clock menu more-arrow no yes

Filed under:

Let’s Talk Scoring: Dawg Stats Looks at Metrics + Points

If you have been reading my stats posts all season, you are know that the “success rate” and “EPA” are the base for a lot of the metrics that I break down. But, I’ve added a wrinkle to the metrics that may help bring home some of this “efficiency” talk.

EPA is a value based on a play’s result which is calculated from like a billion plays from college football. Success rate is a yes/no formula of a plays which gain 50% on yards to gain on 1st, 70% on 2nd, and 100% on 3rd and 4th downs.

But, instead of looking at EPA – lets take the success rate formula and calculated it with actual points scored and allowed.

Here are the teams which have the highest number of successful plays per game in Power 5+ (with AAC).

Mississippi State while running a lot of successful plays on offense, they are short quick passes, and don’t translate into points. Also, this rank shows some teams which have good offenses and not so good defense. I’ll get to successful plays allowed if you stick with me.

Now, if I take points scored and average successful plays run, we get a Dawg Stats version of explosiveness replacing Expected Points with Actual Points. Now we can really see offensive efficiency. So, while UGA’s Metrics Explosiveness lags behind, the actual points scored per successful play is right there with the Buckeyes and Bama.

No surprise that Georgia’s defense, shall I say “badass” by any stat one can crunch. The Bulldogs allowed the second fewest successful plays per game behind only Wisconsin. Goes without saying that teams who blowout out opponents might take foot of defensive gas. That doesn’t seem to drag down this defense.

Continuing with the theme of pairing actual points to successful plays allowed, Georgia is absurdly good, aka efficient, allowing .35 points per successful play. It’s almost twice as hard to score to score on this team even when you have good plays.

Here are a couple of early games with these stats compared.