Wednesday, December 10, 2008

Information Advantage

The NFL's tie-breaking system for playoff spots is admittedly confusing--apparently even for bookmakers.

Here's a sample of some 'to win division' lines that were available around the internet earlier this week:

Miami +297
Pittsburgh -320
Minnesota -240
Tampa Bay +275

All of these lines are substantially worse now, largely because I've been betting them heavily. (I posted my estimates of the true odds over here.) Still, the point is that these are the lines the bookies thought were "correct".

What do these four teams have in common? They all own substantial tiebreaker advantages, and the books apparently didn't realize it in each case.

The exact scenarios are complicated, but here's one example: The Vikings split their head-to-head matchups with the Bears. They're 4-2 in the division, so Chicago (3-2) can't gain the advantage there. Minnesota currently owns the edge in games against common opponents, 7-4 to 4-5. The only way the Bears can catch up in that column (without overtaking the Vikings in the standings) is for Chicago to win out while Minnesota goes 2-1, losing to the Falcons. In this scenario, the Vikings win the division on the basis of their 8-4 conference record to the Bears' 7-5. The Bears win this tiebreaker exactly never.

In the NFC North, there's another factor at work. The Vikings still have to face the Giants in Week 17, but it looks likely that New York will be taking it light that game--the early line is Minnesota -7.5. Perhaps the bookmakers aren't taking this into account when setting the futures lines.

The bookies could always consider hiring consultants. These guys know how to implement tiebreakers in their simulations. Hell, when I told a Bears fan living in Chicago that his team is boned in all tiebreaker scenarios, he replied "Yeah, everyone here knows that."

Well, apparently not everyone.

No comments: