Website founded by
MatPlus.Net Forum General FIDE albums, quality and titles
You can only view this page!
|Page: [Previous] [Next] 1 2 |
|(1) Posted by Marcel Tribowski [Wednesday, Oct 6, 2010 14:24]|
FIDE albums, quality and titles
As there might be a consensus among problemists, FIDE albums are the most known and perhaps the most important collections of contemporary chess problems and studies, edited in 3 year periods. Three kinds of composition titles, FIDE Master, International Master and Grandmaster, are awarded to composers, depending on how many of their problems and studies have been reprinted there. In each section, each of three International Judges gives 0 to 4 points to each entry. Usually, a composition belongs to the selection of the best with at least 8 judgement points.
There have been controversial discussions if 1) chess composers as artists do need any titles for their work at all and if 2) these albums both can fulfil the intention of being documentation and the only basis for composition titles.
But apart from this, album judges often are confronted with a flood of entries of poor average quality. In 1994, K. WIDLERT, an “official” of the former PCCC, already had described this phenomenon [Die Schwalbe, 08/94, p. 496]:”The number of problems to judge: This is a real problem in some sections […] we have not found a good solution to this.” Now, 16 years later, the situation seems to be unchanged. The announcement to the album 2007-2009 contains the following section:
“Composers are asked to number their entries in each section in order of preference, starting with the one that the composer considers best. This will ease the work of the judges and may also be of benefit to the composer.”
It’s doubtful if this appeal will solve a structural problem:
1. The actual title system is purely based on quantity at a minimum level of quality, i.e. 8 (sometimes even 7.5) points are sufficient for a selection. To achieve one title point, it doesn’t matter if a selected composition received the minimum of 8 or the maximum of 12 points.
2. The entries are submitted by composers themselves. Generally, one has to presume the possibility of gaining titles as one of the main motifs for an entry: composer who don’t care about them don’t send anything to the albums.
Today, it cannot be seen why the announcement’s call to authors’ self-criticism will be honoured. The actual title system encourages every potential title candidate to submit as many entries as possible, according to the shotgun principle: one bullet may hit. Even supporting this, the latest “Handbook of Chess Compositions” contains a ranking of the “FIDE album point order” in where one album problem counts as one point. In these charts, the more, not necessarily the best seems to be of benefit to the composer.
In order to restrict the number of entries, i.e. to receive only the best compositions for the albums, it will be necessary to reform the title system of the PCCC/ICCU. The aspect of quality needs to be considered in a stronger way.
Therefore I make the following proposals:
In addition to the “FIDE album point order” mentioned above, an “album quality order” should be published ‘officially’. Obviously, the principle of comparison, of competition is one of the main motives for highest human efforts. The album judges’ evaluations of the selected works have made public since the album 1980-82. Since 30 years now, experts’ judgement points could be quoted for quality. But although it does exist, the difference between good and better compositions has not taken into account yet in connection with the title system.
As an example, the following table contains the available album judgement points 1980-2000 for composers who achieved the GM title and are represented with more than 12 judged compositions in albums. The given points are set into relation with the number of judgements (no discrimination here if a problem or a study has been judged); judgements of joint compositions are divided by the number of authors. It may contain some mistakes, but already gives some kind of overall impression.
