USCF sells out again
ABlue892
"I recently had what for me was a very successful tournament (even score in an
Open section), and was dismayed to find that I was only scheduled to receive
about two thirds of the points I normally would have. It was explained to me
that this is a result of the new "improved" rating system.
I had just started to make some real headway towards my goal of an Expert
rating only to find out now that the USCF has intentionally made this twice as
difficult, while at the same time attempting to artificially inflate the
ratings of weak players. This discovery was extremely discouraging and has
significantly dampened my desire to play tournament chess. This is compounded
by the fact that the formula for determining rating points is now so complex
that I cannot even enjoy, as I and many others have in the past, calculating
potential rating gains "on the fly" during a tournament.
It seems you are trying to discourage A players such as myself, as well as
perhaps B players and experts, from playing tournament chess. I am curious why
this is. Are you trying to make the USCF a more scholastic league, or a league
primarily of casual players and Masters, with no in-between? If so, couldn't
you consider separate scholastic and master leagues and retain a league for
competitive tournament players, rather than force them out? Was this the only
alternative for some reason? Just curious."
. I doubt I will receive any response. What I suspect has happened is
that the heavily debt-ridden USCF has realized that there is more money in
bringing in lots of patzers and kids and getting those dues, then in making
their existing players happy. Thus, they choose to artificially inflate the
ratings of new players who have no chess skill but lots of cash, in order to
keep them coming in, and hold down the ratings of competitive players who are
trying to improve such as myself, as this does not bring in any significant $$
for the organization. Presumably the decision to make the formula one which
requires a degree in calculus to decipher is a way they hope to sort of sneak
this decision by, since they know we can't do anything to stop them from
screwing us, but would like to keep us quiet. This is just one of many
examples of the USCF selling out, including making the September issue of Chess
Life a walking advertisement for Garry Kasparov, who has been not nearly the
Saint of chess they would have you believe, and using young superstar players
to promote their internet chess site without those players' knowledge or
permission. I urge one of you out there, perhaps a wealthy dotcom-er who is
one of the A, B or Expert players being screwed to line the USCF coffers, try
to put together a parallel Chess Federation that considers the needs of its
competitve players, and can run itself well enough not to have to sell out to
anyone who can wave 5 bucks in front of its face. At the very least, send some
letters to the USCF letting them know you don't appreciate being pushed around.
The Seattle Chess Federation, which now sponsors the US Championships, is no
better, as they are already suggesting discriminatory policies towards female
players, by 1) attempting to abolish the Womens Chess Championship, and 2)
failing that, sabotaging it by scheduling it during the first two weeks of
school, when the most talented young women, who are still in college, will not
be able to compete.
SPF
Correspondence Chess, the ICCF gets you in contact with lots of interesting people
all over the world. They also do have a "what if..." ratings calculator on their
web site, if you like that sort of thing. The USCF could at least implement that
simple bit of technology, maybe?
ABlue892 wrote:
--
Dr Simon Fitzpatrick, Senior Lecturer in Mathematics,
The University of Western Australia, Crawley WA 6009
Telephone +61 8 9380 3389, Facsimile +61 8 9380 1028.
Sam Sloan
I am wondering why you did not post this to rec.games.chess.politics
as it really belongs there.
The scheme to abolish the rating system in favor of fiddle points is
the brain-child of Tim Redman, who is rapidly becoming the most hated
man in chess.
Did you read the letter to the editor on page 6 collumn two of the
December Chess Life entitled "Concerned Committee"?
Everyone says don't worry about it because Redman's term expires in
August and he says he is not running again (he better not) and
thereafter sweetness and light will prevail throughout the land.
Sam Sloan
dcle...@my-deja.com
ratings. I think the overall goal of each player should be to improve
his play. If you improve your play to the Expert level then it should
be only a matter of time before you reach your goal.
I have been aware of the the "new player/ratings deflation" problem for
some time. After a discussion with my friend Carl Dunn, he sent a
letter to the USCF about the problem. They assured us that major
changes were being worked on.
My viewpoint is that new players come in with low ratings. They usually
work on their game and improve dramatically. Consequently, when they
are no longer provisional, the points they gain are taken from existing
players.
I guess every system has its faults. They correct one problem and
create another. Interesting....
Sent via Deja.com
http://www.deja.com/
ABlue892
>be only a matter of time before you reach your goal.
But the reality is one must play HIGHER than the expert level to reach this
goal. If one assumes that an expert loses to other experts half of the time,
the rating gain for an A player trying to reach expert level is only the number
of points per gain represented by the difference between their rating and that
level, 100 or 200 points. This difference used to be between 4-8 points, and
now it seems like even less, meaning the matter of time you refer to could be
hundreds of chess games, which is frustrating. Saying "one should want to
improve their play, rather than their rating" is meaningless, because so many
of USCF members are self-deluded egotists who rate their "Chess strength" as
200 points higher than their Elo even though this is usually ridiculous, and
the only reasonable measure of someone's actual "strength" is their rating.
ABlue892
>as it really belongs there.
>
I dunno, I thought Id try to liven up this newsgroup a bit : ) . Besides
don't tell me all the "FIDE sucks" stuff doesn't belong in RGCP also : )
Matt Nemmers
"I am wondering why you did not post this to rec.games.chess.politics as it
really belongs there..."
This is r.g.c.miscellaneous, Sam. Anything chess related is fair game.
Besides, it's debatable whether or not most of YOUR posts belong here.
ske...@my-deja.com
(Sam Sloan) wrote:
> On 18 Dec 2000 05:04:57 GMT, ablu...@aol.com (ABlue892) wrote:
>
> >What follows in the quotes is a copy of my recent letter to
USCF/Chess Life:
> >well as perhaps B players and experts, from playing tournament chess.
> >I am curious why this is. Are you trying to make the USCF a more
> >scholastic league, or a league primarily of casual players and
> >Masters, with no in-between? If so, couldn't you consider separate
> >scholastic and master leagues and retain a league for competitive
> >tournament players, rather than force them out? Was this the only
> >alternative for some reason? Just curious."
I don't know. I think a separate SCHOLASTIC rating system would make
sense, as it would prevent abuse such as seen in the "twenty game match"
incident, seen somewhere else in a thread.
> > I doubt I will receive any response. What I suspect has happened is
> >that the heavily debt-ridden USCF has realized that there is more
> >money in bringing in lots of patzers and kids and getting those dues,
> >then in making their existing players happy.
Speculation. USCF needs all its chess players.
> >I urge one of you out there, perhaps a wealthy dotcom-er who is
> >one of the A, B or Expert players being screwed to line the USCF
> >coffers, try to put together a parallel Chess Federation that
> >considers the needs of its competitve players, and can run itself
> >well enough not to have to sell out to anyone who can wave 5 bucks
> >in front of its face.
We could hope.
> > The Seattle Chess Federation, which now sponsors the US
> > Championships, is no better, as they are already suggesting
> > discriminatory policies towards female players, by 1) attempting
> > to abolish the Womens Chess Championship, and 2) failing that,
> > sabotaging it by scheduling it during the first two weeks of
> > school, when the most talented young women, who are still in
> > college, will not be able to compete.
Can you verify that?
> I am wondering why you did not post this to rec.games.chess.politics
> as it really belongs there.
Actually, it's one of those rare posts that belong in *both* forums.
> The scheme to abolish the rating system in favor of fiddle points is
> the brain-child of Tim Redman, who is rapidly becoming the most hated
> man in chess.
I dunno, Sam; I still like Tim. I may not agree with everything, but
he's doing *something*.
> Did you read the letter to the editor on page 6 collumn two of the
> December Chess Life entitled "Concerned Committee"?
Goodness no. I haven't had the time to open the mag yet.
> Everyone says don't worry about it because Redman's term expires in
> August and he says he is not running again (he better not) and
> thereafter sweetness and light will prevail throughout the land.
Oh, I hope not. At least, not the Sweetness and Light of the Schultz
years.
SK
ABlue892
>sense, as it would prevent abuse such as seen in the "twenty game match"
>incident, seen somewhere else in a thread.
AGREED AGREED AGREED. This is the obvious solution to many of the USCFs
problems, a separate league/system for scholastic players, and even perhaps an
internet chess club for kids...I guess it takes a non chess professional to
figure that out.
>Speculation. USCF needs all its chess players.
One would think....
>> > The Seattle Chess Federation, which now sponsors the US
>> > Championships, is no better
>Can you verify that?
My information comes from one of this year's competitors, based on what she
heard at the SCF's board meeting during the US Championships.
Eric Mark
effect on Jan. 1st, if the computer system cooperates.
For those unfamiliar with this proposal, any player who enters a
USCF-rated tournament rated under 2000, and who plays at least three
games in that event, will be awarded two "activity points" for each game
played.
Actually, this means that any A-player with the available time and money
can all but guarantee that he will attain an Expert rating some time
next year, just by remaining very active, and avoiding total meltdown in
his/her results.
(Once you pass 2000, the fiddling ceases to be heard.)
That's why I'm not a fan of fiddling with ratings in this manner.
In my case, I find that I can maintain 1950 'relatively' easily, but
2000 is much tougher. I don't like the idea that next year, I'm almost
certainly going to top 2000 again just by showing up. I'd rather earn
it.
As for the new rating system, there seems to be no clear agreement as to
when it will take effect. OR whether or not it already has done so....
The info on the USCF Ratings webpage has been somewhat less than
transparent, and USCF President Tim Redman has stated on RGCP that he
expects both the new system and the activity points to take effect on
Jan. 1st. Don't bet too much on any time-frame involvng the USCF
computer system, though.
As to the rating result in question here, are you sure that the event
was rated under the new system? Have you checked the tournament
crosstable, or your updated rating at the USCF webpage?
Please consider that the new system features bonus points, to reward
players who achieve much higher than expected TPR's. The apparent
absence of such bonus points in this case makes me think that this event
was not rated under the new system, and that there may have been some
other snafu in the rating process.
If it was in fact rated the "new way," that should at least answer the
question of which system is currently being used.
I hope....
As to the frustration over the reduced rating gain, I feel your pain,
but think about this: had you achieved a very poor result in this event,
your net rating loss would also have been lessened.
If I understand aright the new sliding-K factor concept, (a big "if"),
an active, established adult A-player with moderately stable results
will achieve very nearly the same net rating gain or loss over the
course of say, a year or so, under both systems. (Corrections from the
better-informed gladly accepted.)
Which is probably not going to be much consolation to ANY player the
first time he/she sees a lower-than-expected rating boost from a good
result.
But I hope Chess Life prints this letter, so this issue can be brought
to more players' attention sooner rather than later.
Best,
Eric M.
Chesspride
>Actually, this means that any A-player with the available time and money
>can all but guarantee that he will attain an Expert rating some time
>next year, just by remaining very active, and avoiding total meltdown in
Actually, what it means is that any current 1950 should have been around 2020
if the system were not deflated.
It's a correction, Eric...
Eric C. Johnson
Chesspride
>If I understand aright the new sliding-K factor concept, (a big "if"),
>an active, established adult A-player with moderately stable results
>will achieve very nearly the same net rating gain or loss over the
>course of say, a year or so, under
Only if he plays more games/is more active....K is going down...so the
variability is going down.
That fellow rated 1925 might see his range shrink from 1890 to 1960, instead of
1840 to 2010. Less variability, but less chance that a person on the border of
a class will be reinforced by temporary forays into the next class.
Eric C. Johnson
Włodzimierz Holsztyński
> [...]
>
> I am wondering why you did not post
> this to rec.games.chess.politics
> as it really belongs there.
>
> [...]
>
> Sam Sloan
Detailed issues of a chess organization, its
personal issues, etc. may often belong to r.g.c.politics.
But broad issues certainly belong here, to rgc.misc, too.
I am thinkig about a series of 3 posts:
1. No awards for mediocrity'
2. Decentralised, selforganized global world chess;
3. Implementation of the decentralised model.
The first post will hacve a very narrow scope but
illustrates ethical, principal thinking which should
guide us in general, looking at things from a distance.
The second will show how easy today it is to have
a very healthy, organized chess activity on the global
scale. The third one will be an actual suggestion
plus basic details.
I think, and the tradition of this list supports me,
that these three posts will be proper for rgc.misc
Regards,
Wlod
Eric Mark
(referring to the new USCF rating system)
"That fellow rated 1925 might see his range shrink from 1890 to
1960, instead of 1840 to 2010. Less variability, but less chance that a
person on the border of a class will be reinforced by temporary forays
into the next class."
Good point, Eric, but remember the activity points, if/when they are
implemented.
"That fellow" will get to 2000, if he only plays a whole lot of events.
When we see a sample of results and rating ranges based on the new
system sans activity points, and if the rating ranges are 'much'
narrower than they were, or "should be," you will be proven right, and
we can have another crack at a system fix.
Players currently floored at or above 2000, of course, won't see any
direct "benefit" from the fiddling, but I suspect the effect wil trickle
up.
For instance, suppose you and I play sometime late next year, and I'm
rated 2025 rather than 1970, due to being handed fiddle points I didn't
really want.
That's an extra coupla points rating gain for you, when you crush my
unsound Budapest, and if I drop back to 1990 after that event, all I
need do is play a few more games, avoid any major upsets, and, voila,
I'm an expert again.
What a sense of accomplishment....
Best,
Eric M.
Chesspride
>For instance, suppose you and I play sometime late next year, and I'm
>rated 2025 rather than 1970, due to being handed fiddle points I didn't
>really want.
>
Ah, but it is a measurement scaling issue, Eric...not an issue of "earning" or
"wanting"...
..for if the system had not been deflating for 10 yrs...your 1970 rating would
BE 2025.
It's a correction...view the activity points as a correction.
The rating committee had the chance to endorse a one-shot point correction, and
didn't do it. Now they are stuck with activity points.
So long as the duration of the promotion is short, it amounts to pretty much
the same thing.
ECJ
Eric Mark
"The rating committee had the chance to endorse a one-shot point
correction, and didn't do it. Now they are stuck with activity points.
So long as the duration of the promotion is short, it amounts to pretty
much the same thing."
Not to belabor this point forever, but there will be players next year
who gain HUNDREDS of actvity points.
Even the most extreme advocates of a one-time correction didn't support
such an extreme ratings boost.
Hey, I really hope it works, I just don't like it....
But the original poster was not talking about the relative merits of
activity points.
Does anyone here know for sure which rating system,-"old" or "new"-, is
currently being used to rate USCF events?
