The Bears are big underdogs to the Colts on Sunday. Yet Allen St. John of the WSJ predicts the Bears will win:
Over the last 22 Super Bowls, 18 — 86% of them –
were won by the team that came into the game after allowing fewer
regular-season points. (In the 2004-05 regular season, the New England
Patriots and the Philadelphia Eagles each allowed 260 points.) Not
since the Denver Broncos won 1999′s Super Bowl XXXIII has the team with
the weaker defense emerged victorious.And that’s why the underdog Chicago Bears should
defeat the Indianapolis Colts in Miami on Sunday. Lovie Smith’s NFC
Champion Bears allowed only 255 points during the regular season.
That’s third-best in the NFL — and is a whopping 105 points fewer than
the AFC Champion Colts allowed.
He adds this info:
Savvy fans may dismiss many of these numbers, arguing that the AFC was
the far better conference this year. That may be true for the
conferences, but it doesn’t necessarily hold for this matchup. The
Bears and Colts played five common opponents this season — the New
York Jets, New York Giants, Patriots, Buffalo Bills and Miami Dolphins.
Indianapolis went 5-0 against these teams, while the Bears were 3-2.
However, in those games, the Bears scored 114 points and surrendered
only 75, outscoring their opponents by 39 points. The Colts? Despite
their perfect record, they scored 128 but gave up 107, for a 21-point
differential, just over half that of the Bears.
This somewhat savvy fan would point out that that was then and this is now. The Colts defense has been playing much better. But it is interesting how easy it is to think that the Colts are unstoppable and that Peyton Manning is a genius. We felt the same way about the Rams and Kurt Warner against the Patriots.
I have been thinking a lot about the difficulties of forecasting based on the past because I am reading Fooled by Randomness. Very thought-provoking.
As a final pre-game prep, listen to the last half of the Michael Lewis podcast. (And listen to the first half, too, where we talk about Moneyball. Pitchers and catchers report in less than two weeks.) Lewis’s insights into the role of the left tackle have changed the way I watch football. Or read his book, The Blind Side. You still have two days.
Podcast RSS Feed
Full EconTalk Text
{ 15 comments }
Well, you can't argue with the statistics, but I've read other "studies" that point to the Colts looking back at similar trends.
The argument that sways me for this game is the disparity in QBs. When you have a sure-thing Hall of Famer against an average (or worse) quarterback, the latter is in trouble.
I believe the most recent such matchup was John Elway v. Chris Chandler in the Broncos-Falcons matchup. Steve Young v. Stan Humphries in the 49ers-Chargers matchup was a disaster. Going further back in history, you can look to Terry Bradshaw v. Vince Ferragamo in Steelers-Rams in January 1980.
With all apologies to Michael Lewis' take on the importance of the LT, I think a wide talent disparity at QB has much more of an effect. Now, with my "luck," watch Manning go out and stink up the joint with 4 interceptions….
The Super Bowl is a great "mind" game.
There have been many Super Bowl games that were blow outs because one team let the event overpower them and they just plain played flat and below their potential. We've been spoiled over the last few years by several exciting games,but that is the exception not the rule for the Super Bowl.
So going back to the idea of how hard it is to forecast, because of all the media and mind games around the Super Bowl it is particularly hard to forecast and more then likely it will be won by the team that handles the psychological pressure the best– and I have no idea which one that will be.
Not sure where to send this, so put it here.
Could you get rid of the snap popups on the links? I find them really annoying
Thanks!
Tim,
Thanks for the feedback. I'm torn. I know some people really hate them, so we're thinking of getting rid of them. But I really like them. They give me an advanced look at where I might be headed before I click through. For me, it's an improvement on just seeing the URL at the bottom left–a feature I really like on Firefox.
There are probably some people who click on almost everything and are happy to hit back if they don't like it. Others rather have the hassle of the delay from the pop-up to save the hassle of having to click back.
