With Super Bowl XLVI less than a week away, armchair quarterbacks are flexing their biceps to throw hard-earned cash into the office pool. You will probably do what a lot of people do, and rely on your gut to make your picks. You might step it up a notch and research the odds being laid in Vegas. You could try consulting a Ouija board. Or, you could apply the laws of statistics like the pros.
What would that look like? We asked a few experts to construct a winning game plan with a big Super Bowl pool on the line. It’s up to you to decide which strategy will give you the edge you are looking for, or if it’s better to just pick your bets at random.
The grid pool is based on 100 squares, each one representing the final digits in both teams’ scores. Pool participants pick squares by random or draft, and winners are awarded quarter-by-quarter, with the grand prize going to the person who nails those digits in the final score.
If your office pool is based on a random draw, fuggedaboutit. But if you can pick the digits, then “there is someone handicapping what goes into this, and 3,4,7 and 0 are the magic numbers,” says Bill Ordine, an editor of http://PhillyGambles.com, which covers fantasy sports and sports betting. “Some pools even charge a premium for those spots.”
Those numbers work, Ordine says, because they reflect point totals from the most common ways of scoring: touchdowns and field goals. If those numbers get taken, Ordine also suggests 1, 8 and 6 as second-tier picks.
You’ll find no end of grid gurus on the Internet, including those with models that reflect the score of every Super Bowl quarter in history. Trouble is, “the game has changed a lot, and there’s a lot more offense than in the 1960s and 1970s,” says Ben Motz, a full-time lecturer at Indiana University who plans to inaugurate a new sports-cognitive science course this fall. Its name? “Prediction, Probability, and Pigskin.”
“So it’s not so much taking the digits from past Super Bowls, but the digits from recent games involving the Giants and the Patriots,” Motz says. “Neither of these teams have great defenses, and it’s likely to be a high-scoring game.”
PICKING THE FINAL SCORE
San Diego State University information systems professor Dr. Jim Lackritz, a statistician and sports nut, calculates the probability of picking the Super Bowl’s final score at no better than 400-to-1. That’s assuming you start with a realistic range of point totals for both the Giants and Patriots, run through all possible combinations and then throw out unusual or impossible tallies, such as ties.
While mathematically possible to nail the score, “you’re not going to be able to do it,” Lackritz says. “I used to do a lot of research and modeling on NFL games and never got models that were sufficient to beat Vegas on a consistent basis.” He laments that one time he tried to number-crunch his way to victory in a college football pool, but was vanquished by a fellow academic who didn’t know a pigskin from a platypus.
“He was from the Middle East,” he recalls.
Despite the Hail Mary odds, it could be worse – much worse. Lackritz once calculated the odds of picking a perfect March Madness bracket at one in 10,000,000,000,000,000,000. (That’s 10 quintillion, for those of you who can count that high.)
Another form of football gambling, sometimes used as a tiebreaker in pools, involves picking the winner and whether the point total goes above or below a threshold determined by Las Vegas oddsmakers. That number currently stands at 55, and it almost boils down to guessing the outcome.
“When you’re looking at numbers and trying to do predictions, you need lots and lots of numbers, lots of data,” says Dr. Irvin Lustig, director of mathematical software for IBM and a football fan who attended the 2003 Super Bowl. “Now if the Patriots and Giants played 30 times a year, you might have enough data to make a prediction. But one time a year? That’s not enough.”
“There is evidence that bettors like the ‘over’ bet because they prefer high-scoring games,” says Tracy Rishel, associate professor of management at Susquehanna University in Selinsgrove, Pennsylvania. She should know: Rishel studied 190 NFL games in co-authoring the paper, “The Determinants of Scoring in 2010 NFL Games and the Over/Under Line.”
“Bookmakers are fully aware of bettor biases and may adjust the lines to take advantage,” she says.
Then there’s a family history of beating the odds. Motz, a contributor to the NFL online video series “Football Freakonomics,” remembers when his father took home Super Bowl pool winnings in the late 1990s. This inspires him to at least take a crack at concocting a method for a repeat, and indeed, Motz sounds convincing – that is, up to a point.
“I would take New England and their record for the past three seasons, and what they are like when they play teams like the Giants,” he says. “And I would not take the average score, but the median score. With the average, you get a decimal. And it should be the case that the mode is the median, but who knows?”
Which brings us back to the way all Super Bowl games start in the first place.
Asked how he would make sense of all those digits in his calculations, Motz replied: “I’d have to take a look at it, and flip a coin.”
(Editing by Beth Pinsker Gladstone; Editing by Dan Grebler)