POWER Rankings: 2007 vs. 2015 Patriots

By Harrison Chase and Austin Tymins

With the Patriots off to a dominant 6 – 0 start, people are already comparing them to the legendary 2007 team that went undefeated in the regular season. Local Boston radio personality Michael Felger has declared that he “would rather have this team over the ‘07 team,” and ESPN is already doing a weekly breakdown of the Patriot’s chances of remaining undefeated. The countdown to  16 – 0 is real. Still, Aaron Schatz contests that this year’s team is nowhere near as dominant as the ‘07 team. So who is right?

We can look at win expectancy curves to test this, as they directly measure the dominance of a team throughout a game. Using the same process as we did last week, we can create average win expectancy curves for both teams and compare them. Note that for the 2007 Patriots I looked at their curve prior to Week 8 for an equivalent comparison. The curves are below.

Clearly, both teams have been very dominant. However, it is clear that the ‘07 Pats have, so far, been much more dominant than this year’s Patriots. To put some numbers on all of this, this year’s Patriots have had a win expectancy of, on average, 86.2% at the start of the fourth quarter. The 2007 Patriots, eight weeks in, had a win expectancy of 98.7% going into the last quarter. That is just ridiculous. The 2007 Patriots won each of the first seven games of the season by at least 17 points while the 2015 team has only done so three times this year. This complete dominance is reflected in the win probability graphs above.

Of course, there is the argument that maybe this year’s Patriots have played harder teams. To look at this we can generate power rankings not only for right now, but also for what they would have looked like going into Week 8 in the 2007 season. Doing this, we see the 2015 Patriots have a ranking of 0.7788, while the 2007 team comes in at a staggering 0.8926. Again, that is just ridiculous. Because of this we can safely say that up to Week 8 the 2007 Patriots were far more dominant than the 2015 Patriots. We can’t necessarily say that they were better – as they are in different years and a secular shift in the talent level of the game could have changed things over this period.

However, we think it is reasonable to conclude that the talent level in the league is fairly similar. By making this assumption, we can use their rankings to generate the probability of the 2007 Patriots beating the 2015 Patriots if they were to play. On a neutral field, our rankings suggest the ‘07 squad would have a 70.2% chance of winning, meaning that would be nearly 7.5 and half point favorites. Think about how good the Patriots have been this season. Now think about a team that would be favored by over a touchdown against them on a neutral field and likely by double digits with home field advantage. That’s how dominant the 2007 Patriots were seven weeks into the season.

We have also updated our POWER Rankings going into Week 8. Despite the Pats win, we can see that they dropped to 2nd as they didn’t beat the Jets in dominant fashion. The Packers are now the top team in our model.

Houston, fresh off first quarter destruction at the hands of the Dolphins, drop into a far last place. Denver, another undefeated team, ranks seventh – behind the other undefeated teams as well as Seattle and Arizona which is a product of their less-than-inspiring wins. There are several deviations from our rankings compared to other methods, for example, Kansas City is ranked very high, Pittsburgh a little low, Indianapolis very low, etc.

As mentioned before, there are several things we are considering adding on to the model. We have thought about weighing more recent results more heavily, perhaps using an Exponentially Weighted Moving Average, but we first need to figure out how to optimize that. We also are a little wary of how the Vegas lines affect the rankings – although they might make our rankings make more sense, as they are acting effectively as a prior, but it might perhaps be overstating how dominant teams have been. For example, comparing the 2007 and 2015 Patriots, the 2007 team started as much higher favorites than this year’s team. Is that alone making them appear more dominant?

As always, let us know if you have any thoughts/suggestions/concerns, either over Twitter or via email.