It gets harder and harder to avoid learning the outcome of a sporting event before you are able to get home and watch it on your DVR.  You have to stop surfing news web sites, stay away from Twitter, and be careful which blogs you read.  Even then there is no guarantee.  Last year I stopped to get a sandwich on the way home to watch a classic Roddick-Federer Wimbledon final (16-14 in the fifth set!) and some soccer-moms mercilessly tossed off a spoiler as an intermezzo between complaints about their nannys.

No matter how hard you try to avoid them, the really spectacular outcomes are going to find you.  The thing is, once you notice that you realize that even the lack of a spoiler is a spoiler. If the news doesn’t get to you, then at the margin that makes it more likely that the outcome was not a surprise.

Right now I am watching Serena Williams vs Yet-Another-Anonymous-Eastern-European and YAAEE is up a break in the first set.  But I am still almost certain that Serena Williams will win because if she didn’t I probably would have found out about it already.

This is not necessarily a bad thing.  Unless the home team is playing, a big part of the interest in sports is the resolution of uncertainty.  We value surprise. Moving my prior further in the direction of certainty has at least one benefit: In the event of an upset I am even more surprised.  This has to be taken into account when I decide the optimal amount of effort to spend trying to avoid spoilers.  It means that I should spend a little less effort than I would if I was ignoring this compensating effect.

It also tells me something about how to spend that effort.  I once had a match spoiled by the Huffington Post.  I never expected to see sports news there, but ex post I should have known that if HP is going to report anything about tennis it is going to be when there was an upset.  You won’t see “Federer wins again” there.

Finally, if you really want to keep your prior and you recognize the effects above, then there is one way to generate a countervailing effect.  Have your wife watch first and commit to a random disclosure policy.  Whenever the favorite won, then with probability p she informs you and with probability 1-p she reveals nothing.