Best,
Eric M
ABlue892
>can all but guarantee that he will attain an Expert rating some time
>next year, just by remaining very active, and avoiding total meltdown in
>his/her results.
>
Considerable issues of available time and money aside, I happen to think trying
to cram in 300 games a year is just not conducive to good chess. Chess
requires a clear and relaxed mind, and I know from experience that playing 4-8
games a week simply deteriorates the quality of my play.
>In my case, I find that I can maintain 1950 'relatively' easily, but
>2000 is much tougher. I don't like the idea that next year, I'm almost
>certainly going to top 2000 again just by showing up. I'd rather earn
>it.
Well if the amount of points I gain under the new system is cut by a third, at
best it seems the fiddle points are helping me break even, while still keeping
the problem of artificially inflating the total fish. (correction my ass, some
people just can't face the fact that they are just not good. I have never met
a person U1800 who didnt think they were underrated by at least 200 points, and
if EVERYONE thinks that way, the system works, it's the players egos that don't
work. THIS is the correction the USCF is catering to)
>As to the rating result in question here, are you sure that the event
>was rated under the new system? Have you checked the tournament
>crosstable, or your updated rating at the USCF webpage?
The tournament website rating changes list clearly says across the top
"According to the NEW USCF rating system" Whether they just did this to
delight and amuse the players and in fact the tournament is being rated under
the normal system, I do not know since the USCF hasn't gotten around to rating
it yet.
>Please consider that the new system features bonus points, to reward
>players who achieve much higher than expected TPR's. The apparent
>absence of such bonus points in this case makes me think that this event
>was not rated under the new system, and that there may have been some
>other snafu in the rating process.
>
I have been trying to figure out if I was entitled to bonus points or not.
Unfortunately I do not have a degree in Advanced Calculus.
>If it was in fact rated the "new way," that should at least answer the
>question of which system is currently being used.
>
Doesn't seem to have.
>As to the frustration over the reduced rating gain, I feel your pain,
>but think about this: had you achieved a very poor result in this event,
>your net rating loss would also have been lessened.
>
Yeah, the point is, I plan to win. : ) Seriously, this is what I mean about
being discouraged. The new system creates LESS incentive for me to win. If I
win, Im not that much closer to my goal, but if I lose, Im not that much
further away. Ho hum.
>If I understand aright the new sliding-K factor concept, (a big "if"),
>an active, established adult A-player with moderately stable results
>will achieve very nearly the same net rating gain or loss over the
>course of say, a year or so, under both systems. (Corrections from the
>better-informed gladly accepted.)
To paraphrase Homer Simpson: "A year??? But I want it NOOWWWWW" : )
>Which is probably not going to be much consolation to ANY player the
>first time he/she sees a lower-than-expected rating boost from a good
>result.
You got that right : )
cavema...@my-deja.com
ablu...@aol.com (ABlue892) wrote:
> What follows in the quotes is a copy of my recent letter to
USCF/Chess Life:
>
> "I recently had what for me was a very successful tournament (even
score in an
> Open section), and was dismayed to find that I was only scheduled to
receive
> about two thirds of the points I normally would have. It was
explained to me
> that this is a result of the new "improved" rating system.
The new rating formula does not go into effect until next year.
>
> I had just started to make some real headway towards my goal of an
Expert
> rating only to find out now that the USCF has intentionally made this
twice as
> difficult, while at the same time attempting to artificially inflate
the
> ratings of weak players.
Unfortunately, this shows a great misunderstanding of the new rating
system, since neither of the statements is true.
> This discovery was extremely discouraging and has
> significantly dampened my desire to play tournament chess.
The author should reconsider. If anything, part way into next year,
the "anti-deflation" of lower ratings will help to ease his task.
> This is compounded
> by the fact that the formula for determining rating points is now so
complex
> that I cannot even enjoy, as I and many others have in the past,
calculating
> potential rating gains "on the fly" during a tournament.
>
Again, not a true statement. Except for the calculation of K, the
system is very similar. Consequently, just find a good way to estimate
your K factor, and you'll be in the same place as before!
> It seems you are trying to discourage A players such as myself, as
well as
> perhaps B players and experts, from playing tournament chess.
Exactly the opposite. Unfortunately, the author terribly
misunderstands the changes.
> I am curious why
> this is. Are you trying to make the USCF a more scholastic league,
or a league
> primarily of casual players and Masters, with no in-between? If so,
couldn't
> you consider separate scholastic and master leagues and retain a
league for
> competitive tournament players, rather than force them out? Was this
the only
> alternative for some reason? Just curious."
>
You left our the correct alternative -- that your concept of the
changes was mistaken.
SNIP
--
Kevin Bachler
Caveman
"Caveman chess is chess without finesse."
cavema...@my-deja.com
ske...@my-deja.com wrote:
SNIP
> I don't know. I think a separate SCHOLASTIC rating system would make
> sense, as it would prevent abuse such as seen in the "twenty game
match"
> incident, seen somewhere else in a thread.
>
> > >I urge one of you out there, perhaps a wealthy dotcom-er who is
> > >one of the A, B or Expert players being screwed to line the USCF
> > >coffers, try to put together a parallel Chess Federation that
> > >considers the needs of its competitve players, and can run itself
> > >well enough not to have to sell out to anyone who can wave 5 bucks
> > >in front of its face.
I missed this before. What does $5 have to do with anything??
SNIP
> > The scheme to abolish the rating system in favor of fiddle points is
> > the brain-child of Tim Redman, who is rapidly becoming the most
hated
> > man in chess.
>
LOL!! The original author complained about the new rating formula,
which recognizes the greater stability of higher ratings, and so awards
points at some levels more slowly, barring bonus and feedback (which
the author apparently didn't know about.) Then Sam tries to fault
Redman, who brought in a proposal to award EVEN MORE points.
What comic relief.
SNIP
--
Kevin Bachler
Caveman
"Caveman chess is chess without finesse."
cavema...@my-deja.com
> AGREED AGREED AGREED. This is the obvious solution to many of the
USCFs
> problems, a separate league/system for scholastic players, and even
perhaps an
> internet chess club for kids...I guess it takes a non chess
professional to
> figure that out.
If you'd think it through, you'd realize this would actually make the
problem worse, not better. FRom the rating perspective, such a
solution has already been considered by the ratings committee and
rejected.
Good young players would play under both systems -- which is
essentially what you have now. But you would exclude the kids who are
not quite as good, and who play in open tournaments from time to time
and leave points there.
--
Kevin Bachler
Caveman
"Caveman chess is chess without finesse."
Kathy Berkley
ABlue892
>
>The new rating formula does not go into effect until next year.
As I have explained before, the "New System Rating" changes, were posted on a
website. If the rating system is not in effect, I would imagine the USCF
should not encourage such postings.
>while at the same time attempting to artificially inflate
>the
>> ratings of weak players.
>
>Unfortunately, this shows a great misunderstanding of the new rating
>system, since neither of the statements is true.
Just because you say something is not true, does not make it so. Do you deny
that weaker players will gain many many more points for a successful tournament
than stronger ones under the new system? On what basis?
> by the fact that the formula for determining rating points is now so
>complex
>Again, not a true statement. Except for the calculation of K, the
>system is very similar.
Right, except for the complex calculation part, the system is very similar.
You are quite good at doublespeak. You must work for the USCF.
>Exactly the opposite. Unfortunately, the author terribly
>misunderstands the changes.
Don't you think there is a problem when the players misunderstand the changes
so terribly, a situation which, if it is true, your denials have done
absolutely nothing to alleviate?
ABlue892
>essentially what you have now. But you would exclude the kids who are
>not quite as good, and who play in open tournaments from time to time
>and leave points there.
>--
The problem seems to be that good scholastic players, who have artificially low
ratings, are kicking butt on adults, who dont like having ratings in the triple
digits as a result of these losses. Having these students play in scholastic
leagues would alleviate this problem.
cavema...@my-deja.com
I understand the problem. The ratings committee has discussed it, and
it has previously been debated here.
Your proposed solution DOES NOT solve the problem. You will still have
the same players kicking butt on adults, and adults will, as a result,
still lose rating points. Scholastic leagues do nothing to alleviate
that.
You have to understand that whenever ANY player improves, that
improvement causes the system to deflate. Young players are often
improving, so their very existance in the system will cause deflation.
Ex: Suppose 4 players form a closed pool, and all 4 are rated 1500
exactly. Further suppose that one player instantly jumps up to 1700
strength, and that the players all play a large number of games. What
will the ratings look like in a year?
Answer: I can't give an execat answer with only the above information,
but I can tell you what it will look like and not look like. You will
not have 1 player at 1700 and 3 at 1500. What you will have is 3 at a
number, perhaps 1400, and the 4th will be 200 points above that, at
1600. Because the one player is improving, the system will determine
their relative strength, but in a deflated way. (And the numbers
chosen above are examples only, you might have 3 players at 1475 and
one at 1675, but the concept is the same.)
Scholastic leagues will not fix that problem. The new rating system,
which awards feedback and bonus points, will.
cavema...@my-deja.com
> As I have explained before, the "New System Rating" changes, were
posted on a
> website. If the rating system is not in effect, I would imagine the
USCF
> should not encourage such postings.
>
Why not? People still need to understand the system. Originally, it
was to be in effect by now, but it was delayed, I believe due to the
added addition of Activity Points.
SNIP
>Just because you say something is not true, does not make it so. Do
you deny
> that weaker players will gain many many more points for a successful
tournament
> than stronger ones under the new system? On what basis?
>
Yes, I deny it. On what basis? Based on the ratings formula. As a
blanket statement, what you say makes no sense. It depends on many
factors including the actual results, the k factor of the players,
whether bonus and feedback points are involved, etc.
As a matter of fact (tested through significant statistical analysis by
the ratings committee) the ratings of higher rated more established
players are more stable than those of lower rated less established
players. Tne new k factors are set to equate with the actual rates at
which ratings change. They are reflective of reality, not an
artificial construct.
Note that you also keep ignoring bonus and feedback points.
> >Again, not a true statement. Except for the calculation of K, the
> >system is very similar.
>
> Right, except for the complex calculation part, the system is very
similar.
> You are quite good at doublespeak. You must work for the USCF.
>
Nope, I do not work for USCF. If you pay attention, and listen to what
I said, instead of just trying to pick a fight, you'll get it.
Much of the system is the same. The same rules of thumb you used
before will basically still work. You just need to adjust k.
Masters, senior masters, some experts, etc., had to do that already
under the CURRENT system, because they had a lower k factor. Just
estimate your k and you'll be fine doing what you did before.
> >Exactly the opposite. Unfortunately, the author terribly
> >misunderstands the changes.
>
> Don't you think there is a problem when the players misunderstand the
changes
> so terribly, a situation which, if it is true, your denials have done
> absolutely nothing to alleviate?
Yes, I do think there is a problem. The changes have been poorly
communicated, which is sad, because the changes are really very good.
As a result, chessplayers have, as they often do, assumed the worst.
Neither of those statements make your numerical complaints correct
(they aren't). I encourage you to volunteer to help write an
understandable explanation of the system so that other chessplayers
don't run into the same problems you've run into.
cavema...@my-deja.com
Improvement causes deflation.
--
Kevin Bachler
Caveman
"Caveman chess is chess without finesse."
cavema...@my-deja.com
ablu...@aol.com (ABlue892) wrote:
> >If you improve your play to the Expert level then it should
> >be only a matter of time before you reach your goal.
>
> But the reality is one must play HIGHER than the expert level to
reach this
> goal.
This is a false statement.
>If one assumes that an expert loses to other experts half of the time,
> the rating gain for an A player trying to reach expert level is only
the number
> of points per gain represented by the difference between their rating
and that
> level, 100 or 200 points. This difference used to be between 4-8
points, and
> now it seems like even less, meaning the matter of time you refer to
could be
> hundreds of chess games, which is frustrating.
Some of your analysis above is roughly correct, and then you come to an
incredibly wrong conclusion.
If, against mid-experts, an mid- player who is expert strenghth, is
gaining roughly 4 points per game, it will take roughly 25 games, not
hundreds, to reach the expert level.
> Saying "one should want to
> improve their play, rather than their rating" is meaningless,
No, it is the most meaningful statements. Ratings measure performance.
--
Kevin Bachler
Caveman
"Caveman chess is chess without finesse."
Chesspride
>Good young players would play under both systems -- which is
>essentially what you have now. But you would exclude the kids who are
>not quite as good, and who play in open tournaments from time to time
>and leave points there.
And an active rating administrator couldn't fix this with annual point
adjustments?
Eric C. Johnson
Chesspride
>Not to belabor this point forever, but there will be players next year
>who gain HUNDREDS of actvity points.
Yes...perhaps 1 percent of the playing population. I suggest that the rating
officer monitor the pool and take a hard look at these particular cases, and
recommend corrections at year's end.
I favor an active rating administrator.
Eric C. Johnson
li...@ork.net
Chesspride
>> I favor an active rating administrator.
>
>Nothing beats attentive subjectivity, right Eric?
>
Elo himself favored an active rating administrator.
Eric C. Johnson
li...@ork.net
cavema...@my-deja.com
Eric,
The old system is currently being used. For events held (rated!?)
after 1/1/01 the new system -- perhaps we should call it ELO ME --
will be used.
The current plan is that Activity points will also be implemented
then. However, my understanding is that this has not yet been
programmed.
ABlue892
>will be used.
for ELO MEssed up? God...why couldnt they leave well enough alone? Dont they
get enough complaints?
cavema...@my-deja.com
underrated. And testing showed that yes, much (but not all) of the
rating population is underrated (There is a segment that is actually
overrated)
So they changed the system to raise ratings, and to raise them more
quickly. You misunderstand that, and complain that it is harder for
you to gain points.
I'm uncertain how to respond to that complaint.
cavema...@my-deja.com
ablu...@aol.com (ABlue892) wrote:
> >Actually, this means that any A-player with the available time and
money
> >can all but guarantee that he will attain an Expert rating some time
> >next year, just by remaining very active, and avoiding total
meltdown in
> >his/her results.
> >
>
> Considerable issues of available time and money aside, I happen to
think trying
> to cram in 300 games a year is just not conducive to good chess.
Chess
> requires a clear and relaxed mind, and I know from experience that
playing 4-8
> games a week simply deteriorates the quality of my play.
The example given is extreme. Understand that all of your opponents
will also be gaining points, and there should be a "trickle-up" effect.
