## Super Bowl Game Probability and Voodoo Analysis

The game probability for the Super Bowl is up at the New York Times' Fifth Down.

This week I talk about momentum and 'voodoo' analysis:

A boulder rolls down a hill and gains momentum. A spark sets a fire, and soon it has built into a blaze. The rains come and soon the river is rushing over its banks. Momentum is everywhere in nature, but applying it to abstractions like team win-loss records in a relatively small sample of football games is what I call voodoo analysis.

Voodoo analysis is the application of apparently intuitive patterns beyond their natural settings. A football team is not a boulder rolling down a hill. It’s not a river bursting through a damn. It’s not a spreading fire. Our brains are continuously looking for patterns like these, and often see them even when they’re not there. That’s why we’re better off taking a disciplined look at the numbers from the full season.

### 11 Responses to “Super Bowl Game Probability and Voodoo Analysis”

1. Chris says:

2. Ian says:

Hey Brian,

The probabilities you give for the super bowl didn't match my expectations. Given that SF has a GWP in the final rankings of .69, and BAL has .48, I expected SF to be about a 70-30 favorite.

How is the discrepancy explained?

3. Brian Burke says:

Baltimore's GWP has improved in the post-season. SF has been expected to do well, and they did. But BAL has done unexpectedly well against 2 highly-ranked teams, especially the defense--Don't forget 2 of DEN's TDs were ST returns.

4. Ian says:

Thanks Brian. One more question: on the 5th down article, the efficiency table shows SF's defensive run SR% as 52%. On the matchup page on ANFLS the same parameter is 62%. I think it should probably be 62% based on end of season values.

Does that alter the super bowl projection at all?

5. Brian Burke says:

Ian--No. I made a typo in the table. (I have to manually type out the SR numbers.) It should be 62. Sorry for the confusion.

6. bigmouth says:

7. Anonymous says:

First I was thinking Ravens had the best chance of winning because of last year's matchup. Sure its now a different QB but you look at the history of scrambling QBs in the Super Bowl, only Steve Young has won. And yes a very similar company with the Niners. Everyone else says 49ers even Pete "More Wrong than Not" Prisco of CBS Sports but the most of the rest of the CBS staff is picking Ravens.

Probably the best gauge is Mike Golic of ESPN. This guy is an astonishing 9-0 on his ESPN straight up picks in the playoffs since getting the opening game wrong. His pick for the winner: Ravens

8. Anonymous says:

Kind of like the hot hand, somehow hot hand exists in individual sports but not in team sports.

Momentum is heavily priced into the live betting markets of some individual sports like tennis. If momentum did not exist it would not be heavily priced into it. Sort of like 52% of all games are decided by chance, if true one would only need bet every underdog every week. Sort like how hot hand exists in individual sports like bowling, golf et al but not in team sports therefore it is random chance.

9. Anonymous says:

Delete posts, good response!

Momentum is actually priced into live betting markets of certain sports like tennis but largely doesn't exist?

Like hot hand exists in individal sports but not team sports and therefore doesn't exist.

Although regardless of entity(team, player, horse, dog, whatever) last out is the most predictive performance followed by second last etc...
what happened 5 months ago is more relevant?

10. Mike M says:

Totally disagre with this. Pointing out the 2 Giants teams that got lucky and that luck continued in the SB, well, you kind of forgot a number of other teams.

Patroits/Rams was a huge statisiticall mismatch in which the Patroits were lucky again in the SB.

As was Arizona, again a huge statistical mismatch which Arizona had no bussiness getting to the SB and if not for some extreme bad luck in the SB we would be talking a different winner.

The Ravens have pretty much elavated their game since changing the offensive coordinator and some offensive lineman changing positions and getting healthy.

The Ravens have proven the flaws in this model, and by the way this is not a critisism as all model have a flaw or 2, but I suspect the Ravens will keep this thing close with a chance to win in the end and should be the right play getting 4.5 points.

11. Mike M says:

Must have been the WORST TIMING IN THE HISTORY OF THE INTERNET to write an article prior to the wild card round as to how right the model was on the Ravens.

Stating their 9-2 record was flawed and the fact the Ravens went 1-4 over the final 5 games showed the power of the model, ouch, that one hurt.

When the model does something good out come the supporters with their rose colored glasses, and when it blows-up like it did on the Ravens they are no where to be seen.