points : judgements = Ø
(1 problem = 1 study)
Rehm........... 1156.58:122.53 = 9.44
Loustau.......... 630.25:68.33 = 9.22
Abdurahmanovic... 900.58:98.17 = 9.17
Kricheli............... 219:24 = 9.12
Vukcevic............. 1091:120 = 9.09
Lobusov.......... 601.42:66.58 = 9.03
Pervakov......... 402.92:44.83 = 8.99
Caillaud....... 1782.33:198.92 = 8.96
Mladenovic........536.25:60.33 = 8.89
Petkov............ 2139.25:241 = 8.88
Vladimirov...... 1017.5:114.75 = 8.87
Kuzovkov........ 998.42:112.67 = 8.86
Nestorescu........ 199.25:22.5 = 8.86
Keller.............. 634:71.83 = 8.83
Kovacevic........ 556.92:63.33 = 8.79
Zappas............... 377.5:43 = 8.78
Rusinek............. 307.25:35 = 8.78
Cheylan.......... 557.67:63.67 = 8.76
Avner............ 487.42:55.67 = 8.76
Heinonen............... 725:83 = 8.73
Marandyuk.......... 675.5:77.5 = 8.72
Rudenko.......... 672.45:77.25 = 8.70
Dobrescu........... 352.5:40.5 = 8.70
Lindgren......... 336.92:38.75 = 8.69
Chepizhny............ 603.5:70 = 8.62
Pachl............... 696:80.92 = 8.60
Kralin........... 267.92:31.17 = 8.60
Goldschmeding........ 180.5:21 = 8.60
Selivanov........ 169.75:19.75 = 8.59
Alaikov............ 697.5:81.5 = 8.56
Gurgenidze....... 511.83:59.83 = 8.55
Macleod............. 410.25:48 = 8.55
Grin................ 136.75:16 = 8.55
Janevski............. 1067:125 = 8.54
Slesarenko.......... 585:68.83 = 8.50
Degener........... 623.75:73.5 = 8.49
Feoktistov....... 248.38:29.25 = 8.49
Garai................ 694.5:82 = 8.47
Shanshin............ 516.42:61 = 8.47
Goumondy............... 161:19 = 8.47
Gamnitzer............ 448.5:53 = 8.46
Bakcsi.............. 362.75:43 = 8.44
Ahues.................. 310:37 = 8.38
Tura................... 506:61 = 8.30
Shavyrin........... 306.5:37.5 = 8.17
As a consequence, the PCCC/ICCU statutes concerning composing titles should be changed in the following way:
“§ X. Award of titles
a) “Grandmaster of the FIDE for Chess Compositions“
A problem composer whose first FIDE album problem has been published in 1980 or later must achieve at least 600 judgement points in FIDE albums. For a study composer the minimum number of judgement points for studies required in albums is 360.
b) “International Master of the FIDE for Chess Compositions“
A problem composer whose first FIDE album problem has been published in 1980 or later must achieve at least 200 judgement points in FIDE albums. For a study composer the minimum number of judgement points for studies required in albums is 120.
c) “FIDE Master for Chess Compositions“
A problem composer whose first FIDE album problem has been published in 1980 or later must achieve at least 100 judgement points in FIDE albums. For a study composer the minimum number of judgement points for studies required in albums is 60.
In counting the points, for the mentioned master titles the value of a joint composition will be divided by the number of composers collaborating. If a composer competes with problems as well as studies then 1 study = 1 2/3 problems.”
There are rumours (confirmed by reading not only the last year’s meeting’s report) that some problem commissions are too engaged with internal affairs to conclude any substantial decision. So it might be rather naïve to believe that proposals by outsiders will change anything. But maybe MAT PLUS FORUM is the right place to present them.
Are they anticipated?
|(2) Posted by Frank Richter [Wednesday, Oct 6, 2010 17:35]|
Interesting proposal. I hope that these and other ideas will be discussed contructively during the congress.
On the other side it doesn't solve the big problem of too many entries by several composers. In my opinion it must be possible to find here a useful restriction. It is simply not real to send more than 200 problems for a single album section!
|(3) Posted by Harry Fougiaxis [Wednesday, Oct 6, 2010 21:04]|
Are they anticipated?
I am afraid, yes. A similar proposal was submitted by Reto Aschwanden and Jean-Marc Loustau in 2005 and it was finally not approved (can't recall right now in which congress). The proposal is available at http://users.forthnet.gr/ath/harryfou/pccc_2005_proposal_aschwanden_loustau.rtf However, like Frank above, I have not fully understood how this would help the poor album judges to avoid having to examine all these mediocre works sent for evaluation by certain composers.
