Sports News Without Fear, Favor or Compromise
Sports News Without Fear, Favor or Compromise

Reminder: Betting On Big NFL Favorites Is Always Dumb

Illustration for article titled Reminder: Betting On Big NFL Favorites Is Always Dumb

The Broncos didn't cover their preposterous -28 (-27 at kickoff) spread. Of course they didn't. Because big NFL favorites usually don't cover.

In some ways, this should run parallel to the Any Given Sunday line of thinking—a few fumbles here, a muffed punt there, and all of a sudden you're in a one-score game with a historically bad team. Still, it's easy to be swayed by screaming columns about unstoppable offenses or gross and encompassing incompetence. There are a few some compelling reasons to not listen, though.


For one, they just don't come through. Since 1978, 12-or-more-point favorites are 220-275-9. That's pretty bad. Since 2005, they're 64-94-2. Also bad.

Phil Brinbaum broke it down even further, to 6-or-more-point favorites, and found that while the effect is diminished, even those favorites tend to lose against the spread:

Does the effect disappear for "less heavy" favorites? Again going back to 1978, and looking at teams favored by 6 to 11.5 points ... they were 1361-1475-57 (.480 excluding pushes). Teams favored by 0.5 to 5.5 points were 2197-2330-156 (.485). So, yes, it seems the effect is more pronounced for the heavy favorites.

Yes, at a rest, lines are meant to incite action on both sides, so you could see some dogs getting a few extra points to get anyone to bet on them. But enough to make that big a difference? What else would be going into that number?

One factor could be pace. A common tactic for underdogs in possession-based sports is to limit the number of possessions in a game. Play slowly, use the whole play clock or shot clock. There's solid statistical evidence that this is effective—to a small degree, at least—at increasing your chances at winning outright as a big underdog.


Brian Burke at Advanced NFL Stats illustrated this with the number of possessions between the 2007 Pats and Giants for their Super Bowl matchup that year. At 12 possessions per team (the most common number), the Pats won 75.6 percent of the time, and the Giants won 20.5 percent fo the time (3.8 percent went to overtime) over 10,000 simulations. Lowering that number to 10 possessions changed the win percentages to 72.7 percent and 22.4 percent, respectively. That's a small bump, but it tracks as you add or subtract possessions form both teams. And it's not like it's something you're going to have a hard time selling to coaches, since "slow the game down and hope you get lucky" (and keep the dominant offense off the field) isn't exactly a new idea.

And of course, if you're slowing the game down, there's less room for gigantic blowout scores. It's harder to hang 40 on a team when they're milking the play clock and running the ball (if they're actually making first downs). So even if they don't buy into the projections, it probably helps for players and coaches trying to avoid being humiliated that slowing things up will deflate the scoring potential.


So is it surprising that the Jaguars stayed within 16 of the Broncos? Given how they've played this year, sure, a little. But in the broader picture, it would have been a surprise in itself if Denver had managed to cover its ridiculous spread.

Regressing is Deadspin's new home for sports science, statistics, medicine, and other nerdy endeavors.


Share This Story

Get our newsletter