One of the most important points in analyzing the concept of ratings inflation is to know how many total rated players there are now vs. previous years. As far as I can tell, the lowest rated person on FIDE's current list is: http://ratings.fide.com/card.phtml?event=619760 so apparently right now there are about 100,000 rated players, just under 60,000 of whom are active. I can't find complete old lists though, does anyone know where I could look to figure out how many rated players were on FIDE's list 5, 10, 20, 30 etc... years ago?

I think it may be flawed thinking to view ratings as a pool where more players inherently inflate the ratings. Ratings are nothing but a skill measurement. If you're 'n' points above your opponent, then you are expected to win with a probability of 'p' and he's expected to win with a probability of '1-p'. If you add a hundred opponents who are all at the same skill level, then nothing changes. You still win/lose with the same frequency and the ratings remain the same.

Actually, I just wiki'd the topic and this seems to oververify 'my' theory. In fact a major problem that makes a whole lot more sense is, as mentioned, rating deflation!

There's some extensive practical examples of rating deflation on the wiki page and they make a whole lot more sense than rating inflation.

I think the fundamental theory behind rating systems is that if your rating is accurate, and your opponent's rating is accurate, then you two could both play hundreds of games and, completely regardless of your respective ratings, you'd never lose or gain any rating points in the long run.

More players don't *inherently* increase the ratings, but presumably playing strength is some sort of bell curve. If you add 100 players who are all of the same skill level then that will change nothing, you're right. If you add 10,000 players though, they won't all be at the same skill level. If you had 90,000 players previously, and 100 of them were rated 2600+ (so in order to be among the 100 best players in the world, you needed a rating of 2600), then after adding 10,000 more players to the pool you will probably end up with about 111 players rated 2600+, and it will now take a rating of perhaps 2610+ to crack the top 100. So when people argue that rating inflation has occurred because the 100th best player has a better rating now than the 100 best player used to have, it's possible that this effect is due merely to a larger player pool.

Of course it's also possible that's not the cause, and I'm not even entirely sure if there IS a larger player pool now, which is why I'm curious how many FIDE rated players there were on previous complete lists.

I completely agree with you there, but I believe that's actually an example of rating deflation! The max rating increases as the players at the very top are that much stronger, but the guy who was previously 2600 may now only be rated 2580 or whatever as his rating readjusts to the new and increased level of competition, so a modern 2580 would be equivalent to a previous 2600 - rating deflation. The guys at the very top aren't inflated, they're just that good (that sounds quite humorous, but I can't think of any other way to phrase it!)

http://en.wikipedia.org/wiki/Elo_rat..._and_deflation

There's a specific example of this on the wiki page.

Good point. From a technical perspective of what rating inflation actually is I agree fully with you. Most people have a mistaken view of what really constitutes "inflation" however, and it is that colloquial definition that I am directing my comments to. I don't actually know whether rating inflation or deflation (in the proper sense) is occurring, I haven't investigated it. I do know though that people frequently point out that the top rated players have higher ratings now, and use this to claim "Inflation!" All I'm trying to point out is that this particular phenomenon could purely be an artifact of more players, and might not be evidence of actual inflation or deflation.

Many people stopped their chess careers when being on rating "upswing". They took other things (college, poker being common reasons). For such people it's very reasonable to say their "true strength" is higher than their ELO rating.

I was FIDE 2186 when I quit playing in 2001, with USCF of 2200, though I was borderline 2300 US in mid-late 1980s.

Some thoughts on above post: last time I looked at FIDE Top 100 list, a few months ago, one had to be c.2630 in order to make it. Compare this to the early days, when Fischer topped out at 2785 (after dismantling Petrosian at Buenos Aires), after which came several players in the 2640-2670 range: Spassky, Larsen, Petrosian, Korchnoi. I'm doing this from memory, so I may not have have named all the others who were in that range at the time. In those days, 2600 was the floor for the 'Super-GM' category, which, nowadays, means you're 'only' a very good GM, though not a title contender.

hit 2200+ uscf/fide when i was 15 or 16, quit around then, became a poker degen, and now a broke college kid. Typical story.

something maybe stupid:

Inflation maybe because of some cheating going on ( like in myanmar ) a GM going there and crushing those guys would have a huge boost in his rating, then back home beating him would give a massive number of points.

I mean : can the inflation ( if it exists ) be caused by the artificial raising of some ratings putting more points in the "rating pool" than it should be?

I'm clueless obv but just an idea.

FYP for me imo