As for the indication of rank of preference by the composers themselves, when submitting their works, the original proposal was suggesting that "Composers should have to submit their entries with a rank order of preference. A FIDE Album judge should be entitled to stop judging an individual composer's submission when he has found 5 problems in sequence to which fewer than 2 points have been given". The first part was approved (as a recommendation and it was included in the announcement of the previous album 2006-08, too), but the second was rejected in 2006 in Wageningen.
|(4) Posted by Marcel Tribowski [Thursday, Oct 7, 2010 11:06]|
I wouldn’t underestimate the psychological effect of a new ranking list described as ‚step 1’. The value of a title will be not an absolute, but a relative one.
I’m sure there still will be some candidates with lack of sensitivity. But fewer of them might be interested in problems selected ‘by accident’. Perhaps it will lead to the effect to work longer on each problem for better quality.
The actual title system promotes mass production, with album judges as victims.
|(5) Posted by Hauke Reddmann [Thursday, Oct 7, 2010 13:07]|
And here is MY modest proposal to find the quality of a work :-)
Points are given every time a problem is reprinted elsewhere
(for sheer adoration; of course problems used to illustrate
a theme or as bad example are exempt). Details of the scheme
can be lifted from science, where this system is widely in
use. (It can also be used to rank the quality of a *publication*).
I admit that the scheme has its shortcomings (which are discussed
in science journals, of course), e.g. fad overrating (neither
the proof of Fermat nor the first Babson I would rate as high
as they are quoted), citation daisy chaining etc., but the
advantage is obvious: no judges needed.
|(6) Posted by seetharaman kalyan [Thursday, Oct 7, 2010 13:08]; edited by seetharaman kalyan [10-10-07]|
"In my opinion it must be possible to find here a useful restriction."
I think a restriction of 50 problems for each Album (not per section !) per composer can be imposed. As for myself, I have not been composing for decades and when I was active, I found that copying the problems 5 times was too much of a bother !
|(7) Posted by Georgy Evseev [Thursday, Oct 7, 2010 14:16]|
The current system based on the percentage of sent problems de facto stimulates the composers to send more garbage.
Let us imagine that I've composed 3 good problems (7+ points) and a dozen of poor problems (5- points). I want to maximize the number of problems included. I cannot affect the marks given to my problems, but I'm able to change the number of problems selected. Sending also a dozen of garbage problems I generally increase the number of problems to be selected (by approximately 2) and so the probability of selection of my better problems if they will make a boundary case.
And at the same time there is no negative side for me in it. In the worst case it simply somebody else who will be lucky. But i have done everything I could to increase my chances).
I see only one way to avoid this difficulty: fix the _absolute_ number of problems to be included into Album (per section, probably). There are some methods which allow to minimize deviation of number of selected problems from desired value.
|(8) Posted by Marcel Tribowski [Thursday, Oct 7, 2010 14:43]|
PS: The proposal of Jean Marc Loustau and Reto Aschwanden from 2005 doesn’t mean quite the same. Without the first step described above, the second just would lead to a faster distribution of titles. (But as we now, one of a problemist’s most important characteristics is .. patience.)
I would prefer a ‘soft’ way of restriction to a hard one, to make a cut at, let’s say, 50 entries per author/album. Because some are really more productive than others, on a high level.
|(9) Posted by Kevin Begley [Friday, Oct 8, 2010 21:02]; edited by Kevin Begley [10-10-08]|
>As there might be a consensus among problemists, FIDE albums are the most known and perhaps the most important collections of contemporary chess problems and studies, edited in 3 year periods.
I do agree with this, but I have one important question, which might help your proposal...
What is the purpose of the FIDE Album?
I am not aware of any consensus as to the present mission...
Some would say it exists to provide a long-term (why not timeless?) chronicle of the best problems [over a given period]
Others would say it exists to provide a timely accounting, for the purpose of bestowing titles upon the best composers.
Both are noble goals, in my view; but, contrary to popular mythology, the album can not fulfill both missions.
