STDL: Imperfect numbers continue to say Broncos are among NFL’s best

No matter their focus, whether on sports, politics, or anything else, statistical models have always been targets of intense criticism.

The number crunchers can test their methods all they want, and even show their work as they did for their high-school geometry midterms, yet ultimately, the old-schoolers will say there's no match for experience and intuition.

But by now, even those data-allergic folks have to admit the 2012 Broncos are pretty good, right?

With the squad sitting at 1-2 after three weeks, Brian Burke's data said the Broncos were the third-most efficient team in the league. Three games later, following the historic comeback in San Diego that evened Denver's record at 3-3, they had reached the top of Burke's rankings.

In the two games since, the Broncos shut down the Saints' typically explosive offense and notched a road win against the Bengals when they didn't have their best stuff - overcoming several big mistakes along the way.

These serve as significant markings of a quality team.

Now, whether this was more a matter of the bounces finally starting to go Denver's way, or of Peyton Manning & Co. gaining acclimation with each other, there's no way of knowing.

The advanced metrics could have been wrong all along, and it's just that the Broncos have gradually become a better team as the year has gone on. That's a very real possibility. Even likelier is that the Broncos' three-game winning streak is a product of all those factors - luck, increasing familiarity, and real improvement.

Stats and predictive models are never perfect when assessing real world events (football games too) - not even the one behind Nate Silver's FiveThirtyEight blog, despite his remarkable performance in predicting Tuesday night's outcome.

But as Dave Cameron of FanGraphs stresses while discussing Silver's work, we'd be foolish to ignore the metrics:

But, hopefully, we can note that a model does not have to be perfect to be useful, and perhaps we can move away from the idea that imperfect — and even biased — data should be discarded until it can be perfected.

That’s essentially where the pundits went wrong with Nate’s model. They didn’t like the conclusions, and some of them raised valid concerns about polling data and whether Nate’s adjustments added or subtracted from simpler, more transparent techniques. But to discard the model entirely was silly, and to pretend like the race was a toss-up was simply wrong. Throwing out the imperfect model with biased data was worse than taking it at face value.

In reality, we shouldn’t do either. The models showed their usefulness last night, but they’re still not perfect, and we shouldn’t just blindly accept every conclusion they spit out in the future. But, we don’t need to discard these models simply because we’ve figured out where their weak points are either. It’s not an either/or situation. We can be informed by imperfect models without being slaves to them.

The advanced figures we cite here each week (and often in the Lard) are far from being without flaw, but they help increase our understanding of what's happening on the field. With that in mind, lets see what the numbers say about the 5-3 Broncos as they prepare for a trip to John Fox's head-coaching stomping grounds.

Advanced NFL Statistics

Denver held their place atop Burke's efficiency rankings, and they remain first on offense and second on defense according to his model.

According to Burke, the Broncos have the most efficient passing offense, are sixth-best at both running the ball on offense and stopping the pass on defense, but have fallen into a tie for 14th against the run.

Looking ahead to Denver's remaining opponents, only the Panthers (4th) and Bucs (16th) rank in the top 20 overall.

Pro Football Reference

PFR's Simple Rating System ranks Denver fifth overall, behind Chicago, San Francisco, New England, and Houston.

According to SRS, the Broncos have played the ninth-hardest schedule, have the fourth-best offense, and are tenth on defense.

As for the rest of the schedule, Tampa Bay (10th), Baltimore (14th), and Carolina (15th) are the only future opponents that rank better than 22nd.

Pro Football Focus

Of course, the PFF figures we cite here weekly are not stats - they're grades that account for how each player fares on every single snap.

PFF's grades still say the Broncos are third-best, behind the Niners and Patriots.

Broken down by unit, Denver is also third in offense behind those two teams, eigth on defense, and fifth on special teams.

Tampa Bay (11th), Baltimore (15th), and San Diego (16th) are the only future opponents better than 25th overall.


Denver's Week 10 host is all over the map here, ranking as high as fourth according to Burke and second-to-last by PFF's accounting, with SRS putting them right in the middle.

Burke's accounting gives Denver a slight advantage, giving them a 56% chance of winning after having called them overwhelming favorites against the Saints and, to a lesser extent, over the Bengals.

The betting public agrees with that, making the Broncos a four-point favorite.

With the metrics and grades in slight disagreement with each other, we'll unconfidently predict a narrow Denver victory. If the Broncos are able to pull off a second straight road win, and the Chargers lose in Tampa Bay, Peyton Manning & Co. will have an opportunity to deal something of a death blow to San Diego's playoff chances a week later at SAF@MH.

One prediction we can make with absolute confidence - if the Broncos win, and exit Week 10 at 6-3 and without major injuries, the punditry will be as vociferous in its praise for Denver's SB chances as the politicos were in their criticism of Nate Silver prior to Tuesday night.

Doug is IAOFM’s resident newsman and spelling czar. Follow him on Twitter @IAOFM

Stats That Don't Lie