Does FIDE need to raise the GM/IM/FM bar?

In 1972 there were only 88 GMs in the world.

http://en.wikipedia.org/wiki/Grandmaster_(chess)

Now there are about 1,000.

http://www.nationmaster.com/graph/spo_che_gra-sports-chess-grandmasters

I'm guessing there were probably around 200 IMs worldwide in 1972, and now there are more than 2500.

http://www.nationmaster.com/graph/spo_che_int_mas-sports-chess-international-masters

Some might contend that there are more and better players now than ever, and this accounts for the vast numbers, but there is good evidence to the contrary.

http://www.chessbase.com/newsdetail.asp?newsid=5608

So, does FIDE need to raise the bar and/or yank some titles or what?

No I don't think they need to raise the bar, the bar is meant to be a constant based on a comparable average of strong play.  Just because there are more titled players doesn't mean that the quality of the play isn't just as good - I would suggest that the quality of the play is actually probably better in general, considering the modern training techniques and the introduction of computers as aids to training.  You also have to actually look at the numbers of people playing Chess; these numbers have gone up and so statistically so will the numbers of people getting titles.

AMcHarg,

I provided a link addressing your argument. Your argument doesn't wash - so it would appear.

From a sporting and scientific point of view, FIDE should probably make it more difficult to get the GM title.

However, from a financial point of view, the more there are, the better for them, as federations need to pay a small fee for every player who gets a title...

Titles are fun. Perhaps they should introduce a Super GM title, requiring 10 norms (2850+ performance), past performance doesn't count.

Edit: In fact, FIDE should run this more like World of Warcraft. Invent norms that you can only get in teams of twenty...

It would help the situation if they just did away with all the bogus/affirmative action titles they give. Look at the ratings of the current top Puerto Rican players for example. There are several "IMs" that have never been over 2400 which is the min required rating for the IM title, so why are they IMs ? Apparently , if you win a National Championship you get the IM title, no matter how weak the players you play may be, this should be stopped.

I would want to see the number of players.  No way you could "yank" a title, but you could make it harder.  It a vast number of higher ranked players are playing, then it is fair.

Remember, the Elo system uses the delta between players.  A top ranked player gains nothing by beating me.  I stand to gain a lot if I beat a titled player.

The only viable system to replace the Elo system, which is wonderful since it is very accurate and uses only basic algebra, is Finite Markov Chains.  I took that course at the doctoral level.  It would be only slightly more accurate.  

The IM title, if you win your country's national title might be a little too easy.  Perhaps you should have to meet a minimum Elo number AND win the championship.

Since FIDE only publishes numbers quarterly, it is hard to get a grip on what happens in between quarters -- although every master level game is recorded.

If you look at the charts, 3000 has never been published.  I think more top level players are playing.  Although someone who is barely a grand master would get killed most likely by a GM 150 Elo points higher.

A level system might work.  GM I, GM II, GM III, GM IV, etc.  It would also give GM's something to work for - to move up the chain.

I think the IM is a problem.  Some countries are very small and the competition is not strong.   City states like the Vatican or Singapore are quite small.  Puerto Rico is technically a US possession.   Some think it should be the 51st state.  They should have to win the US title.

But that is only one example.

It is an interesting discussion and problem.  But every titled player I have met (small sample) has known their stuff.

And what about the gender stuff????  Come on.  The pieces have no idea who is moving them.  Statically,men and women are equal in IQ.  Why then, should their be a lower bar for WGM, etc.  That is ridiculous.  

It would also give GM's something to work for - to move up the chain.

Actually, "average GMs" have to work hard to make a living. I think that's good enough as an incentive

@Jolly

Well, I don't think the Elo system needs replacing, but it can use correcting and/or modification.

As a side note, except for simple lack of will, I don't see why titles could not be taken away.

