I’ve heard a similar story from a number of accelerators these days. It goes something like this: “We’ve decided to try to work with companies that are a little further along. It allows us to have a bigger impact on the companies when there is more meat on the bones.”

This makes some sense to me. And I also understand that these accelerators are increasingly being judged by institutional standards. In the pressure to show success, you want to have a high % of companies get funded at decent valuations and a low mortality rate.

But I’ve been thinking recently that most of the time, when we pass on a company that comes out of an accelerator, it’s not because the company failed to make enough progress. It’s also not because we don’t like the founders – often, you are very impressed by what they’ve accomplished in a short amount of time.

Usually, I pass because I don’t love the idea. Too safe, too unoriginal, not in my area of interest, whatever.

I know that’s probably anathema in a lean-startup world. But increasingly, I’ve been thinking more in the “Deterministic-Optimistic” frame of mind. I’m caring less these days about early traction or getting a product a few people are willing to pay for.  I’m more interested in companies that think they know how to reorganize, transform, or invent entire industries.  I like the lean methodology, but only as a means to an end.

It’s harder to do that with startups that are a bit further along.  It’s too hard to start over and too easy to make the most of what you have.  Plus, part of the magic of accelerators in the startup ecosystem is the ability to take extraordinary people and let crazy ideas bloom a little.