sloan" <[email protected] > wrote in message
news:[email protected]...
> David Kane wrote:
>> "Kenneth Sloan" <[email protected]> wrote in message
>> news:[email protected]...
>>> David Kane wrote:
>>>
>>>> You completely miss the point. Simply, the bonus point
>>>> mechanism will inflate the ratings differently depend on the
>>>> activity level of the players. So it is a possible explanation
>>>> for why the scales do not align exactly.
>>> How does this bonus point system work? It's surely not the same as
>>> the USCF bonus point system. USCF bonus points do not reward
>>> *play* - they reward *good performances*.
>>
>> Sigh. It uses the USCF formulae. But if there are two
>> identical pools, pool A and pool B, but pool A plays more
>> frequently than pool B, then pool A will get more bonus points.
>>
>> Why? Because there will be more performances, and hence
>> more "good" performances.
>>
>
> Oh, I see...we've moved on to THIRD order effects.
The amount of bonus point put into the system is
directly proportional to the number of performances.
Moreover, the effect that it is counterbalancing, improving
players being systematically underated and taking points from
the system, is less of a factor if players play more frequently.
So in the USCF you have x-y and believe x and y to be
roughly equal. We could have 5*x - 0.3*y, so the effect
of using the same formulae could be very different in our pool.
>>
>>
>>
>>>> Personally I don't think bonus points make sense for
>>>> this pool, but I don't see it doing significant harm.
>>> What is it about the pool that obviates the need for bonus points?
>>
>> It is a scholastic rating system with no old-timers to protect.
>
> Now that's just silly.
*You* said that the reason the USCF uses bonus points is
to maintain the ratings of mature players. Did you worry about
the rating distortion that was produced?
>
>>
>> I have seen the following many times: play, get a good result,
>> get bonus points. Next week, play again, underperform
>> the new rating. If we looked at both tournaments as one
>> "performance" there would be no bonus. But with bonus
>> points applied to each week, players come out ahead of
>> where they would have been. Not only does this
>> have the "desired" effect of pumping points into the system
>> (what the USCF wants to keep grandpa happy) but it also *distorts*
>> the system for those playing actively. I am not saying it is a huge
>> or significant distortion, but a needless one.
>
> and, have you measured the entire system to observe the effects with, and
> without, the bonus point system?
>
No.Haven't measured.
> Did you *like* it when 25% of all players were at their 1xx floors?
Don't care much one way or the other. It bothers me slightly that
some active players don't get to see their ratings change. Otherwise
it seems pretty harmless, and is just as good a way to inflate as
any.
>>>
>>> I mean...it *sounds* right - but not everything that sounds right
>>> pans out.
>>
>> I've already given anecdotal evidence.
>
>
> there was a time when anecdotal evidence was good enough. No longer.
>
>> Not at all. This is a well-meaning director trying to encourage
>> USCF play.
>>
>
> So, he's not evil, just incompetent? How does giving "biggest upset" prizes
> in this event *encourage* USCF play. sound to me like he's created a strong
> DISincentive to play in (other) USCF-rated events.
>
First, I doubt the "upset" prizes were an incentive to enter or
not enter any event. His event undoubtedly encourages USCF play, because
it is one of a very small number of USCF-rated events for
scholastic players.
>
>> No one has ever pretended that these are USCF
>> or FIDE ratings. They are clearly ked appropriately
>> on each and every tournament flyer. The vast majority of readers
>> of these flyers will never have heard of the USCF or
>> FIDE, of course. To them, it's just their rating.
>>
>> Your accusation of fraud is unfounded and desperate. I think
>> that most would consider the systems "roughly" aligned by the way.
>
> does anyone consider that a "goal"? does anyone do measurement and tweaking
> to achieve the goal?
Not as far as I know. I would certainly hope that nobody wastes time
on such a useless task.
> If USCF ratings are irrelevant, why "bump up" ratings based on USCF ratings?