SNIP
>
> Well if the amount of points I gain under the new system is cut by a
third,
Where in the world did you come up with that?? That's just wrong.
Question: Suppose you had a base salary of $75,000. Your company gets
bought, and the new owners say:
"Your new base salary is $50,000. Your target bonus for average
perfomance is $50,000. YOu are also now a partner in the firm, and
your annual partnership payout is $50,000."
Would you argue that you were going to be paid at only 2/3rds the rate
you were before? Look at the system and all the available points --
Bonus, Feedback, activity, regular, plus the fact that your opponents
will be the same strength bt rated higher.
The arguments are analogous.
>at
> best it seems the fiddle points are helping me break even, while
still keeping
> the problem of artificially inflating the total fish.
No. The only artificial inflationary item in the above is activity
points. It will be in place for only a year, and should help the
system to self-anti-deflate more rapidly. It is possible that activity
points will be inflationery in some realms. However, don't you see
that if weak players are temporarily inflated, that makes it EASIER for
you to gain points??
> (correction my ass, some
> people just can't face the fact that they are just not good. I have
never met
> a person U1800 who didnt think they were underrated by at least 200
points, and
> if EVERYONE thinks that way, the system works, it's the players egos
that don't
> work. THIS is the correction the USCF is catering to)
>
No, its not. Deflation is measurable, and demonstrable. Elo
recognized that the inherent flaw of this system is that it is
naturally deflationery, and he proposed several tools to combat this --
many of which are part of the changes that you are complaining about.
> >As to the rating result in question here, are you sure that the event
> >was rated under the new system? Have you checked the tournament
> >crosstable, or your updated rating at the USCF webpage?
>
> The tournament website rating changes list clearly says across the
top
> "According to the NEW USCF rating system" Whether they just did this
to
> delight and amuse the players and in fact the tournament is being
rated under
> the normal system, I do not know since the USCF hasn't gotten around
to rating
> it yet.
>
The new system is not yet in place. In a conversation 2 days ago with
the ED, the plan is to make it effective 1/1.
> >Please consider that the new system features bonus points, to reward
> >players who achieve much higher than expected TPR's. The apparent
> >absence of such bonus points in this case makes me think that this
event
> >was not rated under the new system, and that there may have been some
> >other snafu in the rating process.
> >
> I have been trying to figure out if I was entitled to bonus points or
not.
> Unfortunately I do not have a degree in Advanced Calculus.
>
Bonus points aren't that hard.
> >If it was in fact rated the "new way," that should at least answer
the
> >question of which system is currently being used.
> >
> Doesn't seem to have.
>
> >As to the frustration over the reduced rating gain, I feel your pain,
> >but think about this: had you achieved a very poor result in this
event,
> >your net rating loss would also have been lessened.
> >
> Yeah, the point is, I plan to win. : ) Seriously, this is what I
mean about
> being discouraged. The new system creates LESS incentive for me to
win.
Then you misunderstand the new system. There is actually increased
incentive.
> If I
> win, Im not that much closer to my goal, but if I lose, Im not that
much
> further away. Ho hum.
>
Ratings are not rewards. They are performance measures.
> >If I understand aright the new sliding-K factor concept, (a
big "if"),
> >an active, established adult A-player with moderately stable results
> >will achieve very nearly the same net rating gain or loss over the
> >course of say, a year or so, under both systems. (Corrections from
the
> >better-informed gladly accepted.)
The range narrows, appropriately so.
>
> To paraphrase Homer Simpson: "A year??? But I want it NOOWWWWW" : )
>
> >Which is probably not going to be much consolation to ANY player the
> >first time he/she sees a lower-than-expected rating boost from a good
> >result.
>
> You got that right : )
>
--
ske...@my-deja.com
wrote:
> In article <20001221013929...@ng-fj1.aol.com>,
> ablu...@aol.com (ABlue892) wrote:
> > >perhaps we should call it ELO ME --
Based on the previous thread regarding Windows ME and CB 8, I'd leave
off the ME. How about "ELOPlus!" or "ELO+"?
> > for ELO MEssed up? God...why couldnt they leave well enough alone?
> > Dont they get enough complaints?
>
> Yes. And the complaints they were getting were about people being
> underrated. And testing showed that yes, much (but not all) of the
> rating population is underrated (There is a segment that is actually
> overrated)
I think I constitute that segment. :-)
> So they changed the system to raise ratings, and to raise them more
> quickly. You misunderstand that, and complain that it is harder for
> you to gain points.
>
> I'm uncertain how to respond to that complaint.
Me too. Is there *any* possibility that the rating report refered to
earlier (where someone gained fewer points than expected) was QUICK
CHESS rated? Any?
SK
Louis Blair
> Deflation is measurable, and demonstrable. Elo
> recognized that the inherent flaw of this system is that it is
> naturally deflationery, and he proposed several tools to combat this --
> many of which are part of the changes that you are complaining about.
I would be very grateful if specifics were provided
on where Elo's comments on this subject could be
found.
cavema...@my-deja.com
quote the exact section for you, but Elo devotes a fair amount of time
to it in this book.
signthescoresheetoldman
The USCF membership and ratings are like the economic system of a small
country with terrible de-flation. Rating floors don't solve the problem.
We need an Alan Greenspan to come in, take action and infuse more points
into this economy, or eventually the middle class will revolt.
Steve Loring
ske...@my-deja.com wrote in message <91ln8h$4no$1...@nnrp1.deja.com>...
>In article <3a3e0646...@nntp.mindspring.com>, sl...@ishipress.com
>(Sam Sloan) wrote:
>> On 18 Dec 2000 05:04:57 GMT, ablu...@aol.com (ABlue892) wrote:
>>
>> >What follows in the quotes is a copy of my recent letter to
>USCF/Chess Life:
>> >It seems you are trying to discourage A players such as myself, as
>> >well as perhaps B players and experts, from playing tournament chess.
>> >I am curious why this is. Are you trying to make the USCF a more
>> >scholastic league, or a league primarily of casual players and
>> >Masters, with no in-between? If so, couldn't you consider separate
>> >scholastic and master leagues and retain a league for competitive
>> >tournament players, rather than force them out? Was this the only
>> >alternative for some reason? Just curious."
>
>I don't know. I think a separate SCHOLASTIC rating system would make
>sense, as it would prevent abuse such as seen in the "twenty game match"
>incident, seen somewhere else in a thread.
>
>> > I doubt I will receive any response. What I suspect has happened is
>> >that the heavily debt-ridden USCF has realized that there is more
>> >money in bringing in lots of patzers and kids and getting those dues,
>> >then in making their existing players happy.
>
>Speculation. USCF needs all its chess players.
>
>> >I urge one of you out there, perhaps a wealthy dotcom-er who is
>> >one of the A, B or Expert players being screwed to line the USCF
>> >coffers, try to put together a parallel Chess Federation that
>> >considers the needs of its competitve players, and can run itself
>> >well enough not to have to sell out to anyone who can wave 5 bucks
>> >in front of its face.
>
>We could hope.
>
>> > The Seattle Chess Federation, which now sponsors the US
>> > Championships, is no better, as they are already suggesting
>> > discriminatory policies towards female players, by 1) attempting
>> > to abolish the Womens Chess Championship, and 2) failing that,
>> > sabotaging it by scheduling it during the first two weeks of
>> > school, when the most talented young women, who are still in
>> > college, will not be able to compete.
>
>Can you verify that?
>
>> I am wondering why you did not post this to rec.games.chess.politics
>> as it really belongs there.
>
>Actually, it's one of those rare posts that belong in *both* forums.
>
>> The scheme to abolish the rating system in favor of fiddle points is
>> the brain-child of Tim Redman, who is rapidly becoming the most hated
>> man in chess.
>
>I dunno, Sam; I still like Tim. I may not agree with everything, but
>he's doing *something*.
>
>> Did you read the letter to the editor on page 6 collumn two of the
>> December Chess Life entitled "Concerned Committee"?
>
>Goodness no. I haven't had the time to open the mag yet.
>
>> Everyone says don't worry about it because Redman's term expires in
>> August and he says he is not running again (he better not) and
>> thereafter sweetness and light will prevail throughout the land.
>
>Oh, I hope not. At least, not the Sweetness and Light of the Schultz
>years.
>
>SK
Chesspride
>>
>> Well if the amount of points I gain under the new system is cut by a
>third,
>
>Where in the world did you come up with that?? That's just wrong.
No, ain't wrong at all.
K is being reduced...to keep ratings more stable and variability down.
If Player X has an unusually good performance...he gets roughly 2/3 the point
boost from the new formula as the old one.
You will counter by saying that bonus points are available...but they are only
available for certain types of events...of a certain number of rounds.
Are they available for 5 rounders? 4 rounders? 3 rounders?
If they are only available for 4+ rounders...and your local club runs weekly 3
rounders...then the local population will be getting ZERO bonus points.
And the gentleman is correct, that under those circumstances...to get the same
level of point boost as under the old system...under the new system he will
need to either win more games...or score better over a longer patch of games.
>Bonus, Feedback, activity, regular, plus the fact that your opponents
>will be the same strength bt rated higher.
>
Bonus points affect only games played in events of sufficient number of rounds.
If memory serves, a typical one-day three round event (like we run at our club
every week) will not qualify.
Feedback points (same issue??)
Activity points. Here, the committee fought against the idea, and activity
points were not part of the new formula. So it really is disingenuous to claim
them as part of the new (improved?) overall formula. However, activity points
*are* available to three-round events, so they alleviate some of the problem
for clubs that can only run weekly three-round events.
>
>No. The only artificial inflationary item in the above is activity
>points. It will be in place for only a year, and should help the
>system to self-anti-deflate more rapidly.
Ok, then don't include these much-hated activity points in your answer above.
>However, don't you see
>that if weak players are temporarily inflated, that makes it EASIER for
>you to gain points??
>
His point was a general one: it will take him more effort, over a larger
number of games, to climb to a new rating peak.
Your counter is that it would take him more games, and more bad performances,
to fall as far down.
He's not impressed by that insurance.
He wants to have a particular good performance, climb over a plateau...and feel
he can bank the points. Might be misguided thinking on his part, but it is
part of the USCF culture right now.
>
>The new system is not yet in place. In a conversation 2 days ago with
>the ED, the plan is to make it effective 1/1.
Here, I humbly assert that it behooves the federation to trumpet the
implementation date LOUDLY and OFTEN prior to implementation...so that folks
are not forced to rely on private phone calls to the office.
The idea that it is Dec. 21 and USCF is not loudly announcing these plans on
its website yet...is very bad indeed.
They have mis-announced the implementation date on several earlier occasions
too...so it's a bit like the boy who cried wolf.
>
>Ratings are not rewards. They are performance measures.
Agreed 100 percent...which is why I have always said that any necessary rating
corrections need not be tied to future wins/losses/draws by the individuals
involved...but should be made based on the parameters of the entire rating
database.
Yet you came back with an argument that we could not do so...because future
corrections should be "earned" through future play...a point that conflicts
with your (correct) assertion here.
Eric C. Johnson
ABlue892
>Me too. Is there *any* possibility that the rating report refered to
>earlier (where someone gained fewer points than expected) was QUICK
>CHESS rated? Any?
>
No, there is none. The previous poster is trying to mislead you into believing
the new system awards more points to everybody when it is clear to anyone with
eyes that in the case of any player over a certain level (I think around 1600)
that it is not. He is uncertain how to respond because there IS no response.
I am beginning to suspect that the reason that the new formula is so much more
complex is because the USCF establishment like Caveman wants to obfuscate this
issue. The fact is, and I invite the USC establishment to dispute this, is
that under the new system, this previously constant rating factor, K, is
variable. Meaning there will be more variability for those players under 1600,
and less for those over. Meaning that players over 1600, or whatever the
number is, will get fewer points for a win, as well as lose fewer for a loss,
whereas weaker players will gain more for a win and lose more for a loss. This
is all well and good for very weak players, for whom it is easy to get points
to begin with, and for Masters, who dont want much variability because they
dont want to have to worry about staying masters, but everyone else, if they
are intending to continue winning more than they lose and thus improve, is
screwed. This fact that everyone complains about being underrated is simply a
result of ego..all weak players tend to say..Im rated 1400 or 1600 but Im
really 1600, or 1800, or whatever, its just (insert random excuse such as
problems playing tournaments, or bad luck, or whatever) here. If EVERYONE
thinks they are 200 points too low, then of course, the rating system is
perfect. Everyone is in exact relation to everyone else as they should be,
they just would like the number to be higher. For the USCF to pander to these
egotistial weaklings is pathetic, and if you are going to do it, why dont you
just give everyone a 200 point bonus so they can pretend they are good and be
happy and leave us serious players alone?
ABlue892
>bought, and the new owners say:
>
>"Your new base salary is $50,000. Your target bonus for average
>perfomance is $50,000.
Does the USCF hire directly from the IRS?? Are there a group of math majors
hired specifically to further obfuscate delicate issues? Under the old system
I got 50, under the new 35, 2/3s of 50 is a little less than 35. For you to
assert that my statement that 2/3 of 50 is about 35 is just WRONG, and to use
some unnecessarily complicated example about "base salaries" to justify it, is
outrageous, but perhaps, typical of the USCF.
ABlue892
>the ED, the plan is to make it effective 1/1.
>
We will see. The tournament I referred to has not yet been officially rated by
the USCF. Perhaps they are holding it until 1/1 to get around this "effective
1/1" problem
Phil Innes
> The scheme to abolish the rating system in favor of fiddle points is
> the brain-child of Tim Redman, who is rapidly becoming the most hated
> man in chess.
sam - this sort of sentence is very difficult to explain to russians
please explain what are fiddle points
is brain-child the same as idea
how fast is rapid? is tim redman pacing himself so that he becomes exactly
the most hated man in chess by august? (you think he is working on a book? -
pretty catchy title, huh?)
cordially, phil
> Sam Sloan
Sam Sloan
wrote:
Dear Phil,
Are you seriously translating this into Russian? I know you have
friends in St. Petersburg and I met your friend Boris from there.
Sam Sloan
cavema...@my-deja.com
ablu...@aol.com (ABlue892) wrote:
> >
> >Me too. Is there *any* possibility that the rating report refered to
> >earlier (where someone gained fewer points than expected) was QUICK
> >CHESS rated? Any?
> >
>
> No, there is none. The previous poster is trying to mislead you into
believing
> the new system awards more points to everybody when it is clear to
anyone with
> eyes that in the case of any player over a certain level (I think
around 1600)
> that it is not.
