I think that one of the most important changes we’re going to see in lots of professions over the next few years is the emergence of tools that close the gap between the middle and the top–that allow the decision-making who is merely competent to avoid his errors to be reach the level of good.

Malcolm Gladwell writing on his blog.  His musing lately have dealt with finding good basketball metrics, but this post was in response to his article about an algorithm that predicts movie box office receipts from scripts.

Gladwell is right that metrics and decision tools raise the quality of the average decision, but he seems to think it affects most the people right below the top or near the middle.  What it actually does is take the implicit knowledge of the best decision makers and make it explicit for everybody else.  I’m not sure the precise numbers, but if you asked a group of people if they were in the top 10% of their profession, something like 70% will say that they are, even though by definition, only 10% can be.  Everybody thinks that they can make better decisions without structure, advice, or metrics, but reality tells a different story.  This was one of the key findings I took away from Gladwell’s book Blink.  In practice, the biggest issue is that nobody wants to admit that their complex “art,” like picking hit movies or music, making business decisions, or medical services, can be reduced to a simple algorithm or decision tree.  But on average, that algorithm does better than the specialists.

There are two areas where I think you need to watch out with the movie algorithm.  First, movie revenues are heavily influenced by a number of factors that are decided by people today - opening size, marketing budget, timing.  The biases of the studio execs are all inside these factors, which means that they are biasing the results of the neural network.  IT could be argued that in fact, these things don’t matter, but I’m not sure if the current network controls for them.

The second issue is profitability.  Just because a movie will generate revenue, doesn’t mean it’s the most profitable investment.  A studio gets a better return with a $5 million film that generates $50 million in revenue than a $50 million film that generates $200 million in revenue, and if they can make ten of the cheaper films, then they’re able to make more profits with lower risk.  I see Hollywood moving, begrudgingly, in this direction, and entrenched organizational structures, distribution partners, and false preconceptions as the biggest osbstacles.

As an aside, the actual Gladwell article is worth reading for the description of Kamesian thinking and Dick Copaken alone.

Categories: Entertainment Media, Information Asymmetries, Information Economics