Anyway, what I would suggest is first to simply deduct around 100pts from the ratings of all players on the current list to correct for the inflation since appx. 1985. Following that, there should be a review of all titles earned after 1985 using the best statistics on how much inflation had occurred at the time the players made their norms to see if they would have earned those titles adjusting for the inflation (or, if not, whether subsequent results would have earned them the title) and then correct the situation by downgrading to the appropriate title if necessary. My guess is most current GMs would be downgraded to IMs, Most IMs would become FMs and most FMs would become CMs. Then, with the new ratings in place, every year there would be an inflation adjustment to the ratings. This would take the form of simply subtracting some number of points from the ratings of all players on the list - currently inflation is running about 4 points a year according to Sonas. Finally, a new more stringent system of norms would be put in place. For this I would suggest something along the lines of a performance rating goal over a large number of games against opposition of a minimum average rating - e.g. a GM title might require a player to have a better than 2550 performance rating against minimum 2450 average rated opposition over 100 consecutive games. This standard for achieving the title would also eliminate a current flaw in the system whereby a relatively weak player can just play tournament after tournament and eventually title by having a few good results along the way that translate into norms.

Like in most martial arts, ie. Tae Kwon Do, Hap Kido, etc...

Yes, you can "easily" achieve the title of "black belt" but on the master's circuit, a black belt is child's play, where it takes years at a time to earn sequential levels to a black belt, called degrees. My old sensei, Master Yong, was at his 7th degree of TKD and 6th degree black belt for Hap Kido. As far as I know, both of these arts cap the black belt degree's at 9, but this would definitely be a terrific idea for the FIDE.

3rd degree Grandmaster ~ GM*3  =D

well, you could consider 2100 elo = black belt, 2200 = 2nd dan, etc. (2800 = 8th dan - sounds reasonable)

I think that chess play has weakened. I am 13 years old and can beat a lot of the chess players that play here, and in any case playing chess is starting to be just memorizing combinations and tactics. No more original openings.

Correlating the start of the ratings inflation, are you sure it's not caused by the computer and internet revolutions? I feel that many types of study are easier and more efficient now than in the pre-computer past.

For the most part, I see no problem with the rating system as it stands now. The whole thing seems statistically sound.

For competition reasons, I can see where FIDE may want to redefine the classifications. That would be simple to do. You could look at the entire history of rated chess and say, "on average, the grandmaster ranking historically meant the top X% of rated players." You could then adjust the definition of a grandmaster (and other classifications) periodically to account for all of these slow statistical changes.

I don't think it is reasonable that the computer revolution is responsible. First, the inflation started in 1984. There were, of course, computer programs during the '80s, but they were relatively weak with very non-human playing style and few features - little help certainly to FIDE players who were all 2200 minimum at the time. Also, if computers were responsible, there should have been a disproportionate expansion in the west where the technology was much more readily available, but we all know there was no explosion of American and western European GMs relative to the east. More importantly, if computers were responsible, we should have seen an acceleration in the inflation as computers became more widely available, powerful and useful. We didn't and in fact the inflation slowed into the '90s when database programs, internet chess, master strength programs and computer tutorials all became widely available. Lastly, computers are useful tools, but cannot learn and retain information for us, and I think the degree of benefit in productivity of study they do entail is not nearly enough to explain a dozen-fold increase in GM numbers. I would venture to say that a motivated person could get just about as much out dilligent study of books, magazines, Chess Informant and especially the services of a good coach as could a person with Fritz and Chessbase with a million game database.

Adding a title which genuinely compares individuals is probably the way to go.  Something between Grandmaster and World Champion ... such as a WCC (World Championship Candidate) or something along those lines ... the GM title was originially given to people who were competing for world champion. ... that history is too far gone but a new title would work.

I'm not one for stripping titles out of chess... it's part of the game.

(Oooh you play chess? ... are you a master?)

I dont know what discipline you studied, but it is not easy to obtain a black belt.  It takes work, dedication, and years of practice.

 And the top player gets a red belt...at least that's how it works in my martial art style.

Adding a  "World Championship Candidate" title wouldn't be my preference, but at least it's better than the "Elite Grandmaster" title I've heard suggested on another site.

The problem with just adding a title though is it doesn't ultimately fix the problem. If inflation continues, the we will lose all comparison with the past in terms of the meaning of "Grandmaster" or "International Master" etc. Maybe it is too far gone, but I'd prefer to not just give up yet.

Anyhow, I think you may have misunderstood my original post. When I said "yank titles" I meant take titles away from players who would not have earned them absent the inflation (see post #9).

I think the one thing that needs to be kept in mind is that very few of us have these titles anyways... it's still a level of distinction.  I've only beaten a few national masters, never beaten an international master, and never even played a grandmaster ... distinction! :P.

Instead of raising the bar, FIDE could release the Super-GM title, SGM. And 20 years from now, they can release the Super-Mega-GM title, SMGM.