>
This is basically to make USCF players feel better. I doubt it's worth
the effort.
>> The NWSRS is probably much closer to USCF than most
>> online ratings, e.g.
>
> damning with faint praise.
>
Believe it or not, not everybody considers being like the USCF
a good thing.
>>
>>
>>> If you want to avail yourself of the "the scale is arbitrary" argument,
>>> then perhaps you should choose a different scale? Here, I'm repeating the
>>> same argument when I opposed the transition in USCF from the old Postal
>>> Rating scale to a new one which appears to be 'aligned' with the
>>> USCF Regular OTB scale. The problem, of course, is that it is *not*
>>> "aligned" at all.
>>>
>
>
> You forgot to reply to this point.
>>
>>
>>
I'm not sure how to reply. In the system being discussed,
we've already established that the systems are roughly
aligned. I'm not aware of your history with the
great postal rating scandal, so have no comment on it.
|
| | | | | | | | | | | | | | | | | | | | | | | | |
Date: 02 Nov 2006 21:47:32
From: Kenneth Sloan
Subject: Re: nOfMaze.dpo.uab.edu>
|
7572 wrote:
> *You* said that the reason the USCF uses bonus points is
> to maintain the ratings of mature players.
Please show me a quote, or retract this.
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | | | | | | | | | | | | | | | | | | | | | | | |
Date: 02 Nov 2006 20:13:19
From: David Kane
Subject: Re: nOfMaze.dpo.uab.edu>
|
"Kenneth Sloan" <[email protected] > wrote in message
news:[email protected]...
> 7572 wrote:
>
>> *You* said that the reason the USCF uses bonus points is
>> to maintain the ratings of mature players.
>
> Please show me a quote, or retract this.
The quote I was thinking of was:
"Right now, the Ratings Committee has one "design handle" to manipulate
inflation/deflation. That is the "bonus threshold". Currently, it is
set to produce mild inflation. RealSoonNow, it will be re-set to the
neutral position."
The context of the discussion was that there was a perception
that rating deflation had caused ratings to decline.
But my recall was imperfect, and I apologize for that.
The corrected statement will be "*You* said that the USCF
uses bonus points to manipulate inflation/deflation."
I don't see that it has any bearing on my argument, but
there is no harm in being accurate. The point being that
the positive side effect (controlling inflation) has other
consequences (distorting the rating of those who receive
them). Mind you, human nature being what it is, few are
inclined to complain about getting them. But it does have
relevance to the motives and values of those involved in
twiddling with the ratings.
|
| | | | | | | | | | | | | | | | | | | | | | | | | | |
Date: 02 Nov 2006 23:23:01
From: Kenneth Sloan
Subject: Re: .uab.edu>
|
3586 wrote:
> "Kenneth Sloan" <[email protected]> wrote in message
> news:[email protected]...
>> 7572 wrote:
>>
>>> *You* said that the reason the USCF uses bonus points is
>>> to maintain the ratings of mature players.
>> Please show me a quote, or retract this.
>
> The quote I was thinking of was:
>
> "Right now, the Ratings Committee has one "design handle" to manipulate
> inflation/deflation. That is the "bonus threshold". Currently, it is
> set to produce mild inflation. RealSoonNow, it will be re-set to the
> neutral position."
>
> The context of the discussion was that there was a perception
> that rating deflation had caused ratings to decline.
>
> But my recall was imperfect, and I apologize for that.
> The corrected statement will be "*You* said that the USCF
> uses bonus points to manipulate inflation/deflation."
>
> I don't see that it has any bearing on my argument, but
> there is no harm in being accurate. The point being that
> the positive side effect (controlling inflation) has other
> consequences (distorting the rating of those who receive
> them).
This is false.
Ratings are estimates of future performance. When a player
performs "surprisingly well", it is entirely proper to add
points to the system. This does not "distort" the player's rating. On
the contrary, it makes the rating *more* accurate than if the
bonus point mechanism were *not* used.
