Estimating the IQ of Anatoly Karlin

• 

• 1510 on SAT (z-score of 2.55 assuming M=1000 and SD=200 in a nationally rep sample)

• 137 on cattel

• 120-125 on legit online tests

• 100 on cattel, when he was time constrained.

Assuming g-loading of 0.84 of SAT, 0.8 for cattel (ravens has g-loading of 0.8, from page 38), a 0.7 for the test he was time constrained on, and 0.7 for online tests yields a simulation estimate of 132 with an SE of 5.7. The simulation returns a value of 136 when the score of 100 is excluded - averaging the generous and spiteful estimate puts him at a 134.

g <- rnorm(60000000)
ls <- 0.84*g + rnorm(60000000)*sqrt(1-0.84^2)
cs <- 0.8*g + rnorm(60000000)*sqrt(1-0.8^2)
as <- 0.7*g + rnorm(60000000)*sqrt(1-0.7^2)
bs <- 0.7*g + rnorm(60000000)*sqrt(1-0.7^2)

subby1 <- data.frame(ls, cs)
subby1$g = g
subby1$bs = bs

subby2 <- subset(subby1, (subby1$cs < 2.55 & subby1$cs > 2.4) & (subby1$ls > 2.45 & subby1$ls < 2.7) & (subby1$as > 1.2 & subby1$as < 1.8) & (subby1$bs > -0.4 & subby1$bs < 0.4))
mean(subby2$g)
sd(subby2$g)