# Robert Griffin III should pass and scramble but cut back on designed runs

Despite the Redskins’ loss last Sunday, Robert Griffin III has been playing much better than even optimists predicted he would. He’s playing winning football with his arm and legs, racking up 1,601 yards through the air and 486 yards on the ground through Week seven this season. That represents a ratio of well over three-to-one.

But total yards is one of the very worst ways to measure a player’s performance. The best way to illustrate this is by example. Suppose it’s third down and goal from the three yard line. A successful touchdown run yields only three yards. Now suppose a quarterback throws or scrambles for 5 yards on third down and seven. It’s obvious that the play that gained more yards was not nearly as successful as the one that gained fewer yards.

That’s where a stat like Expected Points Added (EPA) can be helpful. EPA measures the change in net point expectancy for any play, accounting for down, distance and field position. In the two examples above, the EPA for the three-yard touchdown run would be strongly positive while the five-yard scramble would be negative.

In the light of EPA, Griffin’s runs contribute to his overall production much more than his total yards would indicate. His passing EPA is 38.2, meaning his pass plays are worth a net 38 points over opponents. His running EPA is 29.0, which represents much more than the yardage ratio of three-to-one.

But Griffin has two vastly different types of run plays. There are designed run plays and there are pass play scrambles. According to the leagueâ€™s play-by-play data, Griffin has run by design 38 times, totaling 13.8 EPA, and he has scrambled 20 times, for 15.2 EPA. Thatâ€™s 0.4 EPA per run and 0.8 EPA per scramble. (This basically means that for every two Griffin runs, he generates somewhere around a full point of net score advantage, turning what hypothetically might be a 20-20 score into a 21-20 score in favor of the Redskins.)

Griffinâ€™s scrambles tend to be more productive than his designed runs, but they also have another benefit which may be considerably more important in the long run. On his scrambles, Griffin made it out of bounds 17 of 20 times. On designed runs he got there just 8 of 38 times. Although getting out of bounds is certainly no assurance of evading a bone-crushing tackle, it does tend to spare a runner the kind of hit that puts a player on injured reserve. Plus, there is a reduced chance of fumbling and losing a fumble near the sideline.

So thatâ€™s my recommendation for Griffin: fewer designed runs replaced by pass plays, perhaps designed with a scramble in mind. Especially if the defense is in man-to-man coverage or if his receivers can clear out one side of the field, Griffin should have an automatic green light to tuck it and run down the sideline. Yes, a sack is a risk on such plays, but on balance itâ€™s worth the risk. Besides, every designed run almost guarantees a hard hit, while Griffin is sacked only about 7% of the time.

Griffinâ€™s scrambles are not only highly productive, they help keep him healthy.

Brian Burke is the creator of Advanced NFL Stats, a Web site about football, statistics and game theory.