3 ms·
I agree that the data is interesting* and I apologize for over-reacting. I also agree with your second paragraph completely. *Performance metrics for many of t
by pxlpshr 17y ago
I agree that the data is interesting* and I apologize for over-reacting. I also agree with your second paragraph completely.
*Performance metrics for many of these early stage startup funds are skewed by the quality they attract. Ycom and TechStars have the creme of the crop, I expect to see better results overtime... 3 companies from S09 raised >$1M. And as you know, the rate of failure in startups is also naturally high - there are far too many factors involved that my decision would be less guided by metrics and moreso toward your second point.
- ig1 17y agoI'm actually doing a similar analysis of UK universities for a non-profit I help run (http://quofoundation.org http://quofoundation.org) and see the exact same issue. With universities you can get around the problem by adjusting for entry grades. I wonder if something similar would be possible in this situation.