Why have ratings gotten inflated?

Fischer and Spassky didnt have the luxury of lots of tournies in which only the world's top players were playing .  If the top GMs of today played more strong Open events they would all lose rating points, and thats why they don't do so , with a couple of exceptions. Something else I would like to see is to allow some of the strongest GMs who do play a lot of Opens be allowed to play these elite RR events . 

It's true that the reverence for the master chess player has been lost because of the chess engine. I became aware of this reading the comments on Chessbomb while watching the Yifan Hou games at this last tournament. Chess masters had this mystique for us in the general public before. I personally thought that they were deep thinkers devoting their lives to a complex game that I could never understand. The most attractive thing about chess to me was it's history. The old photos of grand masters elegantly dressed with beautifully carved-wood pieces, an audience in formal-wear sitting silent and transfixed, like watching a profound art being performed. 

But chess is not like this at all, at least not anymore. Now it is like watching children shouting out risible parodies of contemplation, as if one is in the presence of a juvenile competition of insults directed at each other's comic book hero of the moment. The chess engine has allowed for an audience not of respect and awe, but of know-it-alls repeating engine evaluations and adding their own brand of puerile taunts for the master with each move.

That Wimbledon-type atmosphere of tense quiet followed by approving ovations has disappeared, and we are left with the type of vulgar criticisms saturating the comment boxes of YouTube videos 

@ trysts

Yes, on the web, all that is true. However, one can still sit (in silence) ten or fifteen feet from the big guns, if you spectate an event in person. I watched some games at the recent Hastings Masters, though I have to say, there weren't many youngsters availing themselves of the opportunity. Most of the audience were sporting grey hair, to some degree.

I blame progress.

NOW GET OFF MY LAWN!

That would be nice to some day watch a great tournament in person, Doggy

Doesn't the floor have something to do with inflation also? If your floor is 2000, no matter how badly you play you will not go below that.

Gerry

Fischer didn't play in open tournaments after ~1966, his last US championship. He did represent the US in the Olympiads, so did Nakamura, Kramnik, and numerous other players rated +2750. Fischer confined himself to major international events such as Buenos Aires and Palma de Mallorca. He played in Candidates tournaments and Olympiads. That's about it. 

It's true that there were fewer elite tournaments in the late 1960s and early 1970s than there are today, but even then, the best players rarely played in open tournaments. The quality of chess tournaments has risen along with the quality of chess players. Now, there are dozens of elite grandmasters whereas there was only a handful in the 1960s. 

This is a normal progression. Take a look at any athletic event and you will see an explosion of talented players compared to 40 years ago.

Actually that's a good point. If you lose while at your floor, your opponent still gains points. That clearly causes some amount of inflation. 

FIDE doesn't operate rating floors.

In my area the inflation of rating is also due to the fact that if you enter 3 tournaments simultaneously with a low rating e.g. 1400 and in all tournaments you have to get 2 points in order to get rating points.
if in all tournament you manage to get 4 points. therefore instead of just getting 1400+100 you get 1400+100+100+100= suddenly after three tournaments you are 1700. I call that inflated.
I have tried it myself gone from 1400 to 1613. a lot of players do that in my area like a yoyo. we have no rating floor in denmark.
So if a 1700 enters three tournaments and loses points. in every tournaments he is suddenly rated 1300.
some even do that to get a lower rating for team match.

One aspect that must be understood when comparing players from different time periods through the use of their Elo is that a comparison between two Elo values is only valid if they each earned their Elo by playing in the same general population of players. To understand what I mean, if we constructed a group where only young children could play, and these kids had a separate Elo from adults, perhaps one of the kids could achieve a rating of 2,500 Elo. That would indicate they dominate their peers enough to achieve such a rating. However, that would be only against the population of other young children. If that kid were to face an adult grandmaster (or even an adult international master), they would likely get demolished.

