tigger2010

Since the data points are solely about extreme wealth and Elite MBAs, the only thing that can be proven is some degree of correlation between the two. What you have is insufficient (by a lot) to prove causation.

If your thesis is that the Elite MBA led to the extreme wealth, you have to look at the volumes of Elite MBAs out there. HBS alone turns out 1000 a year. Only a small % of them get to extreme wealth (going back to your point that a few hundred million doesn't count).

As someone else noted, an Elite MBA is a worthwhile investment as a stamp of approval. I have one of these, and I can speak from experience that it helps. Is it going to get me to extreme wealth? No.

This would explain why one should expect to see the largest frequency of almost-psychopathic behavior in politics and business.

"out of the self-made billionaires, 7 out of 32, or 22%, have MBAs from a top business school. Those aren't bad odds."

Correlation vs. causation. Enough said.

Scanning the comments, there's a fair bit of argument about lean vs. cheap and about whether Horowitz is agreeing or disagreeing with the "lean" approach.

Afaik, all he's saying is "it depends". There may be situations e.g., if your competition is cash poor and you are cash rich, where the "fat startup" approach makes sense. It doesn't seem like he's fundamentally arguing for or against "lean" - just that a lean or fat stance only makes sense in specific contexts.

I recently read a book called "Evil Genes" by Barbara Oakley. One of the interesting stories she tells is about how some of the research into Machiavellian behavior got started.

The original research group decided to focus on "everyday" manipulative, lying, power-hungry behavior. Looking for test subjects, they found plenty of instances, but ended up picking a category of people to study - their own PhD advisors.

The point is, mild versions of psycho/socio-pathic behavior are unfortunately commonplace. Changed my perception of people's behaviors in very fundamental ways.

May be stating the obvious here, but why is a probabilistic prediction model primarily aimed at consumers? Wouldn't it make more sense to apply to users for whom probability is useful e.g., truck fleets to predict delays? oil fields to show likely drill spots? etc.