Estimating the IQ of Andrew Tate
quick facts:
1600 elo in chess - which is pretty good, according to the website itself this is roughly the 97th percentile.
His dad was the one of the best Black chess players ever - top 2000 at one point. I will assume that equates to a z-score of 4 or above to make the estimation easier
Net worth between 10 million (according to police) and 300-400 million (according to the internet). The figure of 10 million seems very implausible to me - this is clearly somebody who has a net worth of at least 50M. His Hustler’s University alone probably netted him at least that much based on his subscriber count. I’m going to assume that he is between 3.5 and 5 standard deviations above the mean in wealth due to the uncertainty regarding his fortune.
Claims to have a tested IQ of 148. Do I believe it? No. Not necessarily because I don’t think he’s smart, but because I don’t think he’s trustworthy. But there’s decent odds he’s telling the truth. Maybe P(honest)=0.3?
He takes some general knowledge questions on air… and he claims there are 172 countries and says the Amazon is longer than the Nile. Makes the 148 claim a little suspicious (well, more than it was already).
If we take Tate for his word on his scored IQ, he has an estimated IQ of 143, considering regression to the mean.
If you don’t, and estimate it based on his chess performance (r = 0.4) and his wealth (r = 0.35), you get an estimate of 128 with a standard error of 13.3.
Combining these estimates together using simulations produces an estimate of 133 with a standard error of 13.6. This is done by simply combining the two generated vectors (one from the 143 estimate and the other from the 128 estimate) and then calculating the mean and SD manually.
set.seed(5) g <- rnorm(60000000) c2 <- 0.4*g + rnorm(60000000)*sqrt(1-0.4^2) c1 <- 0.35*g + rnorm(60000000)*sqrt(1-0.35^2) subby1 <- data.frame(c2, c1) subby1$g = g subby2 <- subset(subby1, (subby1$c1 > 3.5 & subby1$c1 < 5) & (subby1$c2 > 2 & subby1$c2 < 2.2)) mean(subby2$g)*15 sd(subby2$g)*15 app <- append(c(), subby2$g) g <- rnorm(60000000) c2 <- 0.9*g + rnorm(60000000)*sqrt(1-0.9^2) subby1 <- data.frame(c2, g) subby1$g = g subby2 <- subset(subby1, (subby1$c2 > 3.1 & subby1$c2 < 3.3)) mean(subby2$g)*15 sd(subby2$g)*15 324*0.3/0.7 324/(324+139) subby2$g[1:139] app2 <- append(app, subby2$g[1:139]) mean(app2)*15 sd(app2)*15ncG1vNJzZmirlZe3prrSnqVnq6WXwLWtwqRlnKedZL1wsdKtoKaZpJ67qHnToZxmoaFivKd5wKebq52nYsGiwMQ%3D