Imagine that you are an avid gambler on your way to Las Vegas. At the top of your list of things to do is to try out the new casino that has just opened.
However, you recently took a course on statistics, so before spending any of your own money at the tables, you observe how other gamblers fare so as to calculate the odds and figure out whether the house is pricing them correctly. It turns out that, at the end of the day, out of roughly 100 people making bets, 1 in 28 lost all their money.
The odds look favorable, so the following morning, after the obligatory visit to the extensive breakfast buffet, you start gambling. Unfortunately, lady luck appears preoccupied with someone else, and you eventually go bust.
What happened?
The easy answer to Nicholas Taleb’s famous thought experiment is that, well, you were simply unlucky. But it was actually inevitable – you could have calculated the probability of going bust and it would have been 100%. Keep playing for long enough and, sooner or later, fatal failure is mathematically guaranteed.
The explanation for the common confusion lies in what Taleb calls the Ludic fallacy and the difference between ensemble probabilities and one-time probabilities. In essence, it means that 100 people each falling one meter is not the same thing as one person falling 100 meters. The probability of success for a large sample does not apply to the individual if there are potential points of no return; a person going bust does not affect the game for other players, but it eliminates the game for the individual.
In strategy, we often forget about the ensemble and focus on the one-time. It is, admittedly, easy enough to do – an external agent such as consultancy or agency usually does not have to carry the cost of error. In analogous terms, we get to play the house odds using someone else’s money. Thus, no matter the result, the game carries on, and we can continue to promote focus, sacrifice and big bets.
For the client that goes bust, however, it is an entirely different matter.
Of course, few external aides would disregard client implications entirely (though I have seen it more than once), so they aim to create solutions that are as robust and failsafe as possible. However, this merely plays into the original problem. In a complex market, nothing can ever be entirely failsafe. With enough repetition of exposure, failure is thus inevitable – and potentially catastrophic.
Consequently, any strategy that deals with the complex should instead aim to be safe to fail and build organizational resilience, i.e., the capacity to rebound from trauma with continued identity; not merely extensibility after a surprise but an ability to adapt to future surprises as conditions evolve.
Indeed, this is what Taleb and, on a much humbler level, I have been referring to when, in each our ways, we have stated that in order to thrive, you must survive.
Another way of explaining it is the difference between improving what you can take before falling down (robustness) and improving your ability to stand up again when you do (resilience).
The traditional belief is that by increasing robustness, one might increase the set of disturbances that the organization can respond to more effectively. But, as Taleb’s casino example and my previous posts on complex adaptive systems demonstrate, that is not just wrong pragmatically, but a priori.
In fact, there is a growing body of evidence (formal, theoretical and empirical) to demonstrate that expanding a system's ability to handle some kinds of events actually increases the systems vulnerability to other kinds of events – any attempt to become optimal with respect to some variations, constraints or disturbances increases brittleness in the face of variations, constraints and disturbances that fall outside of the set.
Put differently, the more we target, the more we focus, the more we sacrifice, the bigger the bet, the larger the risk, the more inevitable the failure and the more consequential it will be.
Of course, in business practice, choices have to be made and there are certain unavoidable formal requirements that can lower resilience. But strategy has to be cognizant of both the odds of success and the cost of error, and ensure that resilience is increased to the extent possible in a given context, not decreased in favor of a robustness that worsens the odds of long-term survivability.
Sometimes that is impossible and you have to, effectively, bet a life. But in most circumstances, it is not. Treating situations as if it were only improves the odds of death.
The key to strategy is ensuring that regardless of the outcome, our organizations or clients can make another bet tomorrow.
That does not mean playing it safe.
It means understanding when one can afford not to.
Onwards and upwards
JP