eli james

Negative Knowledge and Success Stories

1:12am, 3 March 2014

In Antifragile, Nicholas Nassem Taleb points out that negative knowledge (the knowledge that something is false) is more rigorous than positive knowledge (the knowledge that something is true). This is not a new idea: much of the philosophy of science, for instance, talks about falsifiability and the value of negative knowledge.

The core idea behind this is: “a theory that has been proven false today is unlikely to be found true tomorrow. It is, however, more likely that a true theory today will be found false tomorrow.”

Consider the statement “all swans are white”. If you spot a black swan, you can be quite certain that the statement “all swans are white” is wrong. But even if you’ve never seen a black swan, you can never hold that statement as completely true — at best, you may think of it as conditionally true. Since one small observation can disprove a statement, while millions can hardly confirm it, disconfirmation is more rigorous than confirmation.

What’s interesting about this idea is that it may be applied to more things than just scientific discovery. Take our culture’s obsession with success, for instance. Our media is filled with stories of entrepreneurs and companies that have made it big, and they all propose to explain how and why these people have succeeded.

But studying success stories in an attempt to find success falls neatly into the problem of positive knowledge. It’s really hard to tease valuable truths from success. For example, if many stories about founders show how inquisitive they are, it may seem logical to conclude that curiosity is a skill that’s correlated with successful entrepreneurship. But this isn’t true: first, a study of successful entrepreneurs leaves out all the failed entrepreneurs who may have also been curious. Second, journalism is fallible: people think curiosity is an important trait in entrepreneurs, so coverage of successful entrepreneurs often contain some proof of the entrepreneur’s curiosity.

The reality is that we can’t easily conclude that curiosity (or any other factor, really) is correlated to successful entrepreneurship from a study of successful people alone. It is too difficult to tease out all possible factors; worse, some of those factors might have been one-time only events (i.e. being in the right place at the right time). Would Microsoft have won if IBM hadn’t been dumb? Would Facebook be where it was today if Zuckerberg hadn’t been a Harvard student? Regardless of the answers, it’s unlikely anyone would be able to replicate Microsoft’s or Facebook’s success stories in the way that they happened.[1]

So what can you do? The answer that Taleb proposes is that it’s far easier to study failures to conclude what not to do, than it is to study success to figure out what you should.

To my surprise, Warren Buffett and Charlie Munger considered this one of their most important habits. “Invert, always invert.” says Munger, quoting the German mathematician Carl Jacobi. Jacobi believed the solution for many difficult problems in mathematics could be found if the problems were expressed in the inverse; similarly, Munger thought it was far more useful to consider the problem of success in the inverse - what caused failure, and how does one avoid it?

To put this another way: it is easier to figure out what kills companies. If you knew enough to avoid those causes, you’d last long enough to eventually succeed.

This was made very clear to me last summer, when I spent some time reading up on as many startup failures as I could find. What I thought most interesting about the experience was that, after a few of them, I began to see common patterns in the stories — patterns that sometimes ran counter to what the entrepreneur himself concluded from his failure.

One possible reason for this: success stories are often breathtakingly varied in their execution; failures tend to stem from the same proximal causes. It’s easier to figure out why things fail than to figure out why they succeed.[2]

Sadly, it’s not enjoyable nor obvious to seek out accounts of personal failure. I still catch myself glossing over links to startup horror stories, reasoning that ‘no such thing will ever happen to me’. Today, I find that I have to force myself to add them to my reading list, with the knowledge that such stories are ultimately more valuable than the typical hero-worship we see in our industry. And perhaps that little bit of pain is for the better.


[1] Why business schools continue to study success stories is beyond me. I find it enjoyable to read business case studies, but I do so with the knowledge that they are often unrigorous, biased, and overly reductive.

[2] This insight is more useful than it first appears: for instance, it applies just as well to making predictions. Predicting that a brain surgery will go well is difficult, even if we have a surgeon with a good track record. But, as Taleb puts it: “we may safely predict that an incompetent surgeon will cause serious brain damage.”

Other essays


Moving to the least developed ASEAN dragon.

The Most Important Class I Took in NUS

Asian universities and the pursuit of truth.


People buy vinyl for the same reasons they travel to exotic new countries: to find atoms or time in a bit-rich world.