Following up on Ostap's point about college education, as a recipient of high school students who go on to college, many of them are woefully unprepared.  The expectations through high school may indeed be higher, but they just can't write, they can't perform critical analysis, and they seem to think that university involves very smart people imparting the truth so that they can copy it down and then regurgitate it on an examination.

I found this particularly the case in the United States, but also in Canada.  This isn't a reflection on the teachers, but rather the points of emphasis of the curriculum.  High school diplomas and, indeed, university degrees have been devalued even more than the GM title...

Thanks, S_F, for the data. They do suggest an exponential relation between time and the amount of GM's from the very beginning. As the amount of chessplayers is limited, we may expect an S-curve. But we have not reached the bending point yet. The curve might also become linear for a while, though it is very steep. Strictly speaking we must correct the data yet for the amount of active chessplayers, but I cannot imagine that the exponential relation will completely disappear.
So these data do support your argument. I am not surprised, but it would not be the first time, that a reasonable theory does not match the facts.
A theory cannot be true or untrue. Especially when mathematics is involved it is a model, which can be falsified. In that case we must try to improve it. A model describes the data as good as possible. A light example: Ohm's law does not apply anymore, when the voltage becomes too high. The resistor will melt down and the current will drop to zero.
Back to the GM-title problem. While I find the proposal to restrict the GM-title for a certain percentage quite sympathetic, I do not see exactly how we can realize this. Realistically speaking I think we must be glad, if we can flatten the curve to some extent - decrease the increase so to say.
(This post took me much more time than yours - reply nr 18, so I hope you will forgive my previous laziness now).

Maybe this is a good time to segue into a discussion of education.  Ostap berated modern education, especially in the US below the college level.  As a high school teacher, I will defend the honor of my students by pointing out how much they must know now compared to even ten years ago.  

The increased information and the ability to use that information can't be measured on some standardized test on grammar and fractions.  I admire most of my students' abilities to juggle adolescence with the pressures that teachers and parents put on them on a daily basis.  

(I feel comfortable saying that here where no student of mine is likely to tread! )

Smyslov_Fan wrote on 07/16/06 at 00:56:17:


I think we understand each other, and so I'll let this be my final remark.  If chess skill comes reasonably distributed among people (I am deliberately vague here... and also, are ratings normally distributed?), and the upper bound is not reached by humans (both seem like sound assumptions, at least in the near term), then the top 3% players will never be in a class by themselves, because the top .5% will outclass them entirely.  The fact that there are many GMs indicates that the exclusivity of the title has been diminished.  The only way to restore that is to reduce the number of titleholders.  But these two notions (exclusive club, some fixed number of deviations above norm) are not the same; both depend strongly on the total population.

On a personal and practical level, I agree with you that FIDE needs to make it harder to obtain titles, and either perspective will achieve that.  But in terms of the philosophical justification, I am not sure that one can easily escape picking one of the three I listed.

Smyslov_Fan wrote on 07/16/06 at 00:18:45:


Well, when reading your posts, I can't help but feel that you would be unhappy if the GM title had 1000 players, even if those 1000 players met the criteria you have set forth (competitive with the world's best, ranked in the top 1%).  A title which is reserved for players who are in the top ten in the world (even if the title is kept for a lifetime) is still more of a fixed number assignment than anything else, in my view.

I guess my point is simple: if there are 1000 players in the top 2.3% would you be happy, or would you want to make the percentage smaller?  My impression is that you would choose the latter, and my apologies if I am mistaken.  (I understand that it is an empirical fact that the top 2.3% of the group you are interested in does not constitute 1000 people.  But my point here is different: there should be a way to adopt a standard which can stand the test of time, and if one never wants to see 1000 GMs then one can't be satisfied with a percentage solution, since the number of FIDE rated players is probably growing at a nice clip.)

Smyslov_Fan wrote on 07/15/06 at 06:43:16:


Maybe I should have made it more explicit...but that entire bowling thing was completely made up.  I just picked the first activity which came to mind where one had a nice numerical score associated to mastery which would make inter-generational comparison easy.

Quote:


Well, you seem to be a little conflicted over what you want the title to mean...either a percentage or a fixed number.  I suspect you really want it to mean the latter, but if contender just means 'could beat the champ on a good day' and not 'could make a credible claim to be one of the ten best players at some point in their career' I suspect many of the current crop of GMs would keep their title (research proposal: how many people have defeated a top-three player?  I would guess many...how many people have beaten someone who has beaten a top-three player?  I would guess most, if not all, GMs.).  As for me, 99% was a number from the air -- I meant only to suggest the concept, not an implementation.

I think that the entire "GM Inflation" argument comes down to what one wants the title to mean.  Here are three (possible) choices:

1.GM means that one is a contender for the title
2.GM means one of the best 1% in the world (or best 1% with a FIDE rating over 2200)
3.GM indicates a certain level of mastery of chess (which is a bit more abstract than the above)


Let me give a simple analogy: bowling.  Brief summary of bowling: one tries to knock down pins; there is a scoring system, and the best one can do is get a score of 300.  Now imagine that in the 1920's, a rating system was set up: Grand Bowlers (GB) had to have an average score of 270 and have three performances with an average of 290 or better; Intermediate Bowlers (IB) had to have an average score of 250 and three performances with an average of 270, etc.

In 1920, there were only 4 GBs and 20 IBs.  Everyone knew the names of all of the GBs, and most of the IBs were household names as well.  However, the rising popularity of bowling has meant that an increasing number of people have the GB title -- over 1000 now.  What changed?  Does the title mean less than it did back in the 1920's?

It depends on what the title meant in 1920.  If one takes definition 1, then there has definitely been inflation -- there are still only a handful of people who compete for the World Bowling Championship (or the Professional Bowling Championship or the Newton Bowling World Title -- sadly there was a split in the bowling community in the late 80s)

If one takes definition 2, things could have gone either way.  It depends how many "professional bowlers" there are.  If there are 1000 times as many as there were in 1920, then there should be 1000 times as many GBs.  Maybe the GB title has actually undergone significant deflation under this definition -- say there are 10000 times as many bowlers who are internationally ranked, but only 1000 times as many GBs.

If one takes definition 3, then the bowling title system is perfect -- it accurately captures mastery.

*

Of course, it is a bit harder to play this game with chess -- despite all of the questions of how well today's players would do against the players of the 70's or whatever.  If one believes in absolute terms that the GM title has been cheapened ("Today's GMs play chess worse than the GM's of yesteryear.") that is one thing.  One can also have another view ("Today's GMs are as good as GMs in the 1920's, heck they probably play better chess, but I knew the names of all of the GMs then, but now, I can barely keep track of just the GMs who got their title at the last FIDE meeting.").  Whether or not this is inflation depends on what the title means to you.

First people should agree to which of the first three definitions they want the GM title to mean (and I am sure that there are others).  Then they need to decide whether the title means less in absolute terms.  Finally one has to deduce the numerical data (total numbers of players, titled players, professionals, whatever numbers one is using for definition 2).  If people agree to these things (and they may not -- the 'today's GM' versus the 'old GM' debate rages on, it would seem) and *still* can't agree on whether or not there is inflation, then there are other problems.

Here are some more facts, not theories:

In 1996, there were 591 GMs

2001:  768 GMs

2005:  959 GMs
and IMs:  2596.

Make of those figures what you will.



Source:
http://www.angelfire.com/nf/chess/facts.html