Only 13 Players Have Hit 2800. Ding Liren Will Probably be the Next Super 2800 GM

Oh, and I suppose "stealing" rating points sounds a little silly, or weird.

What I mean is... anyone who gains rating points is, in the eyes of the system, underrated (that's why their rating is adjusted upwards).

So in this sense injecting players who become better than average is the same as injecting underrated players, and it's easy to imagine the effect of a lot of new underrated players... as they play us they steal all our points!

Since there are a lot of effects (most I'm sure I don't even know about) pulling both ways (inflation and deflation) I'm not saying this is the whole picture. But because I believe players are better today, on average, than 50 years ago, I'm making the argument that the influx of players cannot have had an inflationary effect.

will any human hit 2900?? when??

If you believe all this rubbish about rating inflation, it's inevitable. The players aren't getting any better. In fact, they're getting worse if you believe this. But their ratings will continue to go up because of the magic of the pyramid. 

The extreme example of a very small rating pool has been known for a long time. But the rating pool was sufficiently large in 1972 to act as a normal population. The size of the rating pool shouldn't have a large effect on ratings once the pool is sufficiently large.

I'm not a statistician, but this doesn't sound right to me. The pool of highly rated players have increased enormously since 1972, because not only has the old Soviet Union players been allowed to flood tournaments all over the world, but chess itself has developed enormously and spread all over the world. Africa has seen its first super GM and and now its first E2700+. Back in those days, every interzonals had a few partipicants with a Fide rating of 2350 to 2400, in the Olympiads it was very common for the world's best players to face several players rated between 2200 and 2400. Bent Larsen was #10 with E2620 in 1973 (today there are about 200 players rated above E2600). These days, the world's best hardly ever play anyone who is not a GM or  below E2600, and the rating list has stabilized with just shy of 50 players above 2700 (the current world #10 is Hukaru Nakamura with E2769). I suppose one could argue that there was a deflationary pressure back then, because all the good players from Eastern Europe (and least of all the players from the Soviet Union) weren't allowed to travel much outside the Iron Curtain.

Until the mid 1990s, the floor of the rating system was E2200, which meant that the pool was indeed very small compared to today. It also meant that only the really good results got into the rating list, while the fair to middling to dreadful just vanished without a trace. Since the floor now is E1000, virtually everyone gets a rating very quickly, and everyone plays Fide-rated games all the time. (I played my first Fide-rated games 10 years after I started to play chess, and only played my next Fide-rated tournament another seven years later.) The system is also much more fluid thanks to the higher k factors. While the floor was being gradually lowered to E1000, for a long time we had many Fide-rated events where a large number of the participants didn't have a rating, who 'fed' new points into system by performances exceeding the floor, but took none away from it (since rated players can't lose points to unrated players). Furthermore, once you've broken E2200, your k factor will never change from 10, which means that over time those players will become permanently over-rated because while their strength diminishes, their rating will stay higher than their peers who've never broken E2200 (if we assume, cetaris paribus, that their real playing strength is actually equal).

That said, on the whole the system is clearly stabilising - the E2700+ club has remained around 45 players for a good many years now.

12 if you don't count Giri, who hasn't yet hit 2800 on an official FIDE rating list.

There's an explanation that those who worship the GMs of 1972 don't like:

Today's best players are that much better than those of the 1970's. Kasparov and others explained the revolution that occurred in the late 1970s and 1980s. 

In the mid70s, there was talk of how weak the generation following Fischer was. Karpov and Korchnoi were clearly the strongest two players, with others from the old guard not too far behind.  But around 1979, a new generation led a revolution.

The reason for that was simple demographics, the players born in the 1950s was the born of the war generation - it was a depleted generation.

However, in consideration of the fact that players of the 1972 generation continued to be highly influential well into the late 1980s, long after the new generation had come in - players like Kortchnoi, Spassky, Hort, Portisch, and Larsen who were all in their late 40s and early 50s. Today's players are 'much better' because they're standing on the shoulders of those giants. It's just a matter of an information gap. Given access to modern information, players like Lasker, Capablanca, Alekhine, Botvinnik, Fischer etc. would have been just as good.

However, that hasn't really got much to do with the ratings going up. The pool of strong players available to play everywhere has increased enormously over the past 20-30 years (not least thanks to the fall of the Berlin Wall). And the greater the pool of highly rated players, the easier it is to increase your rating. It's not just a case of them suddenly becoming so much better: a player like GM Ulf Andersson, world #3 or 4 in 1982, achieved his highest rating when he was nearly 50 years old. The same goes for the Peruvian GM Julio Granda Zuñiga, while GM Nigel Short achieved his rating peak at the age of 39, 11 years after he challenged Kasparov for the World Championship, and at a time when he wasn't even #1 in England (and the same goes for GM Michael Adams, his peak rating came at age 42, 12 years after he was #4 in the world). They did so because the pool of highly rated players kept increasing. Yes, sure, their skills developed as well, but had the pool of players remained consistent in size, they would never have achieved the rating increase.

did nepo ever hit 2800

As of today, Nepo’s highest live rating was 2797.3.

Chessbase recently published a few articles convincingly demonstrating that there has been rating deflation over time.

I didn't realise it was as many as 13. It is easy to remember when it was zero. Then one. Then eventually two.

I put the extension of the top end entirely down to improvements in the highest quality of play, which is quantifiably true.

I recall when Fischer had a published USCF rating over 2800 (circa 1971)  The USCF rating system was designed by Prof. Elo.

@Jamiedelarosa, if you compare oranges to oranges, USCF ratings to USCF ratings, Nakamura has the highest ever USCF rating at 2878. I’m sure you will find a reason to say Fischer was better tho.

Nakamura got choked out after starting a fight, Fischer didn't. hahaha.  So he's got that goin for him which is nice. 

Who else among the elite has that goin for em?

Can you accurately compare players of different eras?  I think their is good mathematical evidence to suggest ratings inflation.

It was the pioneering Isaac Newton who said, "If I have seen further, it is by standing on the shoulders of Giants."  No doubt Nakamura would agree.

Nakamura famously said that Fischer would almost certainly lose to all of the top players today.

https://chess24.com/en/read/news/nakamura-public-forum-appearance

The Elo numbers don't tell you if there has been inflation or objective standards have increased. Computer analysis finds it is the latter.

until he gets access to engines

It's as irrelevant as saying similar for any sport where technology plays a significant role. No-one knows how good Fischer would be if he lived in our time, but it is better than he was (which was rather good, especially at his brief peak).