Truly Useless Info

I usually enjoy Jayson Stark's Useless Information columns. Today's wild card one is a bit silly:

Here's the breakdown of where the 14 wild-card teams since 1995 stood as they plowed into August.
· In first place or wild-card lead: 9 of 14 (64 percent).
* Within 1 game of wild-card lead: 1 of 14 ('99 Red Sox, 1/2 out).
* Within 11/2-2 games of wild-card lead: 1 of 14 ('96 Dodgers, 2).
* Within 21/2-31/2 games of wild-card lead: 1 of 14 (2001 A's, 31/2).
* Within 4-41/2 games of wild-card lead: 1 of 14 ('96 Orioles, 41/2).
* More than 41/2 behind: 1 of 14 (2001 Cardinals, 6).
So only two of those 14 teams (14 percent) came from more than 31/2 back And just three (21 percent) came from more than two games behind.

Basically, sports reporters do not understand the basic concept of probability but they'll spout off predictions based on, say, an 0-for-8 that a batter has against a certain pitcher. Who cares? If you flip a coin ten times and it comes up heads seven of those flips, what does that tell you about the probability of next flip being heads? Nothing. It certainly isn't 70%. The probability is 50% heads always. The probability is determined by a set of circumstances that are independent of the coin flip history.

What does it tell you about the Mets' playoff chances if in 14 tries, no team has come back from a 5-game deficit to win the wild-card? Nothing. The probability is dependent upon how the Mets perform, how the teams they trail perform, intersecting schedules, luck, etc. But since the '96 Rockies trailed the Cardinals by 5 games in the wild-card (I'm making this up) and failed to win it, the Mets have a harder row to hoe?!? At least that's what they'll tell you. Maybe that's why they still don't get OPS.

Numbers can be informative, numbers can be interesting, but very rarely can numbers be good predictors.