clock menu more-arrow no yes mobile

Filed under:

So far, so good ... so what? Do preseason defensive team stats have any meaning?

Our first string defense has looked elite during the preseason. Does that tell us anything about how it will fare in the regular season? Read on to find out.

Chris Humphreys-USA TODAY Sports

As I type this, the Broncos are tied for fifth in the league this preseason in points allowed (14.0 ppg). We are currently third in yards allowed per game, second in passing yards allowed and first in sacks.

That's the good news.

The bad news is that it might not mean jack squat.

Let's look at some numbers to see if preseason defensive performance has any correlation to regular season defensive performance.

First, let's get the explanation out of the way: Preseason stats are mostly put up by the players who AREN'T going to be playing much during the regular season. Therefore, the numbers we see are generated by the second string facing the second string and the third string facing the third string. The instances of our first team defense facing the opponents' first team offense are few. So that is why (as you will see) there is so little correlation.

And in case stats speak is not your thing, any number closely approaching 1.0 would represent a strong correlation between the years being compared while anything below 0.5 is essentially worthless.

Is there any defensive stat that shows a correlation from preseason to regular season?

The answer is no, not really. Even when you find a year with a modest correlation (in 2014 the correlation between preseason yards allowed and regular season yards allowed was 0.52), it doesn't hold up year after year.

Going back to 2006 (earliest that NFL.com has preseason stats), the average correlation for that stat is 0.18. I evaluated every defensive statistic going back to 2006 and found that the best average correlation (still not very good) was with preseason completion percentage allowed and regular season completion percentage allowed, which stands at 0.46.

So maybe there is a correlation at the top?

Let's see if the top 10 teams in the preseason, in terms of points allowed per game, correlates to the top 10 in the regular season. Why is this interesting? The Broncos had the best scoring defense in the 2014 preseason - allowing only 9.2 ppg. Unfortunately for us, we finished the regular season 22nd in the league in points allowed, well out of the top 10.

In fact, of the 10 teams at the top of the NFL in the 2014 preseason, only three finished in the top 10 for the regular season (Detroit, Blatimore and Arizona).

For points allowed here is the percentage of teams that finished the preseason and the regular season in the top 10 (no data available for 2009 at NFL.com):

# of teams finishing in the top 10 in preseason AND regular season in points allowed:

Year #
2014 3 of 10
2013 6 of 10
2012 3 of 10
2011 6 of 10
2010 5 of 10
2009 ?
2008 3 of 10
2007 2 of 10
2006 4 of 10

So in a good year, more than six out of 10 teams are in the top 10 at the end of the preseason and regular season. However, in the average year, six out of 10 who finished top 10 in the preseason, DO NOT end up in the top 10 at the end of the regular season. That's sobering.

OK, so what if we look at another stat - sacks, for example? We like sacks (not just because we have gotten them at an amazing rate in the preseason).

League-wide over the past decade, the correlation between preseason sacks and regular season sacks is almost non-existent (0.10). We could explain this by pointing out that defense is always ahead of the offense at the beginning of the season, mainly because many teams have offensive lines that are in flux - leading to more sacks.

So what about the top 10 teams? They can't have gotten all of those sacks in the preseason because of inept OL and QB play, can they? Here's the table of the top 10 sack teams in the preseason that also show up in the top 10 at the end of the regular season.

# of teams finishing in the top 10 in preseason AND regular season in sacks:

year #
2014 3 of 10
2013 5 of 10
2012 4 of 10
2011 5 of 10
2010 4 of 10
2009 ?
2008 3 of 10
2007 2 of 10
2006 6 of 10

Again we see that even in a good year only six of 10 teams finish in the top 10 in sacks at the end of the preseason and regular season.

Going back to 2006 (again missing data from 2009), we did not finish in the top 10 in preseason sacks but finished in the top 10 at the end of the regular season three times (9th in 2014; 1st in 2012; 10th in 2011). Through three preseason games this season, we have 15 sacks. FWIW, here are our preseason sack totals from the last decade:

Broncos' preseason sacks totals (through four games) since 2006:

sacks
2014 9
2013 8
2012 8
2011 9
2010 7
2009 ?
2008 9
2007 9
2006 8