Contrarian Bridgejumping

The interesting thing that was ticking over in my small little brain some years ago was that while many people are focused on a team's net points on the year (as after all, you're betting against a line that revolves around point margins), not a whole lot of attention is devoted to how a team has done on a net basis against the line set on its games. So if a team was favored by six points and won by eleven, most people think in terms of a +11 number for the team in that game...however, you'd credit it with a +5 for it's performance against the line in that event and by summing up a teams's "linejumping" and averaging it out on a per game basis you find out who's over or under performing versus the spread. This gives you a perspective of a slightly different fashion than just looking at spread records, since a team might have a great spread record but be barely covering when it "wins" while getting blown out in the losses. Obviously there's often a fairly strong correlation between the two, as well as a connection to the overall net points, so while a team like the 49ers might be #2 in net points, they're more likely to be something like #12 in net "linejumping" points, and often even worse (being an obviously good team). As the season goes on though this can change radically as by year's end a lousy team may have the best net against the line.

Starting with this premise, and not considering matchups until week 5 (since the early weeks are too easily distorted by one blowout or "pumpkin smashing"), I the last five years of NFL play through my little program and came up with interesting results -- when you find a game where the performance difference against the line is more than six points (eg one team is averaging +2.5 vs the line while the other is -7) then playing the poor performer is the way to go. This sounds of course very much like the turnover difference which was a superior tool for the previous five years overall, although it suffered an off year in '96. This is also where the "Bridgejumping" part comes in.

Horseplayers have long been familiar with the notion of a "bridgejumper", which is someone who bets a vast sum of money on a dominating favorite to show in the hopes of collecting the guaranteed 5% payout (while horse racing is pari-mutuel, no bet can pay less than $2.10 for a $2 bet, so you can dump in a near infinite amount and still be assured of receiving your 5% profit...so long as the horse finishes in the money). These folks really do exist and from time to time you'll hear about someone having wagered $100,000 on Cigar to show and people will mumble that it's not a bad way to pick up a quick five grand. Damon Runyon did however write a very remarkable little short story entitled "All Horseplayers Die Broke", and while it's far from the truth (why my own mother must have a career net of at least +3 or 4 dollars from her $2 show bets), it tells the sad tale of a fellow who takes the odds-on favorite to show only to see it trip on a hole in the track when leading by ten in deep stretch. The theory is that when one of these wagers goes awry, the wagerer finds a suitably high bridge to jump off of.

Contrarian bridgejumping is what we're dealing with here though, since what it amounts to is we'll find a team that has been playing woefully in terms of keeping close to the line and is facing a "public favorite" who's been dominating, and then we'll root hard for that down-and-outer to show some miraculous signs of life.

Now you know the story we'll take a gander at the historical records in this type of endeavor, for this breakdown only looking at games from week 5 on, since this represents the minimum acceptable number of games "under the belt" to go by --

Wager Type                     0 to 6 pt LJ edge         6.1+ LJ edge

Home Favs                       130 - 120                  46 - 78
Home Dogs                        41 - 34                   10 - 10
Away Favs                        54 - 57                   28 - 59
Away Dogs                       116 - 84                   13 - 38

Totals                          341 - 295                  97 - 185 


(Note:  these stats are based on data from 1992-1996)

So, to repeat the basics, we need to compute the average "net versus the line" for a team (and fortunately as I mentioned we'll be carrying these very same computer generated numbers for you here...and trust me, it's a lot of work to try and do this manually) and then look for a matchup with a big schism between the two sides...if you tell me Minnesota has been besting the line by an average of 6.4 points and the Colts have been losing to the line by an average of 3.1, then I'd be ready, willing and able to play the Indy crew, even if Manning has thrown seven interceptions for every ten yard completion!