No, I'm not trying to mislead anyone. It is true that the k factor for
more established players is lower. More established players do in fact
have their ratings change more slowly.
What you misunderstand is that a lower k factor DOES NOT mean that it
is harder to gain points, for several reaons:
1. Activity points will raise the ratings of the general pool, both the
player and his opponents. Thus, points come more easily. The ratings
committee has already demonstrated this in numerous test runs.
2. Bonus points reward good results better than they were rewarded in
the past.
3. Feedback points stop some of the deflation caused, especially by the
improving youngsters.
4. Rating gains are a stochastic not a deterministic phenomenon. Lower
k factors mean that it will be harder to gain OR LOSE points IN A GIVEN
GAME. OVER SEVERAL GAMES, unless you win all the rest of the games you
ever play, the effect here will simply be to narrow your performance
range, not narrow your ability to gain points due to real improvement.
Your comment shows a fundamental misunderstanding of the way that the
Elo system works.
> He is uncertain how to respond because there IS no response.
No, I am uncertain how to respond to a question about why USCF would
decrease the rate at which players gain points, when there has been
significant discussion in this and other forums for MONTHS about the
FACT that the TOTAL system will lead to point gains.
> I am beginning to suspect that the reason that the new formula is so
much more
> complex is because the USCF establishment like Caveman wants to
obfuscate this
> issue.
For what purpose? Moreover, I would hardly call me USCF
establishment. I've been actively involved as a USCF volunteer for a
year.
>The fact is, and I invite the USC establishment to dispute this, is
> that under the new system, this previously constant rating factor, K,
is
> variable.
That is true.
> Meaning there will be more variability for those players under 1600,
> and less for those over.
That is ALSO true.
> Meaning that players over 1600, or whatever the
> number is, will get fewer points for a win, as well as lose fewer for
a loss,
That is also true.
> whereas weaker players will gain more for a win and lose more for a
loss.
You are still on track. Also true.
> This
> is all well and good for very weak players, for whom it is easy to
get points
> to begin with, and for Masters, who dont want much variability
because they
> dont want to have to worry about staying masters, but everyone else,
if they
> are intending to continue winning more than they lose and thus
improve, is
> screwed.
And this is completely false. You've made a completely incorrect leap
in logic.
The ability to gain or lose points at a greater or less rate PER GAME
impacts variability.
It is like a highly volatile stock.
Question: Which would you rather have -- an investment that returns
10% a year three years in a row (a stable investment) OR A HIGHLY
VARIABLE ONE that goes up 30% in year one, down 10% in year 2, and up
10% in year three.
Notice that if you just ADD the results, they look the same: Each adds
to 30%.
But if you correctly COMPOUND the result, the first result grew 133.1%
over 3 years, while the second grew 128.7%.
Volatility can have a negative impact like that. To look at the
volatility on the basis of a single game and conclude that it is
now "harder" to gain points overall is absolute folly. You need to
remember that it is harder to GAIN OR LOSE, and the result works very
well. There are 2 parts to it, (the gain AND loss) not 1 part (gain
only.)
> This fact that everyone complains about being underrated is simply a
> result of ego..
No, it's not. It is mathematically demonstrable that segments of the
rating pool are underrated.
>all weak players tend to say..Im rated 1400 or 1600 but Im
> really 1600, or 1800, or whatever, its just (insert random excuse
such as
> problems playing tournaments, or bad luck, or whatever) here. If
EVERYONE
> thinks they are 200 points too low, then of course, the rating system
is
> perfect.
But the statement is irrelevant.
A rating of 1500 is meaningless if it doesn't roughly indicate the same
value today as yesterday. And in the Elo system, when players are
improving, they force deflation. He discusses this at length in
chapter 3 of his book "The rating of chessplayers."
> Everyone is in exact relation to everyone else as they should be,
> they just would like the number to be higher.
Your comment is mistaken. There are a number of underrated players.
If I recall right, the average amount of deflation (Tom Doan, correct
me if I am wrong) is about 40 points right now.
> For the USCF to pander to these
> egotistial weaklings is pathetic, and if you are going to do it, why
dont you
> just give everyone a 200 point bonus so they can pretend they are
good and be
> happy and leave us serious players alone?
>
Because that isn't the problem.
--
Kevin Bachler
Caveman
"Caveman chess is chess without finesse."
cavema...@my-deja.com
chess...@aol.com (Chesspride) wrote:
> >
> >>
> >> Well if the amount of points I gain under the new system is cut by
a
> >third,
>
> >
> >Where in the world did you come up with that?? That's just wrong.
>
> No, ain't wrong at all.
>
> K is being reduced...to keep ratings more stable and variability down.
>
You look at one part of the system, K, see that a smaller K is used for
more established players, and conclude that players gain points more
slowly.
The logic behind this conclusion is fallacious. It suffers directly
from the fallacy of compositition.
It is true that in a single game an established player will gain or
lose fewer points under this formula than the old formula. It is NOT
TRUE that it becomes harder for a person to gain points over a series
of games. This is because volatility is removed, so downward spikes
are also removed.
See the investment example in another post. Your assumption here,
besides ignoring activity points, bonus points, and feedback points, is
just wrong.
SNIP
> And the gentleman is correct, that under those circumstances...to get
the same
> level of point boost as under the old system...under the new system
he will
> need to either win more games...or score better over a longer patch
of games.
No, that is not clear at all, Eric.
SNIP
> Activity points. Here, the committee fought against the idea, and
activity
> points were not part of the new formula. So it really is
disingenuous to claim
> them as part of the new (improved?) overall formula.
They are there. Why is that disingenuous?
> Ok, then don't include these much-hated activity points in your
answer above.
Why not? Aren't they there? Won't they raise the ratings of my
opponents? Why are you suggesting that we ignore reality?
> His point was a general one: it will take him more effort, over a
larger
> number of games, to climb to a new rating peak.
>
No. The fact that he gains and loses fewer points per game does not
logically conclude there, Eric.
Consider a concrete example:
Suppose X, a 2000 rated player plays in 2 events. In each event, he
faces EXACTLY the same players in exactly the same order. Each time he
faces them, these 3 players have exactly the same rating. We have done
this to make the calculation simple, but there is nothing "funny" about
the construction, since what we will do is to rate the two events, once
with K=32, and once with K=16, to illustrate the point that IT IS NOT
HARDER TO GAIN POINTS.
Player A is rated 1600, Player B is rated 1800, player C is rated 2200.
In event 1, X scores 1-2. In event 2, X scores 2-1. Now watch:
Approximate winning expectencies for X in event 1:
vs A: .92
vs B: .76
vs C: .24
Total winning expectancy is 1.92. A score of 1-2 means that X is below
his winning expectancy.
So, consider the 2 K factors:
If K=32, then 32 x -.92 = -29. X loses 29 points.
If K=16, then 16 x -.92 = -15. X loses 15 points.
X is now rated 1971 if K was 32, or 1985 if K was 16. (Notice that so
far the smaller K factor produces more points.)
Now consider a second event against the same players, where X scores 2-
1.
If K =32 and his rating is 1971, then his winning expectancies are .9
against A, .73 against B, and .21 against C, for a total expectancy of
1.84.
He scored 2-1, so 2-1.84 is .16 x 32 = 5.12, a gain of 5 points for a
new rating of 1976.
If K=16 and X's rating is 1985, then his winning expectancy is .91 for
A, .74 for B and .23 for C, for a total of 1.88.
He scored 2-1 so 2 - 1.88 = .12 X 16 = 1.92, a gain of 2 points for a
new rating of 1987.
Now Eric, under which K factor, the HIGHER OR THE LOWER, was it easier
to gain points??
> Your counter is that it would take him more games, and more bad
performances,
> to fall as far down.
>
> He's not impressed by that insurance.
>
Now he should be. Unless he plans on winning every single game that he
plays forever from now on, the statement he is making is called into
question.
> He wants to have a particular good performance, climb over a
plateau...and feel
> he can bank the points. Might be misguided thinking on his part, but
it is
> part of the USCF culture right now.
>
Hopefully the education above will help him.
> Here, I humbly assert that it behooves the federation to trumpet the
> implementation date LOUDLY and OFTEN prior to implementation...so
that folks
> are not forced to rely on private phone calls to the office.
>
I agree, and stated this at the time.
> Agreed 100 percent...which is why I have always said that any
necessary rating
> corrections need not be tied to future wins/losses/draws by the
individuals
> involved...but should be made based on the parameters of the entire
rating
> database.
>
The two are unrelated. Moreover, your suggestion demonstrably does not
fix the issues raised, and causes more issues. We've been through this.
> Yet you came back with an argument that we could not do so...because
future
> corrections should be "earned" through future play...a point that
conflicts
> with your (correct) assertion here.
>
I guess I'm missing your point, or you entirely missed mine, because I
have no clue of what you meant with the last statement.
cavema...@my-deja.com
and feedback points. Oh wait, the system does that.
cavema...@my-deja.com
by a certain date are included in the next supplement. It will
probably be rated BEFORE the new system goes into effect.
cavema...@my-deja.com
I do not work for USCF. And try to follow this again.
If you had a $75K salary before, and now you have a $50K salary AND a
$50K bonus, do you earn more, or less, than before??
For you to focus ONLY ON K, and ONLY ON K WHEN GAINING POINTS, is
completely in error. It would be like you focusing ONLY ON SALARY in
the above example, and claiming that that means you make less.
Louis Blair
> Deflation is measurable, and demonstrable. Elo
> recognized that the inherent flaw of this system is that it is
> naturally deflationery, and he proposed several tools to combat this --
> complaining about.
I wrote:
> I would be very grateful if specifics were provided
> on where Elo's comments on this subject could be
> found.
Kevin Bachler writes:
> I don't have a copy of The Rating of Chessplayers here at the office to
> quote the exact section for you, but Elo devotes a fair amount of time
> to it in this book.
Perhaps Kevin Bachler could make note of the
page numbers and bring them in to "the office"
sometime in the near future. Perhaps someone
like "Staunton" would then be kind enough to
let us know exactly what Elo said. It seems to
me that it would be of great interest in this
discussion to know exactly what "tools"
Elo "proposed" and exactly how "many" of
those "tools ... are part of the changes that
[ABlue892 is] complaining about."
cavema...@my-deja.com
Louis Blair <lbl...@pilot.lsus.edu> wrote:
> Kevin Bachler wrote:
> > Deflation is measurable, and demonstrable. Elo
> > recognized that the inherent flaw of this system is that it is
> > naturally deflationery, and he proposed several tools to combat
this --
> > many of which are part of the changes that [ABlue892 is]
> > complaining about.
>
> I wrote:
> > I would be very grateful if specifics were provided
> > on where Elo's comments on this subject could be
> > found.
>
> Kevin Bachler writes:
>
> > I don't have a copy of The Rating of Chessplayers here at the
office to
> > quote the exact section for you, but Elo devotes a fair amount of
time
> > to it in this book.
>
> Perhaps Kevin Bachler could make note of the
> page numbers and bring them in to "the office"
> sometime in the near future.
Your attitude here seems very sarcastic, and I am very confused about
why. I wrote the above while I was at work, and I didn't happen to
have a copy of The Rating of Chessplayers by Arpad Elo with me at the
time.
> Perhaps someone
> like "Staunton" would then be kind enough to
> let us know exactly what Elo said. It seems to
> me that it would be of great interest in this
> discussion to know exactly what "tools"
> Elo "proposed" and exactly how "many" of
> those "tools ... are part of the changes that
> [ABlue892 is] complaining about."
This has been previously discussed online at great length and is easy
to find in Deja. I'm a little surprised again at the attitude here --
why are you expecting others to do your research for you?? And what
does Staunton have to do with anything? Who are you referring to? An
online poster??
At any rate, Elo devotes an entire chapter (chapter 3) to Rating
Administration. In 3.6 he discusses the management of deflation, and
in section 3.7 he discusses various processes to managed deflation.
Several of these ideas have been incorporated into the new rating
system, although not necessarily in exactly the form suggested by Elo.
These included:
3.71 Appropriate treatment of unrated players (modifications were made
to the system for this.)
3.72 Modified processing of provisionally rated players (A form of this
was incorporated)
3.73 Adjustment of K, setting it high when a player is unestablished
and lowering it over time. (A form of this was adopted.)
3.74 Corrective additions for exceptional performances. This means to
add more points if someone does well. (The new rating system includes
bonus points, so they did this too.)
3.75 Feedback process. Elo envisioned this as recalculating the
ratings for an event a second time, replacing the "pre-event" rating
for exceptional performers with their new post event rating, and
calculating the changes in their opponent's ratings using this new
rating. (The new system does have feedback points, although I don't
think it is done quite this way. I'd have to go back and check.)
Those were the 5 ideas Elo suggested to combat deflation, i.e. to keep
ratings up where they belong. ALL 5 ARE INCORPORATED IN THE NEW SYSTEM.
In addition, a 6th idea, Activity points, were added for 1 year.
Given that, its very hard to imagine that anyone really believes that
it is harder to gain points under the new system than the old.
cavema...@my-deja.com
> egotistial weaklings is pathetic, and if you are going to do it, why
dont you
> just give everyone a 200 point bonus so they can pretend they are
good and be
> happy and leave us serious players alone?
>
--
Kevin Bachler
Caveman
"Caveman chess is chess without finesse."
ABlue892
>--
Exactly. A Life master. The rating change doesn't affect you, or non-life
Masters who WANT lower variability because they dont want to lose their master
status. A player like myself, who wants to get better by continually winning
against stronger opposition is out of luck. I am expected to go play class
tournaments against fish to take advantage of their newly inflated ratings when
what I want, is to gain rating points quickly by beating stronger opposition,
which I can no longer do. Now you at the USCF may think this is not a "noble"
goal, or some foolishness, but I imagine most players trying to improve feel
similarly.
ABlue892
>lose fewer points under this formula than the old formula. It is NOT
>TRUE that it becomes harder for a person to gain points over a series
>of games. This is because volatility is removed, so downward spikes
>are also removed.
Meaning, in English, that strong but non master players are discouraged from
competition, because if they have a great performance, they will gain few
points, but if they dont, they wont lose that many, so the entire thrill, fire
of the competition is eliminated. No great loss to the USCF establishment. On
the other hand, new players (read New Membership dues, new tournament entry
fees, maybe new merchandise and internet customers) ARE encouraged to play
because they can get huge rating jumps and be deluded into thinking they are
improving, when in fact the next tournament will bust them back down to 1200 or
whatever, resulting in them, with their inflated egos now brimming with
fantasies that they are really unrecognized experts, will throw even more money
into the USCF machine in order to show everyone how good they really are.
