The President, Luck, and Regression to the Mean

Being particularly lucky or unlucky is sure to interfere with good decision making. It’s hard to tell whether you’re succeeding because of a confluence of favorable effects due to chance or due to your exceptional brilliance. One’s internal model of self probably plays a role in how events are interpreted. Are you special or just really, really lucky?

If success came through chance, the model predicts that the future is going to look more average, for good or bad. It may or may not result in less risk taking. I can take risk knowing that the outcome is largely not in my control. I may win or lose; it’s knowing the odds of success that’s important.

On the other hand, if you’ve climbed to the top based on your merit then the internal model predicts continued success beyond the average, never regressing to the mean.

Case in point? Barack Obama. From Andrew Gelman writing at Frum Forum

More to the point, I don’t think that in January, 2009, Obama had any feeling he was in trouble.  For one thing, he’d spent the previous two years beating the odds and winning the presidency.  (Yes, a Democrat was favored in the general election, but Obama was only one of several Democrats running.)  As I and others have discussed many times, successful politicians have beaten the odds and so it is natural for them to be overconfident about future success.

Losing the 2010 midterms may have been a wakeup call, but then again it’s easy to construct a narrative where we claim responsibility for good outcomes and blame chance or other outside causes for the bad outcomes.

Purely from a statistical point of view, expect success to be followed by failure more often than not. In the end, we’re all average.

Leave a Reply

Your email address will not be published. Required fields are marked *