Happy to hear other thoughts from readers.
While I only took a cursory glance at the article, using point differential seems too simplistic to avoid fallacious conclusions. It doesn't seem to take into account strength of offensive prowess of opponents in a team's schedule , or the fact that with so few data points a couple of outliers within a season can have a large effect on the points allowed differential. Also there is no sense of what a significant points allowed differential is, 3 of the differentials in the chart of 7 seem to imply no strong defensive difference at all(0,-1,-13)
Also it is interesting that he only uses 22 Super Bowls out of 40, perhaps he doesn't have the data, or is cherry picking. Either way I wouldn't bet the house on it.
Fooled by Randomness, I have no doubt, will be one of the best books you read all year.
The Rams should have won in 2002, they just played a terrible game, while the Patriots played solid. Also, if St. John is so sold on Lovie Smith's defense, he might want to keep this little fact in mind: Smith was Saint Louis' defensive coordinator in 2001-2002.
I echo Shakti, Fooled is one of my all time favorite books. I consider it, When Genius Failed (about the rise and fall of LTCM) and Ubiquity (about the ubiquitous nature of the critical state) to be a trifecta that anyone interested in the challenges of studying past data to make future predictions should read.
Any. Given. Sunday.
Betting and stock picking have some similarities in that they are market driven but sit on top of underlying characteristics. The stock market uses company "fundamentals" to analyze companies, but there is a whole heap of speculation and technical analysis thrown on top that shapes the near term.
Betting the spread is similar. The "odds makers" choose a starting point to indicate differences in strength, and everyone has stats available to them. But then the bettors take over. After the initial posting of the odds, subsequent changes in the odds are totally based on which teams get bet on. (The house makes money on transaction costs, and breaks even – for the most part – on balancing the spread.)
A healthy dose of contrarianism can be used when there is a large difference between popular sentiment and odds lines. For example, if the public thought there was a 90% chance that the Colts would win, but the odds were set for the Colts to only win by 7 you'd know that those who are actually betting are significantly less optimistic about the Colts.
In that case, because the "public" was heavily in favor of the Colts, you could say that the "Smart Money" thinks the Colts will by less than 7, and that the lesser-betting but more numerous public thinks it is a 90/10 proposition. In that case, there is a strong argument to choose the Bears.
However, in ESPN's current poll in which 215,000 respondents voted, 64% thought the Colts would win. Which in my experience lines up with about a 4 or 5 point spread. But the actual line is 7 points. This tells me that the betting experts are more pro-Colts than the public is, and therefore, the Colts is the smart bet.
Using this technique, and a Magic 8 ball, I am currently 9-1 in Playoff games against the spread. (The one loss being the Patriots who were covering the spread the entire game until 1:20 left.
)
Statistics in big games is for schmucks……..GO BEARS!
and we pray just decent enough Rex shows up!
I agree with Allen St. John's assessment. The team with the better defense and specials usually prevails.
I Loved Fooled by Randomness, although initially I was a little put off by Taleb's unctiousness.
I loved Moneyball and coach youth baseball with those concepts in mind.
Finally, I check in here as often as I can and have recommended your blog to my economically illiterate (i.e liberal) friends.
Talk is cheap. The Colts are currently favored by a touchdown—would Allen St. John really be willing to place an even-money bet on the Bears? I would gladly accept such a wager. Indeed, I wish I could make a bet with every NFL pregame commentator who picks the underdog to win outright each week. I would of course lose many of those bets (as I might lose an even-money bet on the Colts), but I would end the season a very rich man.
I support the Bears…
But not their mission!
Looks as if Allen St. John was wrong again. This is the same guy who predicted that scoring efficiency (points per yards) was the key in winning the BCS in college football. As such, he predicted Ohio State would win. Had he watched these two teams play, he would've noticed the awesome speed difference between the two teams, especiailly on defense.
You'd go broke listening to Allen St. John.
{ 2 trackbacks }