A desperate move to clean up the USCF debt I suppose. I hope disenfranchising
a big chunk of current loyal members is worth it.
Louis Blair
> Deflation is measurable, and demonstrable. Elo
> recognized that the inherent flaw of this system is that it is
> naturally deflationery, and he proposed several tools to combat this --
> many of which are part of the changes that [ABlue892 is]
> complaining about.
I wrote:
> I would be very grateful if specifics were provided
> on where Elo's comments on this subject could be
> found.
Kevin Bachler wrote:
> I don't have a copy of The Rating of Chessplayers here at the office to
> quote the exact section for you, but Elo devotes a fair amount of time
> to it in this book.
I wrote:
> Perhaps Kevin Bachler could make note of the
> page numbers and bring them in to "the office"
> sometime in the near future. Perhaps someone
> like "Staunton" would then be kind enough to
> let us know exactly what Elo said. It seems to
> me that it would be of great interest in this
> discussion to know exactly what "tools"
> Elo "proposed" and exactly how "many" of
> those "tools ... are part of the changes that
> [ABlue892 is] complaining about."
Kevin Bachler writes:
> Your attitude here seems very sarcastic
No sarcasm was intended. I meant exactly what
I said.
Kevin Bachler writes:
> I'm a little surprised again at the attitude here --
> why are you expecting others to do your research for you??
Kevin Bachler is the one who used Elo to support
his arguments. My belief is that when a person
claims that his position is supported by some
authority, that person should be willing to describe
exactly where that support is to be found.
Kevin Bachler writes:
> And what
> does Staunton have to do with anything? Who are
> you referring to? An online poster??
Someone going by the name of "Staunton" has
posted the texts of various items as contributions
to some recent discussions at rec.games.chess.misc .
Kevin Bachler writes:
> At any rate, Elo devotes an entire chapter (chapter 3) to Rating
> Administration. In 3.6 he discusses the management of deflation, and
> in section 3.7 he discusses various processes to managed deflation.
>
> Several of these ideas have been incorporated into the new rating
> system, although not necessarily in exactly the form suggested by Elo.
> These included:
>
> 3.71 Appropriate treatment of unrated players (modifications were made
> to the system for this.)
>
> 3.72 Modified processing of provisionally rated players (A form of this
> was incorporated)
>
> 3.73 Adjustment of K, setting it high when a player is unestablished
> and lowering it over time. (A form of this was adopted.)
>
> 3.74 Corrective additions for exceptional performances. This means to
> add more points if someone does well. (The new rating system includes
> bonus points, so they did this too.)
>
> 3.75 Feedback process. Elo envisioned this as recalculating the
> ratings for an event a second time, replacing the "pre-event" rating
> for exceptional performers with their new post event rating, and
> calculating the changes in their opponent's ratings using this new
> rating. (The new system does have feedback points, although I don't
> think it is done quite this way. I'd have to go back and check.)
>
> Those were the 5 ideas Elo suggested to combat deflation, i.e. to keep
> ratings up where they belong. ALL 5 ARE INCORPORATED IN THE NEW SYSTEM.
I am grateful to Kevin Bachler for providing these
additional details.
cavema...@my-deja.com
> I wrote:
> > Perhaps Kevin Bachler could make note of the
> > page numbers and bring them in to "the office"
> > sometime in the near future. Perhaps someone
> > like "Staunton" would then be kind enough to
> > let us know exactly what Elo said. It seems to
> > me that it would be of great interest in this
> > discussion to know exactly what "tools"
> > Elo "proposed" and exactly how "many" of
> > those "tools ... are part of the changes that
> > [ABlue892 is] complaining about."
I guess putting "to the office" in quotes threw me. In this medium
quotes can change the meaning. The tone seemed to shift. When you
then asked for poof of what Elo said, it came across as a challenge,
not a request.
Sorry for any misunderstanding.
cavema...@my-deja.com
ablu...@aol.com (ABlue892) wrote:
> >PS I am a serious player. Life Master.
> >--
>
> Exactly. A Life master. The rating change doesn't affect you,
Of course it does.
or non-life
> Masters who WANT lower variability because they dont want to lose
their master
> status.
Variability does not impact life masters. Moreover, as noted
previously, the ratings committee has on several occasions confirmed
what Elo wrote years ago -- that more established players have less
variability.
Variability does not necessarily help your quest to become a higher
rated player.
> A player like myself, who wants to get better by continually winning
> against stronger opposition is out of luck.
No, you are better off. You have more opportunities to gain points.
> I am expected to go play class
> tournaments against fish to take advantage of their newly inflated
ratings
No, the ratings of the general pool will rise.
>when
> what I want, is to gain rating points quickly by beating stronger
opposition,
> which I can no longer do.
Why can't you beat stronger opposition anymore?
> Now you at the USCF may think this is not a "noble"
> goal, or some foolishness, but I imagine most players trying to
improve feel
> similarly.
>
Most players who are trying to improve learn that the most important
thing in chess is to not let their emotions get in the way of reality.
Cold, hard evaluation of a position is important.
The fact of the matter is that the steps that have been implemented
will enhance your ability to achieve and maintain a high rating over
the currrent situation. These methods were documented by Elo long ago,
and I have posted them elsewhere in this thread.
The Ratings Committee has tested these methods by running test "re-
ratings" of tournaments over a several year period. The results are
exactly the opposite of what you claim.
So you have a choice. On the one hand you have the mathematically
devised theory that Elo wrote, you have the mathematically devised
theory written by the Ratings Committee, and you have several years of
test data that say these methods will combat existing deflation, and
make it easier to gain points. On the other hand, you have your
unsubstantiated claims.
If you want to become a strong chessplayer, deal with the reality.
cavema...@my-deja.com
ablu...@aol.com (ABlue892) wrote:
> >It is true that in a single game an established player will gain or
> >lose fewer points under this formula than the old formula. It is NOT
> >TRUE that it becomes harder for a person to gain points over a series
> >of games. This is because volatility is removed, so downward spikes
> >are also removed.
>
> Meaning, in English, that strong but non master players are
discouraged from
> competition, because if they have a great performance, they will gain
few
> points,
Meaning, that you still don't get it. If you have a great result, you
get bonus points. Irrespective of your result, you get activity
points. If one of your opponents has a great result, you also get
feedback points.
I take you back to the analogy of earning a $75K salary vs a $50K
salary and a $50K bonus. YOU KEEP IGNORING THE BONUS and insisting
that in the second case you get paid less money. Wake up.
> but if they dont, they wont lose that many, so the entire thrill, fire
> of the competition is eliminated.
No, it is not. Are you saying you only get thrill out of chess due to
the exchange of rating points? Rule #1, you need for the thrill to be
in playing well, if you want to be a good player.
Moreover, your statement is still mathematically incorrect. All that
decreases is variability.
> No great loss to the USCF establishment.
HUH??
> On
> the other hand, new players (read New Membership dues, new tournament
entry
> fees, maybe new merchandise and internet customers) ARE encouraged to
play
> because they can get huge rating jumps and be deluded into thinking
they are
> improving, when in fact the next tournament will bust them back down
to 1200 or
> whatever, resulting in them, with their inflated egos now brimming
with
> fantasies that they are really unrecognized experts, will throw even
more money
> into the USCF machine in order to show everyone how good they really
are.
>
How sad. You simply don't want to understand the system. New players
will actually have a more conservative provisional rating. The changes
implemented help existing players.
Why do you want to be angry about this? The facts are almost entirely
opposite of what you've heard, and instead of saying "Wow, that's
great, I'm glad its not what I heard" you keep looking for resons to be
upset.
The rating system is mathematically tested. It has nothing to do with
ego.
> A desperate move to clean up the USCF debt I suppose. I hope
disenfranchising
> a big chunk of current loyal members is worth it.
>
Yes, if you quit playing, that will certainly help you to reach your
goal of becoming a strong player.
ABlue892
>> Masters who WANT lower variability because they dont want to lose
>their master
>> status.
>
>Variability does not impact life masters.
Yes, the prefix "non" before a word generally indicates the negation of that
word, you can read it as "not".
>> A player like myself, who wants to get better by continually winning
>> against stronger opposition is out of luck.
>
>No, you are better off. You have more opportunities to gain points.
Well, since every time I win I gain fewer points, maybe you do not know what
the word "opportunity" means either.
>>when
>> what I want, is to gain rating points quickly by beating stronger
>opposition,
>> which I can no longer do.
>
>Why can't you beat stronger opposition anymore?
You can't possibly be this stupid. Obviously the part of the sentence that I
can no longer do is gain rating points quickly, not beat stronger opposition.
Put more simply, playing stronger players in the hopes of gaining many rating
points is downplayed in favor of (according to you) playing weak players who
have been pumped up, which will cause "the ratings of the general pool to rise"
although frankly, the benefit of getting to take points from someone who has
been pumped up from a 1200 to a 1500 for me is exactly zero.
>the most important
>thing in chess is to not let their emotions get in the way....Cold, hard
evaluation....is important.
You may wish to reinforce the stereotype of chessplayers as emotionless number
crunching nerds who expect, due to statistical probabilities, and because
according to this algorithim or that laboratory experiment, that they will lose
exactly 2.4 games out of 5 , and over a period of 7.3 years according to
whatever formula, they will achieve X rating, but I do not. I am a human
being, and a competitive chessplayer, and I am not interested in losing, either
over a long period of time, or in my next tournament. Perhaps you are so
gifted at chess that you were able to become a life master without any work, or
with using some kind of algorithm, but for the rest of us, it requires heart,
emotion, inspiration, and determination, things that you, and perhaps the
voting board of the USCF, with your clinical, disinterested approach, do not
appreciate or understand.
cavema...@my-deja.com
ablu...@aol.com (ABlue892) wrote:
> > non-life
> >> Masters who WANT lower variability because they dont want to lose
> >their master
> >> status.
> >
> >Variability does not impact life masters.
>
> Yes, the prefix "non" before a word generally indicates the negation
of that
> word, you can read it as "not".
Wow. You are so angry, that not only do you ignore facts and good
news, you ignore when you are being agreed with.
>
> >> A player like myself, who wants to get better by continually
winning
> >> against stronger opposition is out of luck.
> >
> >No, you are better off. You have more opportunities to gain points.
>
> Well, since every time I win I gain fewer points, maybe you do not
know what
> the word "opportunity" means either.
>
Yes, I do. Your statement above continues to reflect the error of your
logic.
First, it isn't a true statement. If you are in the realm of receiving
bonus points, the above statement may not be true. Please, answer this
one question. Why do you keep ignoring bonus points? It's an obvious
flaw in your logic, yet you continue to ignore it.
Second, even if it were true, the statement above is only a half
truth. We have already explained, and even provided an example in this
thread that showed that because you also lose fewer points with each
loss, that you can be better off. You also continue to ignore this.
Ignoring such obvious facts can lead me only to conclude that your real
purpose hear is to voice displeasure with USCF, and you are willing to
lie (by ommission) to do so. This is very sad.
> >>when
> >> what I want, is to gain rating points quickly by beating stronger
> >opposition,
> >> which I can no longer do.
> >
> >Why can't you beat stronger opposition anymore?
>
> You can't possibly be this stupid.
I simply responded to what you actually wrote. Are YOU that stupid??
> Obviously the part of the sentence that I
> can no longer do is gain rating points quickly, not beat stronger
opposition.
This may be what you meant, but it isn't what you wrote. Moreover, the
statement is false. If you continuously beat stronger opponents, you
will receive bonus points.
> Put more simply, playing stronger players in the hopes of gaining
many rating
> points is downplayed in favor of (according to you)
No. I did not say that playing stronger players is downplayed. In
fact, I said the opposite. You are no worse off playing stronger
players under the new system than you are now.
> playing weak players who
> have been pumped up, which will cause "the ratings of the general
pool to rise"
> although frankly, the benefit of getting to take points from someone
who has
> been pumped up from a 1200 to a 1500 for me is exactly zero.
You keep missing the point. The question is would you rather beat an
underrated 2000, or a correctly rated 2040?
>
> >the most important
> >thing in chess is to not let their emotions get in the way....Cold,
hard
> evaluation....is important.
>
> You may wish to reinforce the stereotype of chessplayers as
emotionless number
> crunching nerds who expect, due to statistical probabilities, and
because
> according to this algorithim or that laboratory experiment, that they
will lose
> exactly 2.4 games out of 5 , and over a period of 7.3 years according
to
> whatever formula, they will achieve X rating, but I do not. I am a
human
> being, and a competitive chessplayer, and I am not interested in
losing, either
> over a long period of time, or in my next tournament. Perhaps you
are so
> gifted at chess that you were able to become a life master without
any work,
LOL. It took years. I had to work at chess very hard.
>or
> with using some kind of algorithm, but for the rest of us, it
requires heart,
> emotion, inspiration, and determination, things that you, and perhaps
the
> voting board of the USCF, with your clinical, disinterested approach,
do not
> appreciate or understand.
>
Your comment, especially made about me, is one of the most moronic
yet. Why do you think people call me Caveman?? Because my chess is
passionless?? LOL!! My style is EXACTLY THE OPPOSITE of what you
describe.
Insteading of continuing to jump to conclusions and continuing to look
stupid, why don't you actually listen and learn something about the
system? You'll probably like it.
Tell you what. Send me that last result that you are worried about.
Phil Innes
Sloan) wrote:
> Dear Phil,
>
> Are you seriously translating this into Russian? I know you have
> friends in St. Petersburg and I met your friend Boris from there.
>
> Sam Sloan
i am translating it into english, sam
and i do remember a conversation about you
what i am really writing to russians is what we are not doing in the west
-they seem to think that we talk instead of act
-we seem to think they are or were a remote bastion of chess riches
-they think we have great power in the west, and i have to tell them ths is
true, but we don't know it and are pretty frivolous, not having a 2 year
plan between our ears and proud of it
-we think they still have all the best players in the world (actually, as
you know, they have been watching british teams for some time, but now asian
players are appearing in strength)
-they think we will do anything for a short-term buck, whereas, we realise
too late that this is folly
-we think they are 'fidé' fans because it is not corrupted by western
commercialism - whereas they are not fidé fans and wish to be corrupted by
western commercialism to the extent that players have enough money to live
on
-they think they would like some frank talking east-west about the state of
things
-we don't know what we think and won't publish it anyway and are proud of it
glad you asked, phil
ABlue892
>First, it isn't a true statement. If you are in the realm of receiving
>bonus points, the above statement may not be true. Please, answer this
>one question. Why do you keep ignoring bonus points?
