Why stats show Dallas Cowboys will make playoffs
First, let me start by addressing by the fact that I have not, uh, addressed the Cowboys-Lions game. Over the last couple years, I have not hidden the fact that I am not a “media” source but rather simply a Cowboys fan with a love for statistical analysis and a drive to study the ‘Boys in as much depth as possible. When they lose, the cold, analytical side with which I tend to run this site disappears. During the most difficult losses, I find that my ability to analyze the team in a manner that is both rational and fit for Dallas Cowboys Times is compromised. At this time, I have just one thing to say about the Cowboys’ loss on Sunday. . .
On 4th and Goal from the half-yard line, Jason Garrett called a play from “Double Tight Right I Right.” It was a strong side dive. It got stopped. The Cowboys lost the game by four points. While I realize Tony Romo played one of the worst games of his career (not in terms of actual performance, but rather decision-making), the Cowboys will not be an elite football team until this kind of predictability stops. Garrett thinks his offense should always execute, even if the defense knows what it is coming.
While it is true they “should” perform up to a certain standard, it is impractical to think they will. There are professionals on the other side of the football, and calling plays is about putting your guys in the proper position to maximize their ability to use their talent and succeed. Until then, the Cowboys will simply be an underachieving, poorly-coached group of very talented players. Does Garrett have the potential to succeed? I think so. Must he change? You bet.
Last year, I wrote an article called “Why stats show Dallas Cowboys will make playoffs” following the team’s 1-3 start. It didn’t exactly work out as I expected. Today, I am posting an article called “Why stats show Dallas Cowboys will make playoffs.” Let’s see how it plays out this year.
Over at Advanced NFL Stats, the Cowboys are currently ranked as the No. 1 team in the NFL. No. 1! While I certainly don’t find that ranking to be true, it shows that the Cowboys have at least played well in relation to their 2-2 record. Remember, record is a rather poor indicator of how teams are performing (fortunately or unfortunately, depending how you view it). Points are superior, but even they can be misleading. Through four games, the number of points a team has scored or yielded could differ by as much as 50% or more simply from “luck” (special teams touchdowns, field goals, penalties, and other things which have historically shown to not be repeatable. . .or at least not to a significant degree).
Advanced football analytics show that the statistics which actually correlate best to future wins are not past wins or points, but rather adjusted yards per pass attempt and run success rate (the percentage of runs which increase your expected points on a drive) on both offense and defense. The former certainly weighs more heavily than the latter, but run success rate is important too. Actually, the portion of a team’s runs which are “successful” correlates for more strongly to winning than attempts or yards-per-carry.
According to the numbers, the Cowboys have the eighth-worst running success rate in the NFL. That matches up well with other statistics, as Dallas simply hasn’t been able to get much going on the ground, at least not consistently. So how is the team so highly-rated in terms of their win probability? They have been lights out in each other category. No team is running on Dallas, succeeding just 34% of the time–worst in the NFL (making Dallas the best, obviously). Opposing offenses also haven’t been able to throw, racking up just 5.7 adjusted yards per attempt–making Dallas the eighth-most efficient passing defense. On offense, Romo & Co. have generated 7.8 adjusted yards per attempt, ranking sixth in the league.
Certainly factors other than running and passing efficiency go into winning football games, but these statistics correlate better than any others we have. Plus, things like turnovers, sacks, play-calling and so on are all implemented into these statistics (adjusted yards per attempt accounts for the first two, and the numbers are simply a reflection of player talent and play-calling). The numbers have also given Dallas a 67% win probability, ranking them first in the NFL, and a 55% opponents’ win probability, also good for first (meaning they have played the toughest schedule thus far). When you add in the fact that teams have actually struggled against the Cowboys (meaning they have played at least one poor game, statistically speaking, out of four), you can see why the numbers point to the Cowboys’ record improving as the season rolls along.
I’ll end this post with a quote from last year’s article on the same topic, altering just the team’s record:
Through only four games, however, the sample size of wins/losses just isn’t large enough to be conclusively indicative of a team’s talent, nor can it be used as a strong barometer for future success.
Think about it. There seems to be a big difference between a 2-2 team and one that is 3-1 (at least emotionally), but what is that difference in reality? Maybe a single play in just one game? A shoestring tackle here, a fingertip catch there.
Still, there are those who will claim that the “should haves” mean nothing–the Cowboys are 2-2, and that’s it. How could they be anything other than their record? While I generally disagree with this assessment, it is true in some sense. The Cowboys’ record may or may not be representative of how they’ll play in the future, but whether it is or not does nothing to alter the fact that they are 2-2.
For that 2-2 record to change for the better, the Cowboys need to disregard the “should haves” and focus on improving today. If they do that, they should find themselves playing into mid-January (at least). The stats never lie.