MatPlus.Net

 Website founded by
Milan Velimirović
in 2006

16:42 UTC
ISC 2024
 
  Forum*
 
 
 
 

Username:

Password:

Remember me

 
Forgot your
password?
Click here!
SIGN IN
to create your account if you don't already have one.
CHESS
SOLVING

Tournaments
Rating lists
1-Jan-2024

B P C F





 
 
MatPlus.Net Forum General number of album points and title: how were these determined? and is there a function(curve)...
 
You can only view this page!
Page: [Previous] [Next] 1 2
(1) Posted by Eugene Rosner [Sunday, Aug 25, 2013 16:10]

number of album points and title: how were these determined? and is there a function(curve)...


that can relates titles(master-12,international master-25, grandmaster-70)with their corresponding Elo ratings (2200,2400,2500, such that numbers below 12, in between 12-70, and beyond 70 would make sense?
 
(Read Only)pid=10800
(2) Posted by Steven Dowd [Monday, Aug 26, 2013 01:30]

My goodness, of course not. But I do know how many angels can dance on the head of a pin. And the length of the Kaiser's beard.
 
 
(Read Only)pid=10807
(3) Posted by Kevin Begley [Monday, Aug 26, 2013 01:47]

@Steven,

I presume the Angel moves like an Amazon (R+B+N combo)...
And, I suppose "dance" must refer to the criteria from the famous 8-queens problem...
But, I'm unfamiliar with a board shape called, "the head of a pin."
Help me out. :-)
 
   
(Read Only)pid=10808
(4) Posted by Joaquim Crusats [Monday, Aug 26, 2013 08:34]; edited by Joaquim Crusats [13-08-26]

Maybe Steven was referring to Syadristy's chess board in Kiev: http://microart.kiev.ua/en/shahy.html

Edit: I should have used an "emoticon" so as not to disappoint Hauke (next post), but then I would have disappointed myself. In any case, the meaning of both sentences in Steven's post is obvious.
 
   
(Read Only)pid=10809
(5) Posted by Hauke Reddmann [Monday, Aug 26, 2013 16:46]

Ah, I'm always disappointed when problemists, the elite of the culture :-),
can't source an obvious gag. The "angels dancing on a pin" refer to the
sophist tradition in the Medieval of heatedly chatter about completely
pointless questions. (Whether there *actually existed* any sophists who
argued about *this question* would need incessant googling and wikiing,
and I'm too lazy :P - in any case the pindancing angels became proverbial.)

The answer to the original question *is* uncannily alike to the angel
problem, when seen from a mathematic point of view. :-) Shall the angel,
eh, the function be monotonous and infinitely differentiable? Sounds
reasonable (although only integer-values are needed). Shall it be a
n-degree polynomial? Maybe even linear? Given that the ELO system assumes
a Gaussian distribution of hemi-demi-semi-mathbabblian...

I better stop :-)

Hauke
 
   
(Read Only)pid=10810
(6) Posted by Kevin Begley [Monday, Aug 26, 2013 18:23]; edited by Kevin Begley [13-08-26]

@Hauke,

The cultural super elite never presume that an obvious reference has been missed.
For example, I presume that you must have noticed the emoticon, before you let fly the disappointment boomerang.

@Joaquim,
Thanks for sharing that amusing link!
 
   
(Read Only)pid=10811
(7) Posted by Steven Dowd [Monday, Aug 26, 2013 23:01]

The Elo system, well-intentioned as it is, has already made much discussion of OTB chess meaningless. Instead of "Did you see how he won the pawn ending in game x?," discussion on boards such as chessgames.com is always about how many Elo points are won or lost, tournament performance ratings (we even have the handy abbreviation TPR), and other similar nonsense. It's to the point where some will even use decimals, as in, "Wesley So gained 6.78 points in his...."

I mean poor Eugene no harm (he's probably just trying to explain to some unenlightened soul what it means to be a grandmaster in problem chess), but in my opinion, anyone who attempts such a further discussion of quantification of chess problem artists should be whisked off to some deserted island, and never heard from again.