[In fact, I do not believe it presently fulfills either.]
Do we need two separate albums?
Of course, there are a number of issues surrounding either option... many of which you have noted.
* self-promotion is required,
* it is not a single-album, but a collection of several sub-albums (with sub-genres more politically-, than logically-divided),
* etc etc etc
Your proposal is interesting... at the very least, it helps acknowledge that improvements are needed.
But, this being a remarkably complex issue, it might be helpful to first establish some consensus on the simple questions.
|(10) Posted by Dan Meinking [Friday, Oct 8, 2010 21:40]; edited by Dan Meinking [10-10-08]|
I think the real solution to this dilemma is to PENALIZE a composer for submission of sub-par entries. If we say "6 total judge points" for a given problem is AVERAGE, then below-average problems get NEGATIVE points, and above-average ones get POSITIVE points. The composer's TOTAL points for each section will determine the number of Album selections for that section.
Here's a possible scheme: (TJP = total judge points; EP = entry points)
00.0-03.0 TJP = -2 EP
03.5-05.5 TJP = -1 EP
06.0-06.0 TJP = +0 EP
06.5-08.0 TJP = +1 EP
08.5-10.0 TJP = +2 EP
10.5-12.0 TJP = +3 EP
If a composer earns (say) 30 positive points, but also has 10 negative points, they end up with 20 TOTAL points. So... starting with their top-scoring entries, they get enough entries until their 20 points has been reached or exceeded. One could use Marcel's proposed "composer self-ranking" to break ties, in the event two or more equally-scored problems put a composer over his entry-point limit.
In this case, the "10 negative points" will cost the composer some Album entries. That is the PENALTY for submitting mediocre work! In the event that the negative points out-weight the positive, the composer gets ZERO entries.
For joint problems, each co-composer gets an equal share (positive or negative) of the entry-points.
This system achieves the following objectives:
(1) It forces composers to be selective.
(2) It does not limit the number of entries for a prolific, high-quality composer.
(3) It does not penalize entries that are not quite of Album quality, but still average or better.
(4) It allows (and encourages) judges to award points as they see fit, without restrictions.
Of course, this is only a "rough sketch" proposal. I'm sure a more precise scheme could be drawn up.
|(11) Posted by Georgy Evseev [Friday, Oct 8, 2010 22:20]|
I am strongly against any penalty. There are several possible cases when such penalty is obviously bad.
1. A beginner composer is not himself able to distinguish good from bad, and his good problems does not score in the end.
2. A mediocre joint composition is sent by one of authors and the other one receives penalty.
3. Simply a difference of opinion between the author and the judges.
4. The desire of judges to penalize or not penalize authors may affect their marks.
ad so on...
|(12) Posted by Dan Meinking [Saturday, Oct 9, 2010 00:48]|
I agree with you on #1 and #2. But #3 and #4 already exist, in my opinion.
|(13) Posted by Georgy Evseev [Saturday, Oct 9, 2010 10:06]|
Yes, you are right concerning pp.3 and 4, but your suggestion still adds another level of effect, like giving 2 points to a problem (instead of 1, for example) to avoid penalizing the author.
There may another question, which I consider valuable for this discussion.
Currently the system is very simple and transparent. Every composer may ask why specific problem did not make it into Album and (probably) receive from director the marks for his problems and immediately make necessary conclusions.
What will you think about the system based on some average calculations made by computer? The rules remain formal, but knowledge of the marks for any specific problem is not enough to explain why it was or was not included.
|(14) Posted by Kevin Begley [Saturday, Oct 9, 2010 10:55]|
What would you think about establishing some kind of progressive penalty?
The more points you have, the more you are expected to avoid an excessive number of weaker entries.
|(15) Posted by Joost de Heer [Saturday, Oct 9, 2010 11:28]|
For joint problems, each co-composer gets an equal share (positive or negative) of the entry-points.
