Since 2001, New England had made the playoffs twelve times, reaching the Super Bowl, incredibly, in half of those seasons. The Patriots have won the Super bowl four times over this fourteen-year span, which made me wonder: how many Super Bowls *should* the Patriots have won?

This could be measured in a few ways. We could look at say, the team’s pre-season odds of winning it all each year. I don’t have that historical data, but we can be sure that New England significantly overachieved by that measure. We could also look at the team’s Super Bowl chances at the start of each post-season. For example, at the end of the regular season, Bovada had the Patriots at 3/1 to win the Super Bowl. That would imply a 25% chance of winning it all, although after adjusting for the vigorish, the Patriots’ true odds would have been 21.8%. I don’t have historical data of this sort, although I am sure one could use a combination of SRS and home-field advantage to come up with something similar. Hey, if you have ideas, present them in the comments.

Instead, I used the same methodology I used a couple of weeks ago to determine the randomness of each post-season. Remember, a point spread can be converted into an expected winning percentage using the following formula in Excel (if you put the point spread in cell L2):

=(1-NORMDIST(0.5,-(L2),13.86,TRUE)) + 0.5*(NORMDIST(0.5,-(L2),13.86,TRUE)-NORMDIST(-0.5,-(L2),13.86,TRUE))

In the playoffs, New England was a 7-point favorite against both Baltimore and Indianapolis, which translates to a 69.3% chance of winning. The line in Super Bowl XLIX ranged from New England -1 to Seattle -1, so I will just call that game even and imply a 50% chance of winning. That means the Patriots had a 24% chance of winning the Super Bowl. ^[1]Note that this jibes with New England’s Super Bowl odds from before the playoffs began, as the books factored in the more-likely scenario that Denver, not Indianapolis, would have been New … Continue reading

Now, things have not always been so easy for the Patriots. New England was a 3-point favorite against Oakland in the Tuck Rule game (58.6%), but was a 10-point underdog in Pittsburgh in the AFC Championship Game (23.5%) and a 14-point dog against the Rams in the Super Bowl (15.6%). That means the Patriots had just a 2.2% chance of winning the Super Bowl in 2001.

Life was easier in ’03 and ’04. In 2003, the Patriots were favored by 6, 3.5, and 7 against the Titans, Colts, and Panthers, which means New England had a 27.7% chance of winning it all that year. The following season, New England faced a murderer’s row of Indianapolis, Pittsburgh, and Philadelphia, and the Patriots were favored by just 1, 3, and 7 points, for a cumulative championship probability of 21.5%.

So in 2001, 2003, 2004, and 2014, the Patriots should have been expected to win 0.754 Super Bowls. Instead, the team won 4.000. That’s pretty good. But what about the other eight years?

The two other Super Bowl years are the easiest to analyze. In 2011, the Patriots had a 33.9% chance of winning it all: New England was a 13.5-point favorite against the Tim Tebow Broncos, a 7-point favorite against the Ravens, and a 3-point favorite against the Giants. In 2007, New England had an insane 57.5% chance of winning it all! Those Patriots were favored by 13.5, 14, and 12.5 points against Jacksonville, San Diego, and New York.

The other six years require more guesswork, although in three of those seasons, the Patriots made it to the AFC Championship Game.

• In 2013, New England was favored by 7 points against the Colts in the division round and then 5.5-point underdogs in Denver in the AFCCG. There was just a 24% chance New England would win those two games. What would the spread have been in the Super Bowl? It’s tough to say, of course, but a Patriots win over Denver would have surely enhanced the New England mystique. Given that Denver was a 2-point favorite over Seattle and a 5.5-point home favorite over New England, let’s just say this Super Bowl would have also been a push. That means New England had a 12% chance of winning Super Bowl XLVIII.

• The prior year, the Patriots were favored by 9.5 against the Texans and 7.5 against the Ravens. There was a 52.% chance New England would make it to the Super Bowl, where they would have faced San Francisco. In the regular season, the Pats were 5-point home favorites against the 49ers, while San Francisco was 4.5-point favorites against the Ravens. Taking the easy way out again, this game would have likely come in very close to even (although I suspect New England would have probably been a slight favorite). That puts the 2012 Patriots at 0.266 expected Super Bowls.

