As pointed out by @sakai I am surprised there is no mention of sample bias. I wonder if the results where adjusted to account for over- / under- representation

As pointed out by @sakai I am surprised there is no mention of sample bias. I wonder if the results where adjusted to account for over- / under- representation from certain geographies, especially non-english speaking (e.g., Beijing) who probably never heard of SGP, much less used their Startup Compass. Otherwise all good. By far the most detailed comparison I have seen of different locations and beats the pointless VC dollars per capita argument being touted at conferences and all over the interweb. I guess at the end of the day it is a personal or team decision depending on many factors (connections, market, investors, family & friends ...etc.) but the more information to make that decision the better.