The Recency Effect – David Ortiz Edition

Tonight’s World Series game 6 features a truly epic World Series performer in David Ortiz who plays First Base and DH for the Boston Red Sox.  Ortiz is batting .733 in the series against the Cardinals while the rest of his team is batting .144 (sad trombone).  Ortiz has 11 hits while the rest of his team has just 22.  That’s an insane statistic.  Of course, it’s also based on just 15 at-bats.  But Ortiz has done this before.   As ESPN notes, this is a guy who seems to show up when the lights are on him.  In 13 World Series games Ortiz is batting .465.

But is this statistically relevant or is this a case of the recency effect cropping into our heads?  We see this in markets all the time.  People look at recent performance and start to make assumptions about future performance.  Thing is, it’s almost always misleading.  And if we look at the career of David Ortiz we find that he’s not a .733 or .465 hitter, but a .287 career hitter.  Not bad, but certainly not the lights out stud you see in the smaller data sets.

The thing is, when we stretch the data set we reduce the variance and when we reduce the variance we increase the probability of being right about understanding future performance (although, obviously, you don’t eliminate outliers).  So, the question is, should we expect Big Papi to get 3 hits in 4 at-bat tonight and walk off with the World Series ring AND MVP?  Statistically, one game is almost impossible to predict.  That would be like trying to predict the market’s performance tomorrow based on last week’s performance.  That’s just silly.  But if we could stretch this bet out over the course of thousands of future bets the odds are that Big Papi’s batting average will look much worse than it does in just 15 at-bats.  You already see that when you look at his career World Series stats.

Perhaps more importantly, we have to view tonight’s game as its own individual statistic.  This is the problem so many market participants have when they view the stock market.  They don’t view each economic environment and each business cycle as its own unique occurrence.  Instead, they make misleading extrapolations from the past into their current decisions.  Tonight’s game featuring Cardinals pitcher Michael Wacha makes things even more interesting.  Wacha gave up 2 hits to Ortiz in game 2.  Throw in the element of elimination, Fenway Park and you create a confluence of probabilistically impossible events to predict.  As you can see, this all makes for an extremely complex view of the future.  Predictions are really hard.

That said, if I were a betting man, and yes I am a betting man, I would bet that David Ortiz does not get more than one hit tonight.  Statistically, tonight’s game is irrelevant, but if the long-term is a series of short-terms then the wise man would bet into the inevitable future that .733  is not the norm.  He would also bet that .465 is not the norm.  The recency bias should play not play into our future predictions.

NB – I mostly just wrote this post to entertain myself tonight.  Feel free to mock me when David Ortiz hits 3 home runs.  🙂