I also consider emoticons unnecessary when such tongue-in-cheek statements as the above are made, but that will probably get me in trouble.
 
   
(Read Only)pid=10812
(8) Posted by Michal Dragoun [Tuesday, Aug 27, 2013 00:41]

Concerning the very first question (if serious meaning is asked) - titles cannot be compared to ELO in longer period. Titles mean "sometimes (forty years ago or yesterday) I played (composed) so good, that I was awarded by [some] title". ELO means "now I am playing as a patzer (or GM)". And there are no ways, how to measure something like "actual composing strength".
 
   
(Read Only)pid=10813
(9) Posted by Eugene Rosner [Tuesday, Aug 27, 2013 14:55]

no offense intended folks, I was just curious if a correlation could be made between these two groups. I was also curious as to what went into choosing the particular numbers for the composing titles...
 
   
(Read Only)pid=10816
(10) Posted by Steven Dowd [Tuesday, Aug 27, 2013 16:19]; edited by Steven Dowd [13-08-27]

Hey Eugene,

I didn't take any offense to your question, and am glad you didn't to our "answers."

But quantification of composing performance, and a cross-comparison to OTB strength, ouch. I can tell you what I think - that I am about as good at composing as an OTB FM is in over the board play, but I'll probably never perform well enough to earn an FM title in composition. (You can take that as a brag, but I think it is objective, although my American colleagues must disagree since I never even earned the national title of "expert," let alone "master," by these Good Companions)

What I mean by the above is that composition titles are much harder to come by, and typically demand sustained performance over a good number of years to earn them. You can become an OTB FM by simply performing well over 25 games. Some very good composers, much better than me, will probably never earn the FM title even with 25 *years* of composition activity.

If someone wanted to make some kind of comparison, they should look at the number of composers over the years and compare that with the titles awarded, as well as the number of players and titles awarded over a similar time frame. It wouldn't prove a thing, but it would still be interesting.
 
   
(Read Only)pid=10817
(11) Posted by seetharaman kalyan [Tuesday, Aug 27, 2013 20:26]

Interesting... So titles like 'Master' 'Expert' are awarded to US composers.. ? What is the criterion? I think the question of Mr.Eugene was also this !
 
   
(Read Only)pid=10818
(12) Posted by Kevin Begley [Tuesday, Aug 27, 2013 22:07]; edited by Kevin Begley [13-08-27]

US titles are awarded purely on the basis of popularity, at Good Companion Meetings (which occur very rarely).
In fact, you can call for a Good Companions Meeting, stockpile it with friends and family (who need know nothing about problems -- or even the rules of chess), and vote yourself US Grandmaster (of Composition, and Solving).

Long ago, I was awarded the US Master title, in composing.
Later, I felt obligated to reject this title.

In hindsight, it is reasonable to conclude that my award was expedited by two factors:
1) I had complained about how the US Expert title was awarded to a young composer (at the time, his only trumpeted work was a blatant plagiarization).
In his defense, this "US Expert" was too young (and inexperienced) to know that his minor alteration of a classic problem was insufficient to confer originality -- which begs the question: need a US Expert know anything?
2) I was being considered for StrateGems Fairy Editor (which greatly expedited the title process, for the benefit of the journal).

In fact, at the same time, they offered another title (US Master of Solving), to a worthy solver, who was being considered as Retro/PG editor.
That individual smartly rejected their offer, immediately -- arguing that they had no authority to give anyone a boost up the solving ladder.

I was troubled by this, and consulted Dan Meinking about whether I should do the same.
On the basis of Dan's assurances (who remembered clearly having achieved a comparable success, when he was awarded this title), I accepted the award.
It took me several months to realize that I was wrong to accept this award -- and I (very respectfully) rejected the title, under protest of the mechanism.
Furthermore, I petitioned the Good Companions to implement an objective criteria for titles (and, only then, reconsider me for this award).
Not surprisingly, my petition found no support (civil protest never succeeds in problem chess -- which is why problem chess never improves).