IMO this should only apply to the composer(s) who submitted the composition. If three composers create a joint, and only one of them thinks it's good enough to send, why should the other two be penalised or awarded?
|(16) Posted by Georgy Evseev [Saturday, Oct 9, 2010 13:12]|
I still think it is easier to break the link between the number of sent compositions and the number of selected ones). This also allows to take into account average generosity of the judges.
|(17) Posted by Dan Meinking [Saturday, Oct 9, 2010 14:39]; edited by Dan Meinking [10-10-09]|
George: "Yes, you are right concerning pp.3 and 4, but your suggestion still adds another level of effect, like giving 2 points to a problem (instead of 1, for example) to avoid penalizing the author."
True. But... you have to remember that to get "-2" or "+2" or "+3" entry-points requires all THREE judges' consent. I still think the proposed point system does not change this.
George: "1. A beginner composer is not himself able to distinguish good from bad, and his good problems does not score in the end."
A simple way around this would be a proviso: "Any contributor who currently has ZERO Album points will not be penalized." The implication is: If you've already received one or more Album points, you should know what is mediocre or not. :-) Also, the "ZERO Album points" could be "less than 5 Album points", or some other arbitrary criteria. We'd want it to be absolutely fair.
EDIT: I like this idea even better: "A composer who submits more than 10 entries for a given section may be penalized in that section." That forces the beginner composer to be selective, AND allows the experienced composers to submit fewer entries without risk of penalty. Again, the number "10" could be "8" or "12" for example.
Joost: "IMO this should only apply to the composer(s) who submitted the composition. If three composers create a joint, and only one of them thinks it's good enough to send, why should the other two be penalised or awarded?"
Makes sense to me. If the problem is selected for the Album, all three composers still get 1/3 of an Album point as they do now. Seems fair.
|(18) Posted by Vlaicu Crisan [Saturday, Oct 9, 2010 17:34]|
Somehow off-topic proposal: Perhaps the FIDE albums should also group the problems appearing in 2 years instead of 3 years (as it is now).
1. The number of submitted entries will be reduced, so it will be less burden on judges.
2. The number of selected problems will fall within an acceptable threshold for the FIDE Album editors.
3. The results will appear sooner, hence the titles will be also granted sooner.
|(19) Posted by Kevin Begley [Sunday, Oct 10, 2010 01:20]; edited by Kevin Begley [10-10-10]|
>"Any contributor who currently has ZERO Album points will not be penalized."
>The implication is: If you've already received one or more Album points, you should know what is mediocre or not. :-)
It is possible to have more than ZERO Album points, without ever sending a problem to the album.
Why not use a progressive formula?
I do like your proposal that 10 or fewer entries (to a single section) would be exempt...
But, I really don't enjoy the necessity of self-promotion anyway.
Is there no good way to entirely circumvent self-promotion?
Why not base entry upon the date of the award, rather than the publication date.
Then, automatically send all problems which earned a place in the award.
Problems which did not earn a place in the award would require self-promotion (and be placed into a separate pile).
This way, we would establish a kind of supreme court -- which might provide valuable feedback for judges.
Your proposal would disadvantage talented beginners (even more than they already are).
Judgments very often take longer than 2 years.
Without seeing the judgment, newcomers may lack the confidence to send worthy entries.
Also, two-years puts pressure on judges (to locate anticipations faster, and also to search for cooks).
Though, I never understood why there is no recourse -- after a problem is accepted/declined.
It never made sense to me that you could earn a title by sneaking 12 incorrect (or anticipated) problems passed the judges.
I still wish we had two albums:
* One in the timely interest of awarding titles to top composers,
* the other in the timeless interest of collecting the best problems.
|(20) Posted by Jacques Rotenberg [Sunday, Oct 10, 2010 01:27]; edited by Jacques Rotenberg [10-10-10]|
why not each year ?
or, perhaps, another possibility would be to keep the 3years albums, and to divide the 'heavy' sections in 3 teams of judges - one for each year.
|Page: [Previous] [Next] 1 2 |
MatPlus.Net Forum General FIDE albums, quality and titles