> Mind you, human nature being what it is, few are
> inclined to complain about getting them. But it does have
> relevance to the motives and values of those involved in
> twiddling with the ratings.
>
The Motivation of the USCF Rating Committee is to provide
accurate ratings. Bonus points do that.
You do your technical arguments no favor when you try
to score points by impugning other people's integrity.
Simply put - you don't know what you are talking about.
>
>
>
--
Kenneth Sloan [email protected]
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/
|
| | | | | | | | | | | | | | | | | | | | | | | | | | | |
Date: 02 Nov 2006 23:09:39
From: David Kane
Subject: Re: .uab.edu>
|
"Kenneth Sloan" <[email protected] > wrote in message
news:[email protected]...
> 3586 wrote:
>> "Kenneth Sloan" <[email protected]> wrote in message
>> news:[email protected]...
>>> 7572 wrote:
>>>
>>>> *You* said that the reason the USCF uses bonus points is
>>>> to maintain the ratings of mature players.
>>> Please show me a quote, or retract this.
>>
>> The quote I was thinking of was:
>>
>> "Right now, the Ratings Committee has one "design handle" to manipulate
>> inflation/deflation. That is the "bonus threshold". Currently, it is
>> set to produce mild inflation. RealSoonNow, it will be re-set to the
>> neutral position."
>>
>> The context of the discussion was that there was a perception
>> that rating deflation had caused ratings to decline.
>>
>> But my recall was imperfect, and I apologize for that.
>> The corrected statement will be "*You* said that the USCF
>> uses bonus points to manipulate inflation/deflation."
>>
>> I don't see that it has any bearing on my argument, but
>> there is no harm in being accurate. The point being that
>> the positive side effect (controlling inflation) has other
>> consequences (distorting the rating of those who receive
>> them).
>
>
> This is false.
>
> Ratings are estimates of future performance. When a player
> performs "surprisingly well", it is entirely proper to add
> points to the system. This does not "distort" the player's rating. On the
> contrary, it makes the rating *more* accurate than if the
> bonus point mechanism were *not* used.
>> Mind you, human nature being what it is, few are
>> inclined to complain about getting them. But it does have
>> relevance to the motives and values of those involved in
>> twiddling with the ratings.
>>
>
> The Motivation of the USCF Rating Committee is to provide
> accurate ratings. Bonus points do that.
Now you are just contradicting yourself. You've also said
that you "control" inflation. That's not necessary for accuracy.
>
> You do your technical arguments no favor when you try
> to score points by impugning other people's integrity.
>
> Simply put - you don't know what you are talking about.
>
You are the one making simplistic arguments. There could be
many ways to inflate the ratings. You've said that you
adjust the bonus to control inflation. It doesn't sound like
you are modeling "surprisingly good" performances if the
proper way to model it depends on completely unrelated
factors such as inflationary pressures on the rating pool.
I understand that, in the real world, rating algorithms
are designed for things besides accurately predicting
future performances. And in such a world, a rating
floor of 400 is perfectly defensible. As are the USCF's
tweaks.
I don't doubt your integrity in the slightest. Your
technical competence seems dubious because
you don't seem to grasp that features of the
pool (such as how frequently people play) can have
an impact on how the various tweaks work. I suspect
that this is not an inability to grasp the theory of
such points, but more likely a blindness of your
"average adults just having learned the rules play
ELO 1000" variety.
|
| | | | | | | | | | | | | | | | | | | | | | | | | | | | |
Date: 03 Nov 2006 13:32:07
From: alexmagnus
Subject: Re: PROOF a beginner has no rating.
|
In my family, nobody is rated but I would give myself 1400 and m
brother about 1000. I started playing chess 2 years ago and at tha
time i got beaten by my brother in every single game - no matter ho
many games we played.Already after 6-7 moves I was a knight and a roo
down. I would rate myself at that time with something like 30
--
alexmagnus
|
|