• In 2006, the Patriots actually took the long road to get to the doorsteps of the Super Bowl. New England won as a 9-point favorite in the wild card round, a 5-point underdog in San Diego, and then was a 3-point dog in Indianapolis in the AFCCG. New England was favored by 3.5 points in a regular season home game against the NFC Champion Chicago Bears that year, but given that Indianapolis was favored by 6.5 in the Super Bowl, that seems closer to how things would have gone (also, keep in mind the Super Bowl point spread inflation). Whatever number you pick won’t mean much, as New England had just an 11.0% chance of reaching the Super Bowl. Let’s say New England would have been favored by 5-points over the Rex Grossman-led Bears; that would have given the 2006 Patriots a 7.1% chance of winning the Super Bowl.

Those were the easy ones; we have to engage in even more speculation to analyze the team’s other three postseason appearances.

• In 2010, the Patriots had the best record in the NFL at 14-2, and by far the top SRS grade (+15.4). New England, though, lost the team’s first playoff game in the division round to the Jets as 9.5-point favorites. In the AFC Championship Game, New England would have hosted Pittsburgh. In the regular season, New England blew out the Steelers in Pittsburgh, 39-26. The Steelers were 4-point home favorites against the Jets; In 5 home AFC Championship Games, New England has been favored by 7, 7.5, 7, 14, and 3.5 points. I’ll suggest that New England would have been a touchdown favorite over Pittsburgh, and then 4-point favorites over Green Bay in the Super Bowl. That translates to a 32.0% Super Bowl win probability for the year, which may even be low. Before the playoffs began, New England was 13/10 to win it all, which would imply a 33.7% chance of winning it all, after adjusting for the vig.

• In 2009, New England lost as a 4-point home favorite to Baltimore. In the second round, the Patriots would have faced the Chargers, who in reality were 9-point favorites against the Jets. It’s hard to imagine New England ever being that big of an underdog, so let’s go with 6 points. That would have led to an AFCCG date with the Colts, and New England probably would have been 6 point dogs again. A win there would have led to a Patriots/Saints Super Bowl. That would be close, but let’s say the Saints would have been 2-point favorites. Add it all up, and the Patriots would have had just a 3.0% chance of winning the Super Bowl. One source I found says New England was +1100 to win the Super Bowl on January 5, 2010, after it was announced that Wes Welker had a torn ACL. After removing the vig, that probably puts it at close to 5-6%, and that doesn’t account for having to play Indianapolis and New Orleans, so 3% is probably a good enough estimate.

• That leaves just one year left: 2005, the year Champ Bailey returned a Tom Brady interception 100 yards… to the 1 yard line. New England was an 8-point home favorite against Jacksonville and a 3-point road dog to the Broncos in the first two rounds of the playoffs. That translates to a 29.8% chance of making it to the AFC Championship Game, where the Pats would have faced the Pittsburgh Steelers, the team New England beat in the same round a year earlier. New England had also beaten Pittsburgh in Pittsburgh earlier in the regular season, but this game would have been in Foxboro since the Patriots were division winners and the Steelers were a wild card team. The SRS says the Steelers were better, but let’s not forget the Patriots were two-time defending champs, and Brady/Belichick would have been 11-0 in the playoffs. I’ll suggest that New England would have been 3-point favorites against Pittsburgh, since New England was a 3-point dog in Denver and the Steelers were 3.5 point underdogs in the actual AFCCG. Against the NFC Champion Seahawks, Pittsburgh was 4-point favorites. Let’s put New England at that mark as well, since the 12-0 Brady/Belichick hype train for a three-peat would have been out-of-control. That gives the ’05 Patriots a 10.2% chance of winning it all.

And we’re done! Sum the Super Bowl probabilities for New England in each postseason, and the Patriots should have been expected to win about 2.6 Super Bowls from ’01 to ’14, instead of the 4.0 they actually won.

A weird-looking graph, to be sure, but you can get the point: New England massively overachieved from 2001 to 2004 (other than missing the playoffs in 2002), and then the team underachieved in the postseason for about a decade, winning no Super Bowls despite being expected to win about 1.8 of them. Then, New England won it again in 2014. Not exactly breaking news, but it’s still useful to quantify things as best we can.