Peter Zhdanov
-
9 June 2021

On Rating Deflation in Modern Chess

If you check out chess forums where people enjoy bitterly arguing over such topics as "who was the greatest player of all time?", you will often see arguments about modern chess ratings allegedly being inflated. For instance, you will encounter claims that Bobby Fischer's 2785 rating points in 1972 were worth more than Garry Kasparov's 2851 in 1999 or Magnus Carlsen's 2882 in 2014.

However, what we are witnessing in open tournaments right now is gradual rating deflation. Barring bright juniors shooting through the roof, most of my friends and acquaintances are shedding rating points despite feeling like they have improved their game. Of course, one should take the latter claims with a grain of salt: as players age, they seem to better grasp the game's intricacies, but their practical skills deteriorate. Also, chess keeps evolving very fast, so, quoting the Queen of Hearts from "Alice in the Wonderland":

It takes all the running you can do, to keep in the same place. If you want to get somewhere else, you must run at least twice as fast as that!

Still, whereas a few years ago, I more or less knew how a 1700, 2000, 2300, and 2600 were supposed to play on average, now you have no idea what to expect regardless of the rating of your opponent. For example, in one of the rounds of the recent Nezhmetdinov Memorial, I was paired against a 12-year-old kid rated 1747. Typically, this should have been a piece of cake for me, but this guy dispatched me quite confidently. He played the opening well and fast, and all his remaining moves in the game matched the first line of Stockfish. I don't even think he was cheating! You see such "miracles" left and right these days, so they are not something extraordinary anymore.

Another participant of the event, a Grandmaster aged 29 and rated 2462 who usually is quite consistent, lost in the first two rounds to players rated 1985 and 1925, then bounced back in a couple of games, drew a 2011, won a couple more matches, lost to a 2265 and withdrew from the tournament after Round 8. Over these eight rounds, he shed 26.3 rating points, which is quite a lot for someone with a K of 10.

A 40-year-old player from my home city rated 2147, who is considered locally to be quite strong and recently had a rapid rating of 2350+, lost a whopping 81 rating points in this tournament.

On the board next to me I overheard an amusing dialogue between a small kid (1784) and an experienced man (2012). Before the game, they started comparing their Lichess ratings in blitz, and it turned out that both of them were rated between 2300 and 2400 there. These numbers were actually higher than I expected. Guess what? The underdog won the game rather effortlessly as if the ratings were reversed.

I could go on forever bringing up similar examples, but I guess you see my point already.

What are the main reasons for rating deflation in open events?

First and foremost, FIDE has been gradually decreasing the rating floor. Initially, it was 2200. In 1993, the floor was reduced to 2000, then to 1600 in 2004, to 1400 in 2006, to 1200 in 2009, and to 1000 in 2012. On paper, this looks like a great idea: why not allow people of all skill levels to obtain ratings and see how they compare against the world's top players? Also, this is beneficial for FIDE, as they can report a larger player pool to potential sponsors and earn some money from the players themselves. However, in practice, this system has negatively affected both the newcomers and, particularly, the already established players. Here is how it works in a nutshell:

Josh is a 5-year old kid who diligently studies chess in a club and has a great personal coach. To start gaining tournament experience as early as possible, the boy plays a local event against other kids and gets an initial rating of 1038. Since Josh works hard and smart, he keeps improving rapidly and should have been rated about 1500 by age 6. However, his actual rating is, in fact, only 1176 since he keeps mainly playing against his peers, who are also improving quite quickly. By age 8, he is rated 1500+ and starts playing open tournaments, while his real playing strength is more like 1700-1800. As the adults facing him acknowledge quickly, he is much better than the rating suggests, so for them, it is by no means possible to score close to 100% against him, as the rating difference implies. The result is that the grown-ups keep losing rating points against the bright chess prodigy, while he is not gaining them fast enough to obtain an "adequate" rating.

To compensate for this phenomenon, junior players and newcomers in general rated below 2300 have a K of 40 (compared to 20 for most players and to 10 for players who were rated 2400+ at some point in their careers). Still, most of them don't play enough games to close the gap between their real playing strength and the official Elo.

This leads us to point #2: the COVID pandemic made things much, much worse. Right now, there are armies of promising juniors out there who are hungry for progress and who haven't had a chance to play live rated tournaments for the last year. By training hard and smart and playing online, they improved by a few hundred rating points, which is not reflected in their Elos. If you get sucked into facing such guys, getting only a couple of points for a hard-earned victory, and squandering close to 20 (10 for 2400+ players) in case of a loss, be prepared to kiss your rating goodbye.

Let's say we have 25-year old Jessica (2025) playing against 8-year old Mathias (1505). As the rating difference between them is 520, Jessica is expected to score 0.96 points against Mathias (see the table above). If she wins, she will gain (1-0.96) * 20=0.8 rating points. If Mathias wins, he will net 0.96 * 40=38.4 rating points. As you can see, Jessica is not motivated to face such an opponent because she will probably have to work very hard to earn that meager 0.8 points. And in the long run, she will lose a lot of rating points by taking on such opposition.

What can be done in the long term?

Of course, I am not the first person to raise this concern publicly. There have been other cries for help before. For example, in 2019, before the pandemic, Walter Wolf wrote a column for ChessBase, addressing this issue and advocating for a tricky mechanism:

To better account for the increasing playing strength of young players one could slightly raise the K-factor of the winner and slightly lower the K-factor of the loser in a game between two juniors. The winner then would gain more rating than the loser would lose.

A similar plus is necessary when adults play against juniors because the overall Elo level of junior players can only rise in games against adults.

In his recent FB post, GM Alexander Moiseenko estimated the average rating deflation to be somewhere around 200 points and suggested the following measures:

  1. Increase the rating floor to 1600.
  2. Assign the initial rating according to the player's performance. According to him, right now, FIDE uses a weird formula that takes into account the average rating of the opposition, so "even if you score 9/9 against 1200-rated players, you will receive a rating of around 1500."
  3. Bring back the "350 points difference" rule and award 1.1 rating points (for K=10) to the winner if s/he is rated 350+ points above the loser.
  4. Introduce the "350 points difference" rule in rapid and blitz.

One can come up with many other suggestions. It is a matter of acknowledging the problem and initiating an expert discussion within FIDE.

Any short-term tips?

Since FIDE is unlikely to mitigate this issue in the nearest future, here are some of the possible options that my friends and I came up with during internal discussions:

• Play only events with a relatively high rating floor. For instance, 2000+. In such tournaments, the percentage of underrated kids is smaller than in open tournaments where, let's say, everyone, is allowed to participate.
• Stay out of live tournaments for a while and focus on playing online (a questionable decision since these are two very different formats).
• Find "chess reserves" where the players are relatively old. This is what some norm hunters do, by the way: prey on the overrated opponents in round robins instead of dueling underrated kids in open events.
• Quit competitive chess (haha…or not?!).
• Ignore your rating entirely for now and focus on chess improvement if you genuinely love the game.

Anyway, if you are frustrated about youngsters chipping your hard-earned rating points away, I want to let you know that you are not alone!

Are you tired of endlessly facing cheaters online and of getting shortchanged by underrated juniors in live events? Check out the Play Magnus app, where you can challenge such top players as Magnus Carlsen, Judit Polgar, Wesley So at a fixed age and playing strength. Isn't it a pleasure to get a fair fight and to know what to expect from your opponent?

Don’t Miss Out!

Get monthly product updates, exclusive content, and special offers!