US titles say plenty about the GCM attendees (what subjective lengths they will endure, for ego appeasement and companionship); they say nearly nothing about a composer.

If they offer you a Master title, Steven, I would encourage you to reject it.
You are WAY better than any title based upon popularity.
A talented composer's primary obligation must be to objective truth (titles hold no value, except that conferred by the honesty of their process).

ps: had they implemented the objective criteria that I had suggested, you would have *automatically* been awarded the Master Title, many years ago (by now, maybe much more)...
 
   
(Read Only)pid=10819
(13) Posted by Steven Dowd [Wednesday, Aug 28, 2013 09:51]

Kevin said:

If they offer you a Master title, Steven, I would encourage you to reject it.

Hey Kevin, sinus season is here, when I often need to sequester myself for long periods of time (its not all that bad, sitting in the dark I can come up with some interesting - at least to me - problem ideas). However, I have to thank you for clearing out my sinuses, at least for the moment with this amusing thought - I could not stop laughing. We both know the sun will be a cold hard cinder before that group ever offers me a thing. I have a better chance of being invited to Miley Cyrus' birthday based on my reputation as a cool party dude.

If the Onion ever starts a chess problem corner, I am electing you as columnist.
 
   
(Read Only)pid=10823
(14) Posted by Kevin Begley [Wednesday, Aug 28, 2013 11:56]; edited by Kevin Begley [13-08-28]

The onion... no -- I'm too busy calculating the correlation between US Composing Titles and an objectively quantified facial resemblance to any unit found in (Norway's?) Lewis chess set...
Based upon limited data, it's looking like facial contortions toward the "Berserker" Rook are the most profitable.

But, watch out -- I have a feeling this one will be gaining: http://feastbowl.files.wordpress.com/2011/04/zoom-chessmen.jpg

ps: if you really want a US Title, read this kick-ass new problem chess book -- http://www.youtube.com/watch?v=QnvvBzuEg4k
 
   
(Read Only)pid=10824
(15) Posted by Steven Dowd [Wednesday, Aug 28, 2013 12:58]

Many possible comments on that Beserker, all would get me deeper in trouble.

I do regret hijacking Eugene's thread. But to provide at least some information for him, maybe people could try answering a set of questions like these:

1. What does it mean to be an OTB grandmaster?

2. What does it mean to be a composition grandmaster?

3. Why is (or isn't) it possible/wise to provide Elo-type ratings of chess composers?

4. Can the points earned in FIDE Albums be compared to the points earned in chess OTB competition? Why or why not?

I think it is unwise for us to dismiss his question(s) so easily if we care at all about public relations for our hobby.
 
   
(Read Only)pid=10825
(16) Posted by Kevin Begley [Wednesday, Aug 28, 2013 22:00]; edited by Kevin Begley [13-08-28]

>1. What does it mean to be an OTB grandmaster?
>2. What does it mean to be a composition grandmaster?

I would add:
2a. And, given the plausible answers to both questions 1 & 2, why do problemist still insist upon using incomparable titles, from chess competition (e.g., Grandmaster)?
read: are the dubious comparisons not an intentional distortion (to lure the chess player audience into presumptions of false equivalence)?

>3. Why is (or isn't) it possible/wise to provide Elo-type ratings of chess composers?

To be quite honest, if you are willing to accept subjective judgements (in terms of wins/loses), you can't so easily dismiss the honesty of this question.
The current method is akin to measuring composers based upon the number of "brilliancy prizes" won, over their entire career.
By such a measure, Alexei Shirov probably still outperforms Magnus Carlsen (and with prolonged endurance, and good health, he might even outperform Tal).
This fails as a mechanism to measure compositional skills -- because the measure is subordinate to the individual's health & longevity (along with countless other factors).

Rated thematic composing events, specifically intended to measure compositional skills, could eliminate a dependence upon longevity (and provide a title mechanism which is more compatible with chess titles).
The question is: how quick should they be, and to what degree does thematic selection impact competition?