I responded to this in another thread, which essentially said since I am not a
math major, I asked someone who is to explain bonus points to me, and he
explained to me that if I went 4.5/9 in a tournament against all 2300s, I would
gain 6 bonus points. This is supposed to be some big prize? Ridiculous.
ABlue892
>
>Wow. You are so angry, that not only do you ignore facts and good
>news, you ignore when you are being agreed with.
Yes I am angry. I am angry that you are trying to feed me s**t and tell me its
ice cream.
>> >No, you are better off. You have more opportunities to gain points.
>>
>> Well, since every time I win I gain fewer points, maybe you do not
>know what
>> the word "opportunity" means either.
>>
>Yes, I do. Your statement above continues to reflect the error of your
>logic.
>
>First, it isn't a true statement. If you are in the realm of receiving
>bonus points, the above statement may not be true. Please, answer this
>one question. Why do you keep ignoring bonus points? It's an obvious
>flaw in your logic, yet you continue to ignore it.
Suggesting that the fact that if I have a 2300 performance in a tournament as
an A Player, I will gain 6 extra points is an "obvious flaw in my logic" is
obviously ridiculous. You must see that.
>
>Second, even if it were true, the statement above is only a half
>truth. We have already explained, and even provided an example in this
>thread that showed that because you also lose fewer points with each
>loss, that you can be better off. You also continue to ignore this.
I'm not ignoring it, Im just not interested. Its nice that the USCF is
pandering to losers, but I am not interested in losing.
>Ignoring such obvious facts can lead me only to conclude that your real
>purpose hear is to voice displeasure with USCF, and you are willing to
>lie (by ommission) to do so. This is very sad.
>
My real purpose is to regain the system where I get the rating points I
rightfully deserve. I could care less about voicing displeasure with the
USCF. I have better things to do. I just don't like having establishments
push me around.
>> >>when
>> >> what I want, is to gain rating points quickly by beating stronger
>> >opposition,
>> >> which I can no longer do.
>
>> Obviously the part of the sentence that I
>> can no longer do is gain rating points quickly, not beat stronger
>opposition.
>This may be what you meant, but it isn't what you wrote.
It is exactly what I wrote. Continue to try to discombobulate me on the math,
if you wish, but not the English...that you won't get away with.
> Moreover, the
>statement is false. If you continuously beat stronger opponents, you
>will receive bonus points.
I have already consulted with someone who can decipher the math, and he assures
me for all intents and purposes, I will not.
>> Put more simply, playing stronger players in the hopes of gaining
>many rating
>> points is downplayed in favor of (according to you)
>
>No. I did not say that playing stronger players is downplayed. In
>fact, I said the opposite. You are no worse off playing stronger
>players under the new system than you are now.
It depends on your definition of "worse off" You, as a Life Master who doesnt
care about points, doesnt think being awarded 8 points rather than 16 for
beating a stronger player is "worse off", I, and any sensible player who is not
a master, would. I am much better off trying to squeeze in as many games as I
can in one night against anyone, figuring on gaining the fiddle points and
picking up points from inflated players, and not worrying about many terrible
chess games because the new system will not penalize me as much for losses,
than I am trying to play a number of slow, good games over a long period of
time against strong opposition, which will result in a much slower point gain
even if I am successful.
>
>You keep missing the point. The question is would you rather beat an
>underrated 2000, or a correctly rated 2040?
I would rather get the points I deserve for beating the 2000,. then get two
thirds as many for beating the 2040, if you don't mind.
It is clear that I don't like it, and your continuing to insist that I would
like it if only I understood it is absurd. I obviously understand it
perfectly. I have not made one factual statement about the new system that you
have denied with actual facts. I think I should get the points that I deserve,
and would have gotten under the new system. What you are saying that I dont
understand is that it balances out over the long run, because I get bonus
points, plus I lose less for losses, plus when the weaker players get boosts, I
can scalp points from them. What I am saying is that the new system does NOT
provide any realistic way for me to get bonus points, and I am not interested
in playing inflated fish or having a safety net for losses, I play to win, so
guess what I STILL do not like the system! And furthermore I intend to fight it
any way I can think of , and short of you telling me that I actually like the
system and am just too stupid to understand it, I still havent seen you show
me why we competitive players shouldnt object.
cavema...@my-deja.com
rating. So if you share some information, I will check it.
Second of all, ratings aren't a prize. They are a measurement of
performance. Your comment is consequently ridiculous. It sounds like
you want it to be easy to gain points against strong players. DUH, its
not.
Third, do you realize that experts and masters ALREADY USE a lower K
factor??
cavema...@my-deja.com
ablu...@aol.com (ABlue892) wrote:
>
> >
> >Wow. You are so angry, that not only do you ignore facts and good
> >news, you ignore when you are being agreed with.
>
> Yes I am angry. I am angry that you are trying to feed me s**t and
tell me its
> ice cream.
>
understand the situation.
Do you really think that, for example, the highly educated people on
the Ratings Committee would spend years of analysis, and fight years of
political battles, and push and push and push to get a rating system
change implemented if they didn't feel it enhanced the integrity of the
system? Have you ever listened to these people talk about the
integrity of the system? They view it as sacrosanct. People
consistently complain that they won't loosen up their view a little
bit. Your complaint is so far out of left field, its incredible.
Tell you what. Search deja for complaints on ratings over the past few
years, and see if most of the complaints are similar to yours, or
exactly the opposite direction.
> >> >No, you are better off. You have more opportunities to gain
points.
> >>
> >> Well, since every time I win I gain fewer points, maybe you do not
> >know what
> >> the word "opportunity" means either.
> >>
> >Yes, I do. Your statement above continues to reflect the error of
your
> >logic.
> >
> >First, it isn't a true statement. If you are in the realm of
receiving
> >bonus points, the above statement may not be true. Please, answer
this
> >one question. Why do you keep ignoring bonus points? It's an
obvious
> >flaw in your logic, yet you continue to ignore it.
>
> Suggesting that the fact that if I have a 2300 performance in a
tournament as
> an A Player, I will gain 6 extra points is an "obvious flaw in my
logic" is
> obviously ridiculous. You must see that.
First of all, I never suggested that. Second of all, its not obviously
ridiculous. Tell me your rating and let me walk through the
calculation for you.
> >
> >Second, even if it were true, the statement above is only a half
> >truth. We have already explained, and even provided an example in
this
> >thread that showed that because you also lose fewer points with each
> >loss, that you can be better off. You also continue to ignore this.
>
> I'm not ignoring it, Im just not interested. Its nice that the USCF
is
> pandering to losers, but I am not interested in losing.
Why do you continue to view this as something that is "pandering to
losers". It has nothing to do with that. It has to do with the fact
that if you measure the amount that established players fluctuate in
performance, the rating system actually was causing their points to
fluctuate MORE than what their performance did.
Imagine a child on a swing, who normally swings 20 degrees to each side
of vertical. The USCF rating system was like the bully pushing the
child to 90 degrees from vertical. The system was simply wratcheted to
match the measurement of reality.
>
> >Ignoring such obvious facts can lead me only to conclude that your
real
> >purpose hear is to voice displeasure with USCF, and you are willing
to
> >lie (by ommission) to do so. This is very sad.
> >
> My real purpose is to regain the system where I get the rating points
I
> rightfully deserve.
Rightfully deserve?? RATING POINTS ARE A MEASUREMENT OF STRENGTH, NOT
A RIGHT. And you think USCF is pandering to emotions?? Your argument
completely lacks facts, and is ONLY an appeal to your anger.
> I could care less about voicing displeasure with the
> USCF. I have better things to do. I just don't like having
establishments
> push me around.
>
Good, because the people who got this done were not the establishment.
> >> >>when
> >> >> what I want, is to gain rating points quickly by beating
stronger
> >> >opposition,
> >> >> which I can no longer do.
>
> >
> >> Obviously the part of the sentence that I
> >> can no longer do is gain rating points quickly, not beat stronger
> >opposition.
> >This may be what you meant, but it isn't what you wrote.
>
> It is exactly what I wrote. Continue to try to discombobulate me on
the math,
> if you wish, but not the English...that you won't get away with.
>
I can see that, because you can't read or write it.
> > Moreover, the
> >statement is false. If you continuously beat stronger opponents, you
> >will receive bonus points.
>
> I have already consulted with someone who can decipher the math, and
he assures
> me for all intents and purposes, I will not.
>
Then he is mistaken.
> >> Put more simply, playing stronger players in the hopes of gaining
> >many rating
> >> points is downplayed in favor of (according to you)
> >
> >No. I did not say that playing stronger players is downplayed. In
> >fact, I said the opposite. You are no worse off playing stronger
> >players under the new system than you are now.
>
> It depends on your definition of "worse off" You, as a Life Master
who doesnt
> care about points,
LOL!! Like I don't want to ever reach 2400. What a moron. You have
no clue whether I care about points or not.
> doesnt think being awarded 8 points rather than 16 for
> beating a stronger player is "worse off",
>I, and any sensible player who is not
> a master, would.
No. Any sensible player knows that it is a matter of winning, losing
and drawing over time, and that it is the overall result that matters.
Only someone less than sensible would focus on only a minimal aspect
and declare that the most important.
Do you understand that the actually fluctuation of performance of a
2300 is already half that of a less established player?? Do you
understand that 2300's already have a lower k factor, set that way 10
or more years ago, to match their actual results?
> I am much better off trying to squeeze in as many games as I
> can in one night against anyone, figuring on gaining the fiddle
points and
> picking up points from inflated players, and not worrying about many
terrible
> chess games because the new system will not penalize me as much for
losses,
LOL, you still don't get it. Again, you make a one side argument THE
OTHER direction.
Look. Get this. If YOU EVER consider wins without consider losses (or
draws) it is an error in logic. Period. They must ALWAYS be
considered together.
Right now you are trying to pick up points from deflated players. That
bothers most chessplayers, and they want it set right. You are the
only person I have heard complain that if we put the system back where
it belongs, that it makes the currently deflated players inflated.
> than I am trying to play a number of slow, good games over a long
period of
> time against strong opposition, which will result in a much slower
point gain
> even if I am successful.
>
No, it won't. Answer me a question: Do you really believe that you
will NEVER, EVER again have a result in an individual game that will
cause you to lose rating points in that game.
Do you understand that there is a difference between variability and
performance?
> >
> >You keep missing the point. The question is would you rather beat an
> >underrated 2000, or a correctly rated 2040?
>
> I would rather get the points I deserve for beating the 2000,. then
get two
> thirds as many for beating the 2040, if you don't mind.
>
Huh? Why would you get both sets of points? Did you mean "than"
not "then"?
The guy is rated 2000 today, and is underrated by 40 points. Tell me
your rating, and roughly how many games you've played in USCF. What
makes you think you will get only 2/3 as many points until we actually
look at the calculation?
SNIP
>
> It is clear that I don't like it, and your continuing to insist that
I would
> like it if only I understood it is absurd.
Yes, because you want to complain, whether you understand it or not.
> I obviously understand it perfectly.
ROFLMAO. So far I don't think you've had one fact correct.
> I have not made one factual statement about the new system that you
> have denied with actual facts.
Then you haven't read. I think I've contradicted nearly every
statement you've made, with actual facts.
> I think I should get the points that I deserve,
Points aren't something "deserved." That's a fact.
> and would have gotten under the new system. What you are saying that
I dont
> understand is that it balances out over the long run, because I get
bonus
> points, plus I lose less for losses, plus when the weaker players get
boosts,
No, YOU keep sayin that weaker players get boosts. I don't say that.
I say that deflated players get boosts. There are different amounts of
deflation at different points of the system. Some of the very young
new kids are actually INFLATED and the system fixes that issue too.
Masters can be deflated too.
Listen to what I actually say, "Mr I understand it perfectly."
> I
> can scalp points from them. What I am saying is that the new system
does NOT
> provide any realistic way for me to get bonus points,
I believe you are wrong. Give me your actual rating and let's check it
out.
>and I am not interested
> in playing inflated fish
Good, because the system will work to do away with them! We agree.
> or having a safety net for losses,
Good, because the system doesn't provide one, other than keeping the
current system of floors.
> I play to win,
Playing to win and winning are different. Or are you really 3000, not
an A player?
> so
> guess what I STILL do not like the system!
And as shown by your comments above, you still don't understand it, in
spite of your claims.
>And furthermore I intend to fight it
> any way I can think of ,
It's a done deal.
>and short of you telling me that I actually like the
> system and am just too stupid to understand it, I still havent seen
you show
> me why we competitive players shouldnt object.
>
Then you need to listen to what I am saying better, because I am giving
you very good factual reasons, and you ignore them.
li...@ork.net
>> Yes I am angry. I am angry that you are trying to feed me s**t and
>> tell me its ice cream.
> Then you'd better pay more attention, because you still don't
> understand the situation.
> Do you really think that, for example, the highly educated people on
> the Ratings Committee would spend years of analysis, and fight years of
> political battles, and push and push and push to get a rating system
> change implemented if they didn't feel it enhanced the integrity of the
> system? Have you ever listened to these people talk about the
> integrity of the system? They view it as sacrosanct. People
> consistently complain that they won't loosen up their view a little
> bit. Your complaint is so far out of left field, its incredible.
More twaddle from Kevin the Caveboy.
Give the history at the USCF, and the well-established shortsightedness of
folks there, one would (and should) conclude exactly the opposite.
Chris
action. Since Ablue has a theory about gaining 6 points, perhaps he could
tell us his rating (or the rating that was used in this theoretical
example) and we could see it worked out? Anyone else?
Seems to me that some real world examples will be more informative than any
rhetoric.
cavema...@my-deja.com
I looked at a 1900 playing in a tournament immediately after the
conversion to the new system. If he scored 4.5 out of 9 against
2300's, (and assuming I did all the calcs correctly) he would gain
65.43 points PLUS 17.43 bonus points for a total of 82.86.
For a fully established player, the 1900 could expect close to 6 bonus
points rather than 17.43. The number of baseline points gained would
also be a little lower.
A player with K=32 would have gained about 117 points.
Note that the result would be different for a 1400 playing against a
1900. The K factor calculation is determined in part on number of
games played, and also in part based on rating. This is because
players of lower ratings have greater volatility.
