First time entrepreneurs have 18% chance of success; 2nd time entrepreneurs: 30%

There's, like, four total people in the world who match that description in web-based companies over the past few years.

"The findings are similar if we define success to also include firms that were acquired or merged."

Also numbers seem to be ok (about 1-in-5 for the first time enterpreneurs, 1-in-3 for the already successful ones and 1-in-4 overall), considering these are all venture backed companies, thus pre-selected from a much larger pool.

Plus, if you check the list of 40 VCs (Kleiner Perkins, Sequoia, Benchmark, Bessemer, Greylock, Accel, etc.), these are also skewed toward the top of the distribution, so there is another level of pre-selection.

The real probability based on the article (according to my calculations) is 34.4%, which is actually much better than I would have guessed.

Here is how I came up with that.

The article states a first time entrepreneur succeeds at an 18% rate. And someone succeeds at a 20% rate given they failed the first time.

P(First Time Success) = .18

P(First Time Failure) = .82

P(Success | First Time Failure) = .20

P(Second Time Success) = P(Success | First Time Failure) x P(First Time Failure) + P(First Time Success)

0.344 = .18 + (.82 * .20)

In words: 18 percent of entrepreneurs will succeed on their first try. 82 percent will fail. Of that 82 percent that failed, if they try again, they will succeed 20 percent of the time. To get a number from the original population of first time attempts you have to multiply the 82 percent who failed by the 20 percent who succeeded on the second try which results in 16.4%. That 16.4% is the percentage of the original population who tried on the first try but did not succeed until the second try. The end result being 34.4% of the original population will be successful on their first or second try.

Hopefully that makes sense.

* edit to format the formulas a little better

Check your statistics ;-)