3 ms·
I agree that it's a potentially large problem, though I'd bet it's a systematic under-representation which would draw roughly the same conclusion, albeit at a h
by akharris 11y ago
I agree that it's a potentially large problem, though I'd bet it's a systematic under-representation which would draw roughly the same conclusion, albeit at a higher overall rate.
- birken 11y agoIt is certainly possible, but you really don't know. Especially given that the numbers are so small, any false negatives would affect the data quite a bit. And you are making very strong claims with very precise measurements that are drawn from a very error-filled data source. A better source would be a survey sent to YC companies, or perhaps data from some internal YC database. That will have its own sampling issues, but the accuracy will be leaps and bounds above Crunchbase.