People often make the error in thinking Elo values have absolute meaning when they think of the Elo of chess engines. When a chess engine is said to have an Elo of 3,500, that is against the population of all chess engines competing in engine tournaments under certain constraints (like the strength of the hardware, the time control, any enforcement of certain openings being played, etc.). You can't compare that value to human Elo. It's like trying to compare apples and oranges. If you wanted to find out the Elo of Stockfish relative to human opponents, you'd have to have humans play thousands or tens of thousands or even hundreds of thousands of matches done at the classical time control -- enough for the extremely rare event of a human beating Stockfish to fill out statistically and give enough data to pinpoint the Elo of Stockfish. What I mean by that is if you see Stockfish beats Magnus 10 times in a row, well does it have a 90% chance to beat him? Or a 95%? Or a 99.99%? And what are the chances of a draw, which might actually happen more than you'd think, especially if Magnus were going for one? It's tough to tell, because a large range of human Elo performance could explain that small sample size of wins. You'll know more after the human players get enough wins and draws, giving you the ability to calculate a precise human Elo for Stockfish. The best you can conclude after something like 10 wins in a row is that its Elo is likely much higher than 2800 in terms of human Elo, but you can't be too sure of just how much higher. You effectively would have a lower bound on the rating of Stockfish with a certain probability of being correct, and there would be potential for Stockfish to have even higher rating than that likely lower bound.

Back to the main question, the Elo of someone like Fischer was gotten by playing all the masters that played during that time period. The main thing you can determine is that he utterly destroyed the competition. He was beyond his peers in a way that Kasparov was for much of his career. Magnus, on the other hand, has many peers of similar strength although he has demonstrated he is the best of this current generation by a decent amount.

We can only guess why there are so many players competitive with world champions in this era compared to pretty much any other era, and we can only guess why these elite players have gone from grandmaster to super grandmaster. I'd say it comes down to these points:

1.A) More people than ever before are playing chess, and particularly in areas of the world where a USD can go farther, many people from those nations, if they are gifted at chess, can make quite a good living playing chess. There is more money in chess than ever before, bringing in many gifted minds to the game, especially in countries with a lower cost of living. When you have more gifted minds than ever before playing, you're going to have more people competitive with the best in the world as well. Prior to these conditions, you had to have the perfect storm for an extremely gifted person deciding to dedicate their life to chess. They likely needed independent wealth already or some way to survive without working a full-time job at a minimum. The geniuses were mostly working hard at traditional jobs to support themselves and their family financially rather than mastering chess. Take Paul Morphy, the 0th world champion, as an example. He lived in a mansion, and as far as I know, he never worked a job yet lived comfortably. He earned a degree in law that he never used.

1.B) Due to the popularity of chess worldwide and the ease of communicating with people through the internet, gifted players can attempt to become supreme at the game with more confidence that their work will pay off even if they don't become the best, because tutoring people online is a viable way to make a living. As point 1.A mentioned, this is especially true in areas where the cost of living is lower. A person whose rent is 1/10th that of the average in the US can tutor Americans for US$50/hr, and that might go quite far where they live. The end result is that more great minds choose to study and play the game as being even just good at chess can support more people financially than ever before. And if they end up becoming supreme at the game, they can make even more money than they'd get tutoring new players and experts. A super GM can tutor promising, young grandmasters for even more money than an expert would pay. The rarer the knowledge a person has to give, the more money they will make giving it. It's just supply and demand.

1.C) Streaming gives yet another avenue for a person to make a living by playing chess if they don't end up totally supreme at the game, and this access to making a living wage can only increase the amount of gifted minds that choose to go full-time into chess rather than going full-time into a traditional role to make a living. Once again, this impacts people with a smaller cost of living even greater as the money they can earn goes farther where they live.

2.A) There is a much larger availability of theoretical knowledge (online tutorials, videos of lectures, and books), there is an easy way to get supreme analysis of your games with engines for free where you previously had to be in the presence of a grandmaster to get good analysis (that you had to pay + the higher quality of that GM, the more you had to pay), pretty much every important game ever played and even just played is online for studying, and people can play the game way easier due to online chess -- you are moments away from a game with an opponent around your skill level to push you along the path that takes you to being a super grandmaster. More ways of improving means that those who are gifted with immense work ethic that decide to go full-time into chess can near their theoretical limits easier. Prior, there were likely many extremely gifted players who, due to inaccessibility of the things I just mentioned, never became competitive with the likes of Paul Morphy or Fischer or Kasparov. The fact that the USSR had such a dominant force in chess speaks to this truth as, due to their attitude toward teaching and playing chess, many youthful players had the opportunity to get nearer to their maximum performance whereas people in countries without those types of resources may have fell behind despite having immense natural ability.