The correct way to understand examples is to look at numerous
tournaments over a period of time. Variability is a two-way street,
and only by looking at several tournaments can one get the right
feeling for the overall trend of the rating, as well as its variability.
Quite frankly, an 1900 who is well-established isn't going to suddenly
score 50% against 2300's.
ABlue892
>conversion to the new system. If he scored 4.5 out of 9 against
>2300's, (and assuming I did all the calcs correctly) he would gain
>65.43 points PLUS 17.43 bonus points for a total of 82.86.
>
>For a fully established player, the 1900 could expect close to 6 bonus
>points rather than 17.43. The number of baseline points gained would
>also be a little lower.
More doublespeak from Caveman. He begins by suggesting in a loud voice that I
am wrong by using, I guess, a provisional player, then, presumably hoping
people will stop reading, admits I am right and 6 points is the actual gain.
>Quite frankly, an 1900 who is well-established isn't going to suddenly
>score 50% against 2300's.
>--
YES!! Do you even pay attention to what you write? This is my point. You are
trying to tell me that bonus points are some great boon, and I am trying to
explain that even if a 1900 had some outrageously strong performance, they
would only gain a few points. Is the light of the penalties of the new system
starting to dawn on you? And dont even give me that nonsense about Rating
points not being thought of as a "prize" or more accurately, a reward for good
performance. If people didnt think of it that way, they wouldnt complain, and
the USCF changed the system.
Go ahead, respond by saying "HA HA you are so stupid, the exact opposite of
everything you say is true, you should love the system" and then present your
arguments, which, besides ignoring the issue that I HAVE stated the facts of
the new system correctly, are utterly non-compelling, and I challenge you to
find someone who thinks they are.
Chris
>More doublespeak from Caveman. He begins by suggesting in a loud voice
>that I am wrong by using, I guess, a provisional player, then,
>presumably hoping people will stop reading, admits I am right and 6
>points is the actual gain.
My understanding is that you gain 6 instead of 17 *bonus points*, but this
is on top of the 65.43 points that you gain anyway. So, instead of 82
points you gain 71. Hardly seems like a tragedy of the proportions you
describe.
Todd Durham
cavema...@my-deja.com wrote:
>
> In article <20001221013929...@ng-fj1.aol.com>,
> ablu...@aol.com (ABlue892) wrote:
> > >perhaps we should call it ELO ME --
> > >will be used.
> >
> > for ELO MEssed up? God...why couldnt they leave well enough alone?
> Dont they
> > get enough complaints?
> >
> Yes. And the complaints they were getting were about people being
> underrated. And testing showed that yes, much (but not all) of the
> rating population is underrated (There is a segment that is actually
> overrated)
Pardon, but I'm coming to this discussion a little late. How did the
USCF determine that a significant proportion of the rated population is
underrated? Is there a significant group that are regularly scoring
above their expected performances? If so, why aren't their rating
reaching a new equilibrium point? (Sorry for the bad terminology, but I
hope my questions are clear.)
Todd
>
> So they changed the system to raise ratings, and to raise them more
> quickly. You misunderstand that, and complain that it is harder for
> you to gain points.
>
> I'm uncertain how to respond to that complaint.
ABlue892
>is on top of the 65.43 points that you gain anyway. So, instead of 82
>points you gain 71. Hardly seems like a tragedy of the proportions you
>describe.
Wrong..this is exactly the point. The USCF intends to confuse you so you don't
see the proportions I describe. Under the old system, you would gain ONE
HUNDRED AND THIRTY-FIVE POINTS in that scenario instead of the 71. Now, does
it seem like more of a tragedy? Keep reading Caveman's posts and watch him
explain it away.
Chris
>you don't see the proportions I describe. Under the old system, you
>would gain ONE HUNDRED AND THIRTY-FIVE POINTS in that scenario instead
>of the 71. Now, does it seem like more of a tragedy? Keep reading
>Caveman's posts and watch him explain it away.
Certainly seems like a bigger deal. I would simply like to see these
comparisons:
1900 player scores 4.5/9 against 2300 opposition under OLD rules
1900 player scores 4.5/9 against 2300 opposition under NEW rules
1900 player scores 6/9 against 2100 opposition under OLD rules
1900 player scores 6/9 against 2100 opposition under NEWS rules
two sets of data for each, assuming in one that player is proivisional and
in another that he/she is established.
Louis Blair
Chris wrote:
> I would simply like to see these
> comparisons:
>
> 1900 player scores 4.5/9 against 2300 opposition under OLD rules
> 1900 player scores 4.5/9 against 2300 opposition under NEW rules
>
> 1900 player scores 6/9 against 2100 opposition under OLD rules
> 1900 player scores 6/9 against 2100 opposition under NEWS rules
>
> two sets of data for each, assuming in one that player is proivisional and
> in another that he/she is established.
This seems to me to be the sort of approach
to this subject that is most likely to result
in useful information. However, I wonder how
common it is for a player with a 1900 rating
to score 4.5/9 against 2300 opposition or
6/9 against 2100 opposition. It seems to
me that it would be desirable to have a
table of comparisons that gives rating
point gains for a 1900 rated player scoring
4.5/9 against 2300 opposition, 2250
opposition, etc. all the way down to 1950
opposition.
Chris
>me that it would be desirable to have a
>table of comparisons that gives rating
>point gains for a 1900 rated player scoring
>4.5/9 against 2300 opposition, 2250
>opposition, etc. all the way down to 1950
>opposition.
I agree with this. I asked for a more limited set just because of the
question at hand and because I hated to ask others to pay too much for my
lack of math skills and knowledge of the rating systems...
But that set would make even more sense.
Herbert Kanner
(Chris) wrote:
Some of this discussion seems pointless, the writers missing the
essential fact that a chess rating is supposed to be a measure of the
skill of a player and is supposed to be accurately (within reason)
predictive of the outcome of a series of games between that player and
rated opponents. Any perversion of the system which arbitrarily awards
bonus points will subvert this objective. The only adjustments to the
system which maintain the original intent (Prof. Elo's intent) are those
which alter the weighting factor (the term usually called "K") so as to
speed up the approach of a rating to its equilibrium value.
Herb
--
In addressing email, replace "acmd" by "acm".
--
Herbert Kanner <kan...@acmd.org>
Chris
>Some of this discussion seems pointless, the writers missing the
>essential fact that a chess rating is supposed to be a measure of the
>skill of a player and is supposed to be accurately (within reason)
>predictive of the outcome of a series of games between that player and
>rated opponents. Any perversion of the system which arbitrarily awards
>bonus points will subvert this objective. The only adjustments to the
>system which maintain the original intent (Prof. Elo's intent) are those
>which alter the weighting factor (the term usually called "K") so as to
>speed up the approach of a rating to its equilibrium value.
If this is the case, then it seems equally clear that the rating system is
evolving into something more than just Elo's intent.
cavema...@my-deja.com
Moreover, I did not include the 18 activity points in each situation,
which are added on top of the above. AND, when correctly taken in
conjunction with the impact of lower k-factors on less stupendous
performances, it makes all the sense in the world.
cavema...@my-deja.com
ablu...@aol.com (ABlue892) wrote:
> >I looked at a 1900 playing in a tournament immediately after the
> >conversion to the new system. If he scored 4.5 out of 9 against
> >2300's, (and assuming I did all the calcs correctly) he would gain
> >65.43 points PLUS 17.43 bonus points for a total of 82.86.
> >
> >For a fully established player, the 1900 could expect close to 6
bonus
> >points rather than 17.43. The number of baseline points gained would
> >also be a little lower.
>
> More doublespeak from Caveman. He begins by suggesting in a loud
voice that I
> am wrong by using, I guess, a provisional player, then, presumably
hoping
> people will stop reading, admits I am right and 6 points is the
actual gain.
No. The first playerIS noT provisional, and I did not suggest that.
He is a player IMMEDIATELY AFTER the change to the new system.
You are wrong, yet again.
Also, you never provided a player's rating, I did. We still can't tell
if you were correct or not.
>
> >Quite frankly, an 1900 who is well-established isn't going to
suddenly
> >score 50% against 2300's.
> >--
>
> YES!! Do you even pay attention to what you write? This is my point.
You are
> trying to tell me that bonus points are some great boon, and I am
trying to
> explain that even if a 1900 had some outrageously strong performance,
they
> would only gain a few points.
Yes. Now YOU try listening. K-factors are lower. 6 points ARE
meaningful for a single result like this. Now YOU get it.
You keep arguing that you want' more devalued points. If I give you
more points, and each means less, I say What??
>Is the light of the penalties of the new system
> starting to dawn on you?
No, because there aren't any.
>And dont even give me that nonsense about Rating
> points not being thought of as a "prize" or more accurately, a reward
for good
> performance. If people didnt think of it that way, they wouldnt
complain, and
> the USCF changed the system
The Rating Committee didn't change the system because of complaints.
They examined the system because of complaints. They changed it
because the complaints were valid.
Your complaints are not valid.
> Go ahead, respond by saying "HA HA you are so stupid, the exact
opposite of
> everything you say is true, you should love the system"
OK, well, you've got that part right.
>and then present your
> arguments, which, besides ignoring the issue that I HAVE stated the
facts of
> the new system correctly,
No, I didn't igore that you claim to have stated the facts correctly.
I said that you were wrong in your claim. I still say that you are
wrong in your claim.
> are utterly non-compelling, and I challenge you to
> find someone who thinks they are.
The entire Ratings Committee. Most of the rated population of USCF
that has seen comments on this site have responded postively.
cavema...@my-deja.com
Now consider the other half of the argument, which you keep ignoring.
First of all, a new player during the next year, would gain a total of
100 points.
They would also have less negative volatility for a poor result.
Now,you keep wanting to turn this into some psycobabble argument on how
people feel about points. It's not.
The fact is that higher rated players have less rating fluctuation than
lower rated players.
So the system takes that into account, and when you have the random
excellent result, or the random horrid result, it neither moves you up,
nor down, as far as it did in the past, because in either case it would
have been adjusting your rating too much.
Again, you seem to think of points as a reward. They are not. Points
have a strict predictive capability, and moving a player up too far, or
down too far, diminishes that.
cavema...@my-deja.com
Under this system, there are degrees of provisional and degrees of
established, so your question isn't exact enough. One would need more
info to do the comparisons your request.
However...
The comparisons you are asking for are meaningless. They assume, at
the heart of them, that the OLD rules produced a result that was more
meaningful. What you really need to do is to measure the real
volatility of players at various levels, and then compare various types
of results with the degree of volatility expressed. When that is done,
you will find, as did the Rating Committee in the analysis already
done, that the degree of volatility for high rated players was too
great.
cavema...@my-deja.com
Louis Blair <lbl...@pilot.lsus.edu> wrote:
>
>
> Chris wrote:
>
> > I would simply like to see these
> > comparisons:
> >
> > 1900 player scores 4.5/9 against 2300 opposition under OLD rules
> > 1900 player scores 4.5/9 against 2300 opposition under NEW rules
> >
> > 1900 player scores 6/9 against 2100 opposition under OLD rules
> > 1900 player scores 6/9 against 2100 opposition under NEWS rules
> >
> > two sets of data for each, assuming in one that player is
proivisional and
> > in another that he/she is established.
>
> This seems to me to be the sort of approach
> to this subject that is most likely to result
> in useful information.
Actually, it provides NO useful information. The question is, if you
measure the actual volatility of players, how do the situations
described under the old and new rating systems compare to the actual
change.
The old system causes swings that are too big.
> However, I wonder how
> common it is for a player with a 1900 rating
> to score 4.5/9 against 2300 opposition or
> 6/9 against 2100 opposition. It seems to
> me that it would be desirable to have a
> table of comparisons that gives rating
> point gains for a 1900 rated player scoring
> 4.5/9 against 2300 opposition, 2250
> opposition, etc. all the way down to 1950
> opposition.
>
>
Picking one 50% score is ridiculous. You need to see how the point
swings work versus actual performance.
cavema...@my-deja.com
It isn't. The new system is actually CLOSER to Elo's intent. See his
book, "The Rating of Chessplayers", chapter 3.
cavema...@my-deja.com
Your question is like this:
Take a light blue paint (the old system)
Take a dark blue paint.
Compare them.
What did you find out?
The dark blue paint is darker.
Now, that may seem like useful information,
Unfortunately, it tells you nothing useful. The question you want to
answer is which paint is closer in color to that used on a New York
Giant uniform.
You need to compare each system AGAINST REALITY. This has already been
done, which is why the change is being made to the new system.
cavema...@my-deja.com
Herbert Kanner <kan...@acmd.org> wrote:
> In article <90168FA6...@207.126.101.100>, fn...@my-deja.com
> (Chris) wrote:
>
> Some of this discussion seems pointless, the writers missing the
> essential fact that a chess rating is supposed to be a measure of the
> skill of a player and is supposed to be accurately (within reason)
> predictive of the outcome of a series of games between that player
and
> rated opponents.
Agreed.
>Any perversion of the system which arbitrarily awards
> bonus points will subvert this objective. The only adjustments to
the
> system which maintain the original intent (Prof. Elo's intent) are
those
> which alter the weighting factor (the term usually called "K") so as
to
> speed up the approach of a rating to its equilibrium value.
>
Your second paragraph is false. Elo himself wrote that bonus points
are a key way to combat rating deflation, which is a natural tendency
of the system.
The new system does not "arbitrarily" award bonus points, although
activity points are somewhat arbitrary.
> Herb
>
> --
>
> In addressing email, replace "acmd" by "acm".
> --
> Herbert Kanner <kan...@acmd.org>
>
--
cavema...@my-deja.com
Todd Durham <tdu...@mdo.net> wrote:
> Pardon, but I'm coming to this discussion a little late. How did the
> USCF determine that a significant proportion of the rated population
is
> underrated? Is there a significant group that are regularly scoring
> above their expected performances? If so, why aren't their rating
> reaching a new equilibrium point? (Sorry for the bad terminology, but
I
> hope my questions are clear.)
>
> Todd
>
It was determined by using statistical measurement of various segments
of the rating pool.
Without injecting points into the system, there cannot be equilibrium.
Rating systems like the Elo system naturally deflate over time.
ABlue892
>have a strict predictive capability, and moving a player up too far, or
>down too far, diminishes that.
