What is Viswanathan Anand’s IQ?
One peak rating, two formulas, two answers: 182 and 191. Neither is a test result.

Every figure we could find for the IQ of Viswanathan Anand is either arithmetic or unattributed. The arithmetic: correspondencechess.com (November 3, 2023) runs his rating through 100 + 0.06 × (Elo − 1300) and gets 187 at his 2023 rating of 2754 and 191 at his 2817 peak, and gamelevate.com (January 11, 2024) applies the Levitt Equation to the same peak and gets 182. The unattributed: a grid on worldiqranking.com, seen in a screenshot posted to X on January 5, 2025, lists 190 beside Albert Einstein’s 160, and the opening post of a Chess.com forum thread lists 165–175 for his years as champion; when other members asked for documentation, the poster replied only that it came from a trusted source. None of the four cites a test. For how such figures usually start, see where celebrity IQ numbers come from.

The best-known of these formulas has an author who attached warnings to it. Jonathan Levitt proposed in Genius in Chess (Batsford, 1997) that a player of IQ Y will tend, after many years of tournament play and study, to reach an Elo rating of about 10Y + 1000. He wrote that it should not be taken too seriously and that it cannot be used to deduce a person’s IQ from a rating, because the years of effort it presupposes may not have been put in; his conclusion that a world champion would need an IQ near 180 was drawn “for the sake of speculation”. We could not trace the other formula to an author or a study. Applied to his 2817 peak, the two formulas give 182 and 191; applied to his September 2026 rating of 2739, they give 174 and 186. That is four different “IQs” for one man, and the number follows the FIDE list, not a trait. The check is the one champion on this site with a reported test result: the profile of Garry Kasparov carries a reported 135, while gamelevate.com’s equation gives 185 for his 2851 peak, fifty points higher.

What a published rating can say is real but narrow. Elo is a results-based, continuously updated measure of chess strength against other rated players; it cannot be converted into a general-ability score. The largest synthesis of the research, the 2016 meta-analysis by Burgoyne and colleagues (19 studies, 1,779 players, in Intelligence), found, as originally reported, that cognitive ability correlates with chess skill at r = .24 overall (.22 after a 2018 correction) but at only r = .14 for fluid reasoning among rated players, about 2 percent of the variance; full-scale IQ tests alone explained under 1 percent, from just six effect sizes. Its Elo-rated studies ran from 1311 up to 2607, and the authors flag that rating spreads inside samples were narrower than in the full population, so it says little about a player rated 2817. See what a chess rating can and cannot say about IQ and the wider research on chess and intelligence.

Anand’s own words point the same way. In an October 2023 Forbes India interview he said that when he thinks about other areas of life he asks what his ranking is there, not in chess, and reminds himself that he should “probably not extrapolate too much” from his assumptions. In Mind Master (2019) he retells Adriaan de Groot’s experiment: grandmasters and masters reproduced a real position 93 percent correctly, but on shuffled positions every group, grandmasters included, placed only three or four pieces correctly, which he takes to show that chess players have “contextual memory”, not an inexplicably photographic one. That is a skill tied to chess positions, one reason it can look extraordinary at the board and say little on a general test (how memory relates to IQ). So the range follows the rating, as on the site’s other chess profiles. His 2817 peak is one point above the 2816 of Hikaru Nakamura, whose band is 140–155, so the two share a band; Magnus Carlsen, with a 2882 peak, sits slightly higher at 145–160. As on most pages here, the documented record supports a lower range than the circulating figure.