Exploration, Exploitation, Education

A discussion I often have is whether to eat out at a new place, or stick to old favourites. If it’s one of our few meals in a new town (say we are on holiday) we very rarely eat in the same place twice; there are plenty of great places out there, and it’s unlikely we’ve found the best one early on. Exploration seems like a good idea here. On the other hand, when we are eating locally, we figure we have a pretty good sense of what’s what, and often go for the best thing we already know. In this case, exploiting what we know may be the best strategy.

This is a very helpful characterisation of a certain class of problems Source

The same situation, arises, for example, if you are choosing a new school in a new country, where it might be unwise to accept the first one you like. On the other hand, if you have had two kids go through several happy years in one school, then sticking with it for your third is probably the right thing to do.

This simple conundrum - choosing between what you know and what you don't - is called the explore-exploit trade-off and it’s surprisingly common (so much so that there is even a rather elegant branch of maths - optimal stopping theory - devoted to these types of problems). Should we...

• ...stay living in our hometown or move away?
• ...offer new products or stick with existing ones?
• ...always read the same genre of books?
• ...stick with our existing close circle of friends of seek new ones?
• ...pursue a career in depth or hop between careers?
• ...hire a very good candidate for a job or hold out for an event better one?
• ...deep-dive into specialties, or retain broad but shallow expertise?
• ...stick with the current great job or have a new adventure?
• ...stick with the new school subject or change, when I wasn’t taken by the first 1, 10, or 20 lessons?
• ...marry that boyfriend/girlfriend or dump them?

(The last one is only slightly tongue in cheek - and I feel sorry for those who meet the love of their lives on their first date; how would they ever know?)


Obviously the answer in any given situation will depend on many variables; but the structure of the problem is the same; seek new knowledge and opportunities, or  go ahead on the basis of what you already have. And the crux of the issue is that the very knowledge that would help make the right decision - that is, the experience of the unknown - can only be achieved by actually making the decision in one way, often irreversibly. One way of thinking about the problem is through a metaphor like this: 

The question is, what’s the relative size of the area to the left of the line, and the area on the right of the line? Of course, we can never really be sure, but in case B explore looks to be the right option, whereas case C seems to lean towards exploit. The difficulty arises, of course, in case A, or in cases where you do not even know where the line is.

In his recent book Range, journalist David Epstein makes the case for exploration over exploitation; for trying new stuff a lot; and not sticking with what you know, even if it is pretty successful. He starts, however, with the case of Tiger Woods.  Here, early very narrow specialization paid off handsomely. Pushed and coached by his father, Woods started with a putter at seven months - yes really!  At two he was on national TV, driving golf balls with a club almost taller than he was; at three he won an under-10s tournament and at four “his father could drop him off at a golf course at nine in the morning and pick him up eight hours later, sometimes with the money he’d won from those foolish enough to doubt”. Much has been written about Wood’s legendary success; it fits with Amy Chua’s Battle Hymn of the Tiger Mother, which lauds the relentlessly and aggressively pushy parenting style, and with the idea that early and narrow devotion to a single pathway is the route to success. According to this narrative, it’s all about headstarts for our kids and exploiting what talent they have; exploring alternatives is just a distracting waste of time. It’s an approach that we often see at school; where parents are concerned to plan ahead, and ensure that IGCSE choices align with future dreams of rising to CEO status.

Despite the apparent logic here, Epstein convincingly demonstrates that this is a mistaken approach for most of us; and not just because we lack Tiger’s talents. He shows that many, many people at the top of their field had a much more eclectic approach (Roger Federer being the most striking example); but more importantly, that the narrow approach only works in limited domains. In golf, the problem of hitting a ball with a club is extremely difficult, but it is solved “according to rules and within defined boundaries, a consequence is quickly apparent and similar challenges occurs repeatedly. Drive a ball and it either goes too far or far enough; it slices, hooks or flies straight.” In these constrained, predictable and repeatable situations, deliberate practice with immediate feedback is likely to be highly effective. In terms of the diagram above, it’s suggesting that the field of possible knowledge/skill is finite; the whole rectangle is actually fairly small so it’s possible to ‘move the line’ rightwards by deep drill and practice. But in most cases, life’s not like that. In many cases (see my list of questions earlier) as Epstein says, “the rules are often unclear or incomplete, there may or may not be repetitive patterns, and they may not be obvious and feedback is often delayed, inaccurate or both”. That's especially true in a world that is increasingly characterized as volatile, uncertain, complex and ambiguous; where old habits are being disrupted and the rate of change is accelerating.  This means, I think, that in the rectangles above, the unknown portion to the right is effectively infinite.  Epstein is, I think, arguing that in these cases, it is just not possible to progress by exploiting what you have; that you nearly always need new ideas and to explore further.

Range is a rich book, and while Epstein does not explicitly use the exploration vs exploitation lens, it’s the broader abstraction of his themes. Gladwell-like in his, ahem, range, he tackles:

• the whole idea of giving our kids a head start (what does that really mean in a disrupted world?);
• when you would be better off giving up and starting something new
• our inherent cognitive biases and what we can do about them
• the notion of disruption as central to breakthroughs. 

In each case he marshalls anecdotes and data to show the benefits of being able to make links across disparate areas; to make analogies across disciplines; and to have many sources of creativity. His overall thesis is one any parent thinking of early specialization for their children should consider carefully: “While it is undoubtedly true that there are areas that require individuals with Tiger’s precocity and clarity of purpose... we need...people who start broad and embrace diverse experiences while they progress. People with range.”


References

• Conceptually.org (2019) What is the explore-exploit tradeoff? - Explanation and example
• Epstein, D (2019) Range. Why generalists triumph in a specialised world. Riverhead Books.