This is such a load. At the heart of it, you and I both know that Rating
deflation/Inflation, is an invisible monster. There is no such thing. Rating
is a measure of strength of players in relation to each other. Therefore, if
everyone is 200 points "underrated" the system is perfect. People are exactly
where they are in relation to each other. The bottom line is that anyone would
rather be an 1800 in a world where the average player is 1900 than be a 1550 in
a world where the average is 1400. This is something you may not realize in
your ivory tower of Life Masters, but come out of the cave and meet some people
who have more average ratings, and you will find this to be the case.
cavema...@my-deja.com
ablu...@aol.com (ABlue892) wrote:
> >Again, you seem to think of points as a reward. They are not.
Points
> >have a strict predictive capability, and moving a player up too far,
or
> >down too far, diminishes that.
>
> This is such a load. At the heart of it, you and I both know that
Rating
> deflation/Inflation, is an invisible monster. There is no such thing.
This is just wrong. There is such a thing. It is measurable, and it's
existance can be easily demonstrated. If you are willing, I will walk
you through a simple example that shows it exists.
Arpad Elo, the designer of the system, knew that deflation was a major
issue for the system, and included potential remedies in its design.
The changes made in the system are for the most part the very remedies
he describes.
> Rating
> is a measure of strength of players in relation to each other.
This is a false statement, in that it is only half true. It is
desirable for ratings to have roughly the same meaning over time.
Someone who is 1500 today, who does not have a change in performance,
should be 1500 tomorrow.
>Therefore, if
> everyone is 200 points "underrated" the system is perfect. People
are exactly
> where they are in relation to each other.
In relation to each other yes. But what about in relation to the
ratings? Do you want 2200, which means master, to eventually mean
Senior Master, and 2000, which means expert, to be master? And do you
want class A to be expert? Do you want to constantly shift the titles
around to new numbers, or would you rather just make the ratings
accurate?
> The bottom line is that anyone would
> rather be an 1800 in a world where the average player is 1900 than be
a 1550 in
> a world where the average is 1400.
No, that is not the bottom line. People only think that way
because "1400" has a certain connotation and "1800" has a certain
connotion. Deflation changes what the numbers mean.
> This is something you may not realize in
> your ivory tower of Life Masters, but come out of the cave and meet
some people
> who have more average ratings, and you will find this to be the case.
>
Your argument has been considered and rejected for about 35 years.
Come into the light and learn what ratings are all about.
If you would like an example, I will construct one for you.
Louis Blair
> The only adjustments to the
> system which maintain the original intent (Prof. Elo's intent) are those
> which alter the weighting factor (the term usually called "K") so as to
> speed up the approach of a rating to its equilibrium value.
Chris wrote:
> If this is the case, then it seems equally clear that the rating system is
> evolving into something more than just Elo's intent.
Kevin Bachler wrote:
> Elo
> recognized that the inherent flaw of this system is that it is
> naturally deflationery, and he proposed several tools to combat this --
> many of which are part of the changes that [ABlue892 is]
> complaining about. ...
> Elo devotes an entire chapter (chapter 3) to Rating
> Administration. In 3.6 he discusses the management of deflation, and
> in section 3.7 he discusses various processes to managed deflation.
> Several of these ideas have been incorporated into the new rating
> system, although not necessarily in exactly the form suggested by Elo.
> These included: 3.71 ... 3.72 ... 3.73 ... 3.74 ... 3.75 ...
> Those were the 5 ideas Elo suggested to combat deflation, i. e. to keep
> ratings up where they belong. ALL 5 ARE INCORPORATED IN THE NEW SYSTEM.
> In addition, a 6th idea, Activity points, were added for 1 year.
Louis Blair
> I would simply like to see these
> comparisons:
>
> 1900 player scores 4.5/9 against 2300 opposition under OLD rules
> 1900 player scores 4.5/9 against 2300 opposition under NEW rules
>
> 1900 player scores 6/9 against 2100 opposition under OLD rules
> 1900 player scores 6/9 against 2100 opposition under NEWS rules
>
> two sets of data for each, assuming in one that player is proivisional and
> in another that he/she is established.
I wrote:
> This seems to me to be the sort of approach
> to this subject that is most likely to result
> in useful information. However, I wonder how
> common it is for a player with a 1900 rating
> to score 4.5/9 against 2300 opposition or
> 6/9 against 2100 opposition. It seems to
> me that it would be desirable to have a
> table of comparisons that gives rating
> point gains for a 1900 rated player scoring
> 4.5/9 against 2300 opposition, 2250
> opposition, etc. all the way down to 1950
> opposition.
Chris wrote:
> I agree with this. ...
> that set would make even more sense.
Kevin Bachler writes:
> Your question is like this:
> Take a light blue paint (the old system)
> Take a dark blue paint.
> Compare them.
> What did you find out?
> The dark blue paint is darker.
>
> Now, that may seem like useful information,
> Unfortunately, it tells you nothing useful. The question you want to
> answer is which paint is closer in color to that used on a New York
> Giant uniform.
>
> You need to compare each system AGAINST REALITY. This has already been
> done, which is why the change is being made to the new system
If Kevin Bachler wants people here to understand
and believe that the USCF has chosen the paint that
is closest to the color used on a New York Giant
uniform, then it seems to me that one of the steps
in that process is to help them to perceive the paints
that are being considered.
Kevin Bachler writes:
> This ... sort of approach ...
> provides NO useful information.
Repeatedly it has been stated that critics of the
new system have not properly calculated how
many points a player would receive under the
new system. If a person who thoroughly
understood the system did some sample
calculations of the sort described above, it
might help the discussion to proceed on the
basis of a more accurate perception of what
the new system does.
Kevin Bachler writes:
> The question is, if you
> measure the actual volatility of players, how do the situations
> described under the old and new rating systems compare to the actual
> change.
>
> The old system causes swings that are too big.
It seems to me that if people are to perceive this,
it would help if they have some clear examples of
the "swings" that are produced by the old and new
systems.
Kevin Bachler writes:
> Under this system, there are degrees of provisional and degrees of
> established, so your question isn't exact enough. One would need more
> info to do the comparisons your request.
Then I would suggest doing the comparison only
for the case of players with ratings that are fully
established.
Kevin Bachler writes:
> The comparisons you are asking for are meaningless. They assume, at
> the heart of them, that the OLD rules produced a result that was more
> meaningful.
I do not see how any such assumption is implicit
in the data being requested.
Louis Blair
> Again, you seem to think of points as a reward. They are not. Points
> have a strict predictive capability, and moving a player up too far, or
> down too far, diminishes that.
ABlue892 wrote:
> This is such a load. At the heart of it, you and I both know that Rating
> deflation/Inflation, is an invisible monster.
I could be wrong, but I do not think that Kevin Bachler
was talking about rating deflation/inflation in the above
quote. I think he was referring to the claim that, under
the old system, an individual player might get too much
of a rating boost from a particular tournament result.
Perhaps, under the old system, most 1900 players who
suddenly gained 100 points in one tournament (for
example) might find that their rating went most of the
way back down after a few more tournaments. If
such were the case, might not the predictive ability
of the rating system be improved if the ratings of
such players only went up say ... 50 points ... in the
first place. I don't know if these ideas are valid,
but it seems to me that the issue is separate from
the deflation/inflation issue.
ABlue892 wrote:
> anyone would
> rather be an 1800 in a world where the average player is 1900 than be a 1550 in
> a world where the average is 1400.
I would think that I would be happier 150 points
above average (if only it were true) rather than
100 points below average. I must be missing
something in this argument.
cavema...@my-deja.com
> If Kevin Bachler wants people here to understand
> and believe that the USCF has chosen the paint that
> is closest to the color used on a New York Giant
> uniform, then it seems to me that one of the steps
> in that process is to help them to perceive the paints
> that are being considered.
>
comparison, then they must first demonstrate a sufficient knowledge of
statistical theory for it to make sense. Otherwise the best that can
be done is analogy and description.
> Kevin Bachler writes:
>
> > This ... sort of approach ...
> > provides NO useful information.
>
> Repeatedly it has been stated that critics of the
> new system have not properly calculated how
> many points a player would receive under the
> new system. If a person who thoroughly
> understood the system did some sample
> calculations of the sort described above, it
> might help the discussion to proceed on the
> basis of a more accurate perception of what
> the new system does.
>
It has already been done and the results were already made known, by
the Rating Committee.
> Kevin Bachler writes:
>
> > The question is, if you
> > measure the actual volatility of players, how do the situations
> > described under the old and new rating systems compare to the actual
> > change.
> >
> > The old system causes swings that are too big.
>
> It seems to me that if people are to perceive this,
> it would help if they have some clear examples of
> the "swings" that are produced by the old and new
> systems.
>
In the type of discussion that will follow, it will be difficult for
people to perceive it to this level without a firm statistical
underpinning.
> Kevin Bachler writes:
>
> > Under this system, there are degrees of provisional and degrees of
> > established, so your question isn't exact enough. One would need
more
> > info to do the comparisons your request.
>
> Then I would suggest doing the comparison only
> for the case of players with ratings that are fully
> established.
>
> Kevin Bachler writes:
>
> > The comparisons you are asking for are meaningless. They assume, at
> > the heart of them, that the OLD rules produced a result that was
more
> > meaningful.
>
> I do not see how any such assumption is implicit
> in the data being requested.
>
>
That's my point.
Suppose you get the comparison that you ask for. What do you think it
tells you if the new system produces more points than the old? What do
you think it tells you if the old system produces more points than the
new?
ABlue892
>
>> anyone would
>> rather be an 1800 in a world where the average player is 1900 than be a
>1550 in
>> a world where the average is 1400.
>
>I would think that I would be happier 150 points
>above average (if only it were true) rather than
>100 points below average. I must be missing
>something in this argument.
The point I was trying to make is that I think people in general get too caught
up in the numbers, and THIS is the reason for the rating change. I agree,
logically someone would want to be above average, but I get the sense that most
people are attracted to the higher number. When put in such a bold face as I
presented it, people would see the illogic of it and say "oh, of course I would
want to be 1550" but in real life, when most people are not aware of or only
vaguely took note of what the average is, they seem to just want that
attractive high number.
cavema...@my-deja.com
> The point I was trying to make is that I think people in general get
too caught
> up in the numbers,
EXCUSE ME?? I can't think of ANY poster that has been more caught up
in the numbers (and misunderstanding them) than you. It is you who
raised the issue that if you win game after game after game that you
will gain points more slowly than in the past -- irrespective of the
fact that the likelihood of this event is miniscule, and unimportant
versus the improvement of the entire system
> and THIS is the reason for the rating change.
Uh, no. The reason for the rating change is that the system was
inaccurately measuring the performance of the players.
There have been system changes in the past, whether the players wanted
them or not.
ABlue892
>them or not.
Yes, this is your position, and you have made it clear. "This is the change,
it doesnt hurt me, we don't care if it hurts the players, and if you don't like
it you can take up checkers." Presumably your feeling represents the general
feeling of the USCF voting board. Thanks for clearing this up.
cavema...@my-deja.com
ablu...@aol.com (ABlue892) wrote:
> >There have been system changes in the past, whether the players
wanted
> >them or not.
>
> Yes, this is your position,
Please, stop declaring my position, and then writing things other than
what I said. The below IS NOT my position. It is your straw man
version of my position.
>and you have made it clear. "This is the change,
Yes.
> it doesnt hurt me,
NO.
>we don't care if it hurts the players,
No.
>and if you don't like
> it you can take up checkers."
No.
> Presumably your feeling represents the general
> feeling of the USCF voting board. Thanks for clearing this up.
Apparently, I didn't clear it up, or you are just so hateful that you
refuse to listen to the real position.
The change in rating system helps most players immediately, and the
entire pool in the longer term. In the short term some players may
be "harmed", but the addition of activity points provides a means to
more than mitigate that harm. Consequently, there is no real harm.
Whenever a rating system is changed, some people will undergo temporary
negative effects, that is unavoidable.
I did not say that I am unharmed. I said that right now decreasing the
negative volatility is not a boon to me.
Please get it straight.
ABlue892
>
>--
>Kevin Bachler
>Caveman
>> >There have been system changes in the past, whether the players
>wanted
>> >them or not.
>>
>> Yes, this is your position,
>
>Please, stop declaring my position, and then writing things other than
>what I said.
Ok, I actually copied that from your post, so I don't think I could have
written other than what you said.
>> it doesnt hurt me,
>
>NO.
How does it hurt you? By inflating the ratings of your weak scholastic players
so you look like a Chess teaching guru? By NOT causing you to lose your
Master status no matter what your results are under the new system? Or by
making it harder for other players to achieve master status, making your Master
title more valuable for prestige and attracting new students?
>>and if you don't like
>> it you can take up checkers."
>
Are you claiming you did NOT tell me that if I didn't like it I should take up
checkers? That's interesting.
Tell me what I did not get straight, since my "facts" are so suspect.
cavema...@my-deja.com
ablu...@aol.com (ABlue892) wrote:
> >Please get it straight.
> >
> >--
> >Kevin Bachler
> >Caveman
>
> >> >There have been system changes in the past, whether the players
> >wanted
> >> >them or not.
> >>
> >> Yes, this is your position,
> >
> >Please, stop declaring my position, and then writing things other
than
> >what I said.
>
> Ok, I actually copied that from your post, so I don't think I could
have
> written other than what you said.
Copied out of context.
>
> >> it doesnt hurt me,
> >
> >NO.
>
> How does it hurt you? By inflating the ratings of your weak
scholastic players
> so you look like a Chess teaching guru?
The sarcasm is idiotic at best. Here, try this. My chess teams have
won 6 state championships in the past 6 years, and 3 national
championships in the past 4 years. My students have won several
individual state championships (I think 5) during that period.
Do you think I need ratings to show that I am a good instructor?
The lack of downward volatility doesn't do anything for me right now.
Consequently the decrease in k-factor in an upward direction actually
hurts me MORE than it does you.
> By NOT causing you to lose your
> Master status no matter what your results are under the new system?
I am a life master under either system. Your statement, which I assume
was an attempt at sarcasm, is pathetic.
>Or by
> making it harder for other players to achieve master status, making
your Master
> title more valuable for prestige and attracting new students?
It doesn't make it harder for other players to achieve master status.
Arguably, it may be easier under the new system.
> Are you claiming you did NOT tell me that if I didn't like it I
should take up
> checkers? That's interesting.
No, what I claimed is that you don't understand sarcasm. Your comments
above bear that out.
>
> Tell me what I did not get straight, since my "facts" are so suspect.
Well, for example, you still think it will be harder for you to become
a master. That's wrong.