That's where the idea comes from. Evolutionary change is greatest in the least amount of time in "small populations" though obviously the term is relative. I did not mean it in that sense though but in the sense that the smaller your sample size (statistically) the more it CAN deviate from the norms. What if Cone and Slava were the only ones logged in tonight at 22:05? Then your population (of those logged onto this site) would be 100% synesthetes but the sample population is too small to say anything reliable.

For reference: http://en.wikipedia.org/wiki/Genetic_drift