This rebuttal is at least as badly constructed as the article. Tearing something apart isn’t putting something together. If the author doesn’t believe the article works, a tortured negating of Kauffmann doesn’t make your argument. The real question is whether accelerators (and incubators) create more or better companies.
First, the Warner Yard “research” group is nothing more than the foxes guarding the hen house. Gee, what a surprise they think accelerators “increase survivorship” over companies that didn’t go through one. Duh. But what this fox & hen house “study” doesn’t bother to mention is this. They are comparing a few dozen of what the accelerator folks think were the elite of the elite, to everybody else in the universe. They pick through thousands of startups to find the tiny few they think already exhibit the best chance for survival, and even with that huge head start they barely beat the entire rest of the field.
Harvard picks the elite of the elite and then falsely claims their students are more successful because they attended Harvard. Utter nonsense. Accelerators and incubators use they same false equation. They take a tiny percentage of the elite and try to claim they succeeded because they attended an accelerator.
This blog’s argument falls apart on the same Harvard premise — correlation does not prove causation. If that Warner Yard group really wanted to do research, they would study accelerator attendees against only the first runner-ups who didn’t get in. Study 1,500 accelerator attendees against the 1,500 who were closest to being picked. If you did that, you would see no measurable difference, which is why the Warner Yard guys studied their elites against the entire field of startups — a ridiculous comparison.
There is not real evidence they help, and there is plenty of logical evidence (the Harvard effect) and historical evidence that they don’t.