2.B) Younger players who demonstrate their prodigal nature readily play against the best in the world, accelerating their progress in mastering chess. Even back in the 1970s, it was quite rare for a 16 year old to play a super grandmaster. They were stuck beating people around their own age and physically around them, which isn't the best way to improve.

3.) The best minds can prepare better than ever before, giving them the ability to crush less talented opponents as both players have about equal access to their opponent's games. A person can load up a database and analyze to their heart's content the play style and opening choices of their opponent. More knowledge means fewer random losses, and fewer random losses means obtaining a more supreme rating. In the past, maybe a lesser known player could study the games of a super GM whereas the super GM couldn't get their hands on their opponent's games. That imbalance in knowledge meant people couldn't go as high in Elo, because they'd lose to preparation more often than they do now. These days, both players can prepare for each other the same amount, and you'd bet the better talent who has developed their skills more will win out over their opponent.

4.) There is much more access to performance-enhancing drugs like Adderall. All it takes is for a doctor to say you have ADHD, and a player now has a pill that can spawn endless mental endurance on command. Prior, if you got tired, you maybe had a cup of coffee and / or a cigarette. Amphetamines improve performance and mental endurance way more than coffee and a cig. I believe amphetamines are against the rules of FIDE, but I've never heard of a tournament taking blood samples to find out who is clean and who uses chemicals to enhance their play. This aspect also enables older players to stay competitive with younger players, solving one of the problems older players have as they age: Mental endurance. As a funny example of that in another field, the great mathematician Erdős readily used Ritalin to fuel his insane pace of creating novel proofs by the week. One time, his friend bet he couldn't go a week without it, so he had a momentary moment of sobriety. As the story goes, he spent most of the time asleep during his withdrawals. After winning the bet, he told his friend that he had held back the progress of mathematics by a week.

One reason to think factors like these explain the higher ratings more than rating inflation is that there is some evidence that ratings have not inflated. One statistician did an analysis of player's games, using the most powerful chess engine available, and found that the increase in rating came with more accurate play overall. Now, there are mechanisms in the FIDE rating system for points to enter and exit the system, so there can be inflation or deflation. Interestingly, since older players often retire after accruing a bunch of points, there might even be slight deflation rather than inflation. There is no de facto answer to this question. In either case, the stuff I wrote above explains part of the equation.

The way points enter the system is when new players get a rating. Those points pop into existence out of nowhere. Worse, maybe some of the players lose a good deal of points and then quit. That aspect in isolation does increase the number of points per player, which is a force toward inflation. At the same time, different players gain or lose different amounts of Elo based on stuff like how many total games they've played and whether they have reached high enough a rating. In other words, certain games are not a zero-sum game with some players winning or losing more or less points based on that aspect of the system.

All those words, and the post can be ignored because there is demonstrable rating deflation since Covid, and FIDE is working to limit it.

My post highly indicated there isn't much rating inflation and likely rating deflation instead.

Why can't we put the old games from yester-year into the modern chess engines to evaluate them and see how their games rate on the computer (for accuracy)?

We can, but people don’t like the answers they get so they fiddle with the formula.

A statistician employed by FIDE and this site came up with an “Intrinsic Performance Rating” which showed that prior to the Pandemic there was no rating inflation and in fact there was rating deflation. That means that GM Deac today is stronger than Spassky was in his prime and that Fischer is about 21st all-time.

People don’t like those answers so they go creating less sound methods and claim there’s been massive inflation since 1972.

One way that a rating pool can inflate is a result of natural aging. Players get weaker as they get older, losing some of their rating points to other (younger) players... and they eventually quit chess or die, removing their deflated rating from the pool and thus raising the average slightly.

There are many plausible reasons for deflation or inflation. But that doesn’t mean inflation exists. Deflation exists, and is demonstrated repeatedly. Deflation is such a problem that FIDE is working to combat it. Deflation became a huge problem due to the pandemic, but existed even before that.

The best players in the world are lower rated than they were a decade ago despite playing better!

Many people believe that inflation exist because over the past 70 years the top GMs went from being around 2600 at most to being at 2800 at most. Im not sure if its because people have improved or if its actually inflation