LINKEDINCOMMENTMORE

I'm a nerd. A huge dork who spends all day in front of the computer analyzing football stats. No really, that's all I do. It's great.

There are really two different types of stats out there: explanatory and predictive. Explanatory stats, as the name suggests, explain a past event. So when you hear that Team X wins 90 percent of the time when they run the ball Y times, you're listening to an explanatory stat.

But it's not predictive. You can't predict future winners by analyzing rushing attempts because the latter is an effect of the former; teams typically run because they win, not the other way around.

Explanatory stats are nice, but I'm mainly concerned with finding predictive stats those which can help us better understand the Cowboys, NFL, and what makes teams win. That's one of the reasons I track statistical trends (and think NFL teams should do the same, way more than they do).

One of the errors I see people make when dismissing statistical analysis is thinking that stats should be flawless that if they can't predict everything, they shouldn't be used for anything. But why is that the case? We don't hold those who study film independently of analytics or those who talk about vague, unscientific concepts such as 'heart' and 'savvy play' to such a standard, so why the stat dorks?

All we want in a stat is that it can help us predict the future better than we can right now. That's all. Accurately forecasting the future wins games. And again, we don't need perfection. If a team knows that the Cowboys pass the ball 90 percent of the time out of a particular formation, regardless of the situation, that's useful and actionable information.

With that little primer out of the way, I introduce my first 'team trends' article of the year. These pieces will analyze statistical trends from the Cowboys, their opponent, or both. The goal is to find unique data that will give us insights into the upcoming matchup while also simultaneously and paradoxically increasing both the enjoyment and fury we feel when sitting down to watch the games.

Cowboys vs. Chiefs Statistical Trends

I track every play in Cowboys games, so I have a substantial database of data dating back to 2009. Some of the stuff I'll write about this year is proprietary, but most of it is very much public.

And one of the best public resources I know is Pro Football Reference's Game Play Finder. You can sort through box scores to uncover anything you'd like in seconds. Want to know how frequently NFL teams run the ball on first down inside the opponent's five-yard line? Easy. Prior to games, I spend a lot of time just sifting through that data to see what I can find. Here's some of it.

  • Trend #1: The Cowboys passed on 54.5 percent of their first downs through the first three quarters last year.

    I like to analyze stats through three quarters because it does a decent job of eliminating unusual circumstances. We want to analyze stats when games are point-maximization contests for both teams, not when one is running the ball all the time and the other passing on every play because of the score.

    It might seem like this 54.5 percent pass rate is low, but the league average was just 46.8 percent. And guess what? The Cowboys' first down pass rate should increase substantially. Defenses still usually play to stop the run on first down, so offenses see much, much greater efficiency through the air. It's popular to say that running the ball on early downs can set up manageable third downs, but you know what's even better than third-and-short? Not even facing third down because you didn't blindly run the ball on first and second down.

  • Chiefs Comparison: 33.8%

    This is really a remarkably low number, and one that is going to rise with Andy Reid in town. Reid likes to use the passing game as an extension of the run, and he has a quarterback who can work the underneath passing game in Alex Smith. For what it's worth, Phlly's first down pass rate through three quarters was 53.2 percent in 2012.


  • Trend #2: The Cowboys pass the ball 43.5 percent of the time on second-and-one.


    There's no down-and-distance that represents low risk and high reward quite like second-and-one. Stats show that the down-and-distance is so prized that, in most areas of the field, it's more valuable to gain nine yards on first down than it is to gain 15 yards! Second-and-one allows offenses to take shots in the passing game without much risk, knowing an incompletion sets up third-and-short.

    Yet most NFL coaches are so risk-averse that they avoid passing on second-and-one, opting for 'the sure thing' in a run. They're so scared of third down that they knowingly pass up the optimal play-call to move the chains.

    And even at a 43.5 percent pass rate on second-and-one, the 'Boys are actually one of the better teams in the NFL. The league-wide rate in 2012 was only 31.9 percent.

  • Andy Reid Comparison: 31.1 percent

    Looking at Reid's calls since 2008, we see a second-and-one pass rate that's lower than the league average. That's particularly interesting since he's normally such an aggressive coach. It's a potential area for him to improve in 2013.

  • Trend #3: The Cowboys' play-action rate is increasing, up to 15.7 percent in Week 1.

    Had Romo totaled this rate in 2012, he still would have ranked in the bottom six quarterbacks in the NFL. That shows you just how ridiculous it was that the Cowboys' 2012 play-action rate was 10.0 percent. The difference between Romo and the next quarterback Eli Manning was larger than the gap between Manning and the subsequent 11 passers.

    Oh, but you need rushing success to set up play-action, right? Wrong. Defenders play situations, not past rushing efficiency. There's all kinds of evidence that play-action success isn't at all correlated with rushing success, but here's all you need to know: Romo's 2012 passer rating on play-action passes was 109.1. On regular dropbacks, it was 88.3.

  • Alex Smith Comparison: 23.0 percent in 2012 with a 134.6 passer rating

    Smith has been unbelievably efficient on play-action passes during his career. Last year, he generated the highest play-action passer rating in the entire NFL, according to Pro Football Focus. He attempted a play-action pass on 23.0 percent of his dropbacks.

    The play-action game is a big part of Reid's offense, too. Smith showed play-action on 26.3 percent of his Week 1 passes. If the Chiefs are able to beat Dallas downfield through the air, it will likely be through the play-action passing game.

Jonathan Bales is one of the foremost analytical NFL minds not to have been scooped up by a team yet. Follow him on Twitter for more great insight at @BalesFootball.


For more of our Cowboys content, download WFAA's newest app, Football Dallas! Available free on both Apple and Android devices.

LINKEDINCOMMENTMORE
Read or Share this story: http://www.wfaa.com/story/sports/2014/08/19/14096348/