Chess is growing. The top is shrinking.
The FIDE rated population more than doubled between 2013 and 2026. Over the same fourteen years the number of players above 2000 fell by 15%, and above 2700 by nearly a third. Both of those are counted from FIDE's own lists.
Twice as many players
In January 2013 FIDE published standard ratings for 90,783 players. In January 2026 it published 211,925 — an increase of 133% in fourteen years. By any measure of participation, competitive over-the-board chess has had an extraordinary decade and a half.
The median rated player moved in the opposite direction, from 1904 to 1711. That is what you would expect from rapid growth: new arrivals enter at the bottom, and a pool that doubles mostly at the low end pulls its own middle down with it.
What is not expected is the table below. If a pool doubles, the obvious prediction is that every level gets more crowded — more players overall means more strong players, even if the proportion falls. That is not what happened.
| Rating | 2013 | 2026 | Change |
|---|---|---|---|
| 2,000+ | 32,452 | 27,553 | -15.1% |
| 2,200+ | 10,505 | 8,175 | -22.2% |
| 2,400+ | 2,101 | 1,707 | -18.8% |
| 2,500+ | 778 | 595 | -23.5% |
| 2,600+ | 222 | 164 | -26.1% |
| 2,700+ | 48 | 33 | -31.3% |
Counted directly from FIDE's January standard lists: 90,783 rated players in 2013 and 211,925 in 2026. Each row is the number of players at or above that rating on each list.
Every level above 2000 has fewer people in it
There were 32,452 players rated 2000 or above in 2013 and 27,553 in 2026 — a fall of 15.1%, in absolute numbers, while the pool as a whole grew by 133%. Above 2200 the count fell 22.2%, above 2400 by 18.8%, above 2500 by 23.5%.
And the higher you look, the steeper it gets. Players rated 2600 and above fell 26.1%, from 222 to 164. Above 2700 the count went from 48 to 33, a fall of 31.3%. The very top of chess has roughly two thirds of the people it had in 2013.
The top-100 mean rating tells the same story more quietly: 2702 in 2013, 2682 in 2026. Twenty points off the average of the hundred best players in the world, over fourteen years in which the game was supposedly booming.
What could produce this
The first candidate is deflation, and this site has measured it separately. Following the same players year over year — matched by FIDE ID so the population is held fixed — ratings drift very slightly downward rather than upward, which is the opposite of the usual complaint. A pool that deflates a point or two a year will, over fourteen years, push some players below every threshold you care to draw.
The second is arithmetic that follows from the first. Rating points are conserved in a closed system: a new player entering at 1400 and losing brings points in, but a large influx of underrated newcomers taking points off established players moves the whole distribution down at the top while widening it at the bottom. Growth and thinning are not in tension; one can cause the other.
The third is that the strongest players of 2013 got older. A rating list is not a measure of talent in a country; it is a list of who is currently competing. Some of the 48 players above 2700 in 2013 retired, and the generation replacing them has not yet arrived at the same numbers — which may say more about how ratings are earned now than about how good anyone is.
This page cannot separate those three, and it is not going to pretend to. What it can say is that the shrinkage is real, monotonic in the threshold, and large.
Titles went the other way
The count of titled players rose 52.2% over the same period, from 8,613 to 13,105. Set beside a 22% fall in players rated 2200 and above, that looks like a contradiction and is not one.
Titles are permanent. A player who reached the Grandmaster standard in 2015 and has since dropped to 2400, or stopped playing entirely, is still a Grandmaster and still counted. The titled population can only grow, whatever happens to the ratings underneath it.
Which is the strongest argument against reading a titled count as a measure of anything current. The 52% rise in titles and the 22% fall in players above 2200 are both true, and only one of them describes chess as it is being played today.
What this measures
Standard ratings on the January list, in 2013 and 2026 only. FIDE's rapid and blitz rating lists began later and are not comparable across the full window, so nothing here says anything about them.
A player counts if they held a published standard rating that January, which does not mean they played that year. The companion article on annual rating movement finds that 27.4% of rating-years show no movement at all, so a substantial share of any list is dormant.
And 2013 is the earliest list this site holds, because FIDE's downloadable archive does not reach further back in a form that can be parsed. Fourteen years is a long window and it is not a long history — the counts before it may well have looked different again.
Common questions
Are there more chess grandmasters than there used to be?
There are more titled players — 13,105 against 8,613 in 2013, a rise of 52% — but fewer players actually rated at grandmaster level. The count above 2500 fell 23.5% over the same period. Titles are permanent and are never revoked, so the titled count can only rise.
Is the number of strong chess players falling?
By count, yes, at every level above 2000, while the rated population more than doubled. Above 2000 the number fell 15.1% between 2013 and 2026; above 2700 it fell 31.3%, from 48 players to 33.
Why is the FIDE median rating falling?
Mostly because the pool grew. FIDE published 90,783 standard ratings in 2013 and 211,925 in 2026, and new players enter near the bottom, which pulls the median down — from 1904 to 1711 — without anyone getting worse.
Data updated 2026-08-02
Chess.com publishes no rating distribution. Those figures are derived from the Lichess distribution mapped through the conversion anchors, and are rounded accordingly.