Estimating the IQ of Curtis Yarvin
After some digging, I’m pretty sure I know where the 1/10000 claim for the SMPY 3rd cohoty comes from - the original creator of the study tested 396 gifted students (who were roughly in the 98th percentile of their class), 7 of which scored above a 700 on the SATM. Essentially, a score of 700 would be two standard deviations above the average of a sample that is two standard deviations above the average. Because of the way restricted samples work, this is actually probably somewhere close to 3.5 standard deviations above the mean or so. A 1/100000 is a z-score of 3.7, so their claim isn’t that far off.
Yarvin was in the SMPY, a longitudinal study of talented youth which has various cohorts of varying ability. He was almost certainly selected into the 3rd one of 1/10000th talent, based on his birth year of 1973. This 1/10000 pass mark corresponds to a 3.719 z-score (equivalent to 155.7 on an IQ scale) on whatever test they passed. He also claims his tested IQ is 167 - I'm pretty sure that's just whatever score he achieved on the SMPY entrance exam, but I can't say for certain.
It doesn’t really matter whether he got through on the SATV or the SATM, given that the g-loadings are very similar (0.79 and 0.77 respectively).
I must adjust for longitudinal regression to the mean, which also applies for gifted students.
Assuming a g-loading of 0.8 (which is appropriate, given it correlates at that with ASVAB), the IQ of moldbug is calculated the following way:
(3.96-0.7)*0.765*0.79+0.7=2.67
0.765 figure comes from averaging out the two stability estimates
0.78 is the average of the g-loadings of the SATM and SATV
0.7 comes from the average jewish IQ (~110)
The 3.96 figure comes from the inference that the average student who scores above 3.7 SD is going to be higher than that, based on simulations (and basic intuition).
This results in an estimate of 140.
yeah idk 140 looks fine to me