ESPN 538: SpyGate Hasn’t Affected New England Patriots

Whether it’s the Texans last year, Eagles corner Cary Williams’ sorry-I’m-not-sorry whinefest earlier this week, or Marshall Faulk’s annual “I’m not-saying they did but I’m just saying it’s weird” soapbox every year when the playoffs come around (BARF), the Patriots haven’t been able to shake the shadow of 2007’s SpyGate incident ever since their 17-1 season. Bill Belichick probably still isn’t losing any sleep over it, but in an extremely interesting piece today, the superhuman nerds at FiveThirtyEight decided to fire up their TI-83’s and lay to rest whether taping actually helped New England in their decade of dominance.

Since there are approximately 1 billion variables in a football game, FiveThirtyEight’s Neil Paine first concludes that just comparing Pats wins pre-SpyGate vs. post SpyGate is too sweeping. Amusingly, he finds out that New England’s win percentage before SpyGate was 69.3%, compared to 75.6% after getting slapped by Roger Goodell, but since Paine says there’s way more to it than that, let’s move on.

Next, to account for player turnover and varying levels of offensive talent in the Patriots’ system, like, say, Brady–Moss compared to Brady–Dobson, Paine stacks the points the Patriots scored against the Vegas predictions for how many points they were “supposed” to score. Paine also, to keep his numbers clean, limited the comparisons to regular season opponents that the Patriots played both before and after the SpyGate incident. The result? In both eras, the Patriots beat predictions by exactly 2.4 points. In other words, both before and after they were caught taping signals, on average, the Patriots were racking up 2.4 more points per game than Vegas said they would. Myth busted?

Before the champagne toast, though, there’s that whole groan-inducing argument of “The Patriots haven’t won a Super Bowl since 2004”, so Paine applied the same standards to playoff performances. The results there are, as he describes “right on the edge of statistical significance”; since SpyGate, the Patriots have been under the Vegas predictions by an average of 6.6 points per game. The caveat there is that Paine is only able to compare opponents that the Pats played in the playoffs before and after SpyGate, which limits his sample to 9 games. So games like, say, the Patriots’ manhandling of the Texans a few years ago don’t figure into his calculations.

But before any Jets fans start fist-bumping, Paine’s conclusion is that, as alluded to a second ago, he’s reduced his playoff data to an extremely small sample size. Notice how many times the terms of each calculation had to be clarified? The final verdict (per Neil Paine) is that when you combine regular season and postseason stats, there’s been no difference in the Patriots’ offensive performance since SpyGate. And even using the playoff “filter” we just talked about, Paine concludes that it’s too small of a sample with too many extremely specific conditions to yield any real results, especially when luck and random unpredictability come into play (read: David Tyree). It’s also extremely telling that his penultimate sentence is “And we must always keep in mind that splits happen if we look for them hard enough.”

Translation: when you factor in all the numbers you can, plus take random chance into account, the Patriots literally haven’t changed a bit since SpyGate. The train just keeps a-rolling.