>4. Can the points earned in FIDE Albums be compared to the points earned in chess OTB competition? Why or why not?

All good questions, but all of them ignore the most important question: Can we trust subjective judgement?
If you haven't already seen this video, check out Uri Avner's lecture in Marianka: http://www.youtube.com/watch?v=RHcJ1a3GXKs&feature=youtu.be

I applaud him for having the courage to touch upon this issue (even if he allows the audience to draw their own conclusions).
More careful introspection is needed, on the value of titles (and awards), given the limitations of subjective judgement.
I suspect problemists routinely avoid this subject, because it is in their interest to preserve some deliberative illusions about composition titles.
 
   
(Read Only)pid=10826
(17) Posted by Steven Dowd [Thursday, Aug 29, 2013 01:22]

I was lucky Kevin in that early on when I started composing Chris Feather did not brush me off in the way many others did and took the time to answer my persistent questions about composition, and this idea of trusting subjectivity was one he took on again and again - in fact the reason he self-published so many problems was to avoid this rather irritating "inability" of judges to reward the best problems - and I hope he will excuse me if I am not making the case as he would have.

But if we accept that chess problems are an art, we have to learn to accept this sort of misjudgment - great artists in other fields can go unrewarded until their deaths (at which point, who cares?) while some putting out trash are praised to the skies. Whether you are an actor, writer, painter, musician, and the list goes on - you can expect that it may simply come to pass that your talents will never be recognized to the extent that they should be. At least one talented composer I know of did not submit any problems to the last Album cycle, not because he doesn't believe in the process, but because he felt his problems would not survive the prejudices of the assigned judges.

I think you just have to accept these sort of things and muddle through the best you can, taking care that the work you turn out at least meets your own standards of quality. One thing that amuses me are those who write long articles showing why judge so-and-so messed up on not placing their problem higher in the awards, what they missed, and so on. That sort of hubris can only lead to stomach ulcers.

I also think this pertains to Eugene's original question in that you just have to accept a certain amount of subjectivity in the recognition and awarding of problems and titles - and that means any attempt at quantitative interdisciplinary comparison (OTB to composition) is bound to fail. Even intradisciplinary comparisons are fraught with danger - comparing say Franz Pachl to Christopher Jones is like comparing a landscape painter to a sculptor. Or consider, in closing, the rather fruitless discussion that appeared on why study composers should receive more points for their compositions.
 
   
(Read Only)pid=10829
(18) Posted by Hauke Reddmann [Friday, Aug 30, 2013 11:15]

While we're at numerology, can we throw in the *solving* ELO
in one fell swoop? :-) I participated in Bremen for the first time
and gee whiz, solved to about ~2280 which agrees almost perfectly
with my OTB ELO. Cf. also the article of Blatant X. Lies,
"The use of n=1 statistics in science", in Int. J. Humbug. :-)

Hauke
 
   
(Read Only)pid=10847
(19) Posted by Eugene Rosner [Tuesday, Aug 26, 2014 15:34]

so a quick follow up...and I'm basically interested in the mathematics here, I used an online power regression calculator and found the following curve to be smooth and somewhat close: y=1871x^07
this uses method of least squares. The question is, is it possible to do least squares on y=ax^b+c? using the data in the first post?
 
   
(Read Only)pid=12576
(20) Posted by Hauke Reddmann [Wednesday, Aug 27, 2014 11:27]

To my best of math knowledge (you might ask this on a math forum),
you can always least-square-fit an arbitrary function with
n free parameters to >=n data points. It's simple: Since
Sum[yi-f(xi,pj)]^2=min., d/d(pj) of that =0. Solve. (Numerically
if you have no other way.)
I have access to MATHEMATICA if that helps you. Just give me
the data list as {{x1,y1},{x2,y2},...} and the wanted curve.

Hauke
 
   
(Read Only)pid=12582

Read more...
Page: [Previous] [Next] 1 2

MatPlus.Net Forum General number of album points and title: how were these determined? and is there a function(curve)...