John Wolfe: [Let’s say] there are 100 companies, and you assume that 90% of them are crappy companies — that’s the base rate. Out of our 100 sample, 90 are bad and 10 are good. And now let’s add an analyst that’s working for me. He or she has a 90% success rate. Of the 90 that are bad, she picks 90% right, so she gets 81 and correctly says that those are bad. And she picks wrongly that 9 of them are good. Then of the 10 that are actually good, she says 9 are good and gets one wrong. So if I’ve been given a list of 18 companies, only 9 of them will be good.
Think about that for a second. When there is a 90% that companies are bad, and a 90% accuracy rate being able to identify them, I still have a 50% hit rate of being right. So someone with 90% accuracy is no better than flipping a coin.
So either you can find someone that’s better than 90% accurate… or you find an area that there’s a higher percentage of good companies. And good companies could be themselves, they are intrinsically good, or that no body else is looking and therefore is a function of price. It’s scarcer. I like to go into areas that other people aren’t looking.
Demetri Kofinas: It’s like the Hamptons. It’s fully developed and everything’s priced in. Are you going to go invest there or are you going to go try to find the next Hamptons?
That was from an excellent interview on Hidden Forces with John Wolfe, co-founder of a venture captial firm.
I don’t think it is thought of in this way in the DOD. We do not assume a base rate failure of 90%. We assume that the requirements and competitive proposal process will provide a base rate of 100% success. That’s precisely why so much time and effort is involved up front.
Or perhaps, we assume 90% of all project requirements or technical solutions are bad, but we work hard to get our accuracy of selecting the best ones up to 100%.
Here is one of Wolfe’s defining philosophies for investments:
The first one is randomness and optionality… you want to expose yourself to as many people, as many ideas, with the intellectual humility that you don’t know what the thing is that’s going to matter and then all of those are effectively free options. Free options in that you have spent time, maybe as little money as possible.
Learning is the greatest option there is. Reading lots of different books in lots of different sectors and then finding cohesive ideas between them. Meeting lots of people in different disciplines is another way. And then sharing ideas with other people, saying “hey I’m interested in X,” then they connect you to somebody that connects you to somebody. After-the-fact this constellation of things is obvious, but beforehand I’m intellectually honest enough, you just never know.
That was great, and the correct mindset to recognize hindsight bias. Notice how this nonlinear path of discovery is something a little different than the stage-gate milestone approach we take in the DOD. Wolfe admits at minute 17:30 that he has something akin to faith that his searching and connecting the dots will lead to a good outcome. His decisions are based on tacit knowledge, and a process of rapidly learning and updating expectations. That is a bit different than what is needed to justify a multi-stakeholder defense program.
I think the following part is a great example of a cultural theme in R&D that is an essential part of this blog:
If you question the dogma of a scientific sector or you question the dogma of the market I actually think it leads to great reward.
I love this story. Nobel laureate, Peter Mitchell. He was looking at the chemo-mechanical way that things transported within cells between membranes. Everybody said he was totally wrong, he’s an idiot, and what was amazing is that he was proved right, and when won the Nobel prize, he presented charts that showed all the empirical data, but also showed a list of all the people who said no, that he was wrong, and the date they changed his mind.
I think of John Wolfe as being a great example of a new generation of “Smart Creatives” who thrive of exploring a diverse range of ideas (are generalists) while connecting the ever-expanding set of dots in new ways (combinatorial innovation). It is like the opposite of the index fund stock market, where there is just a flat base consensus and no one is taking speculative or consequential bets.
Leave a Reply