Two games into the season, the run game has looked fairly disappointing so far. I understand that the offensive line and the blocking schemes are an important component in the success of the run game, so when we talk about whether a player is performing well, you must always keep in mind that they are not the only player on the field and there are other circumstances that can lead to the statistics we're left to interpret.
But with that disclaimer out of the way, I do want to take a look at the numbers from a different perspective than perhaps we've seen before. This is all still sort of in development and could possibly evolve into something more complex later, but for now, I think I'm happy enough with what I see to share the fruits of my labor.
Ray Rice and Bernard Pierce appear statistically to be about equal this year. The stat line we'd typically see shows us this:
Rice: 72 yards on 25 carries, 1 touchdown, 2.88 yards per attempt
Pierce: 79 yards on 28 carries, 1 touchdown, 2.82 yards per attempt
That's essentially a dead heat.
From my alternative way of looking at the statistics, we will see quite a divide indeed. First and foremost, you don't always see the fumbles statistic listed on the line, and in that regard Rice (with 1 fumble) is hurt compared to Pierce (with 0).
The primary ways of looking at a running back's performance are typically total yards and yards per attempt. I'm going to flip that on its head a little bit: I will give 1/3 weight to yards per attempt, and 2/3 weight to the value of each individual play from a positive/negative outcome point of view. What this means is that I'm going to score each play (based off statistics, not subjectivity) in order to come up with a value that increases based upon first downs, touchdowns, and yards gained, and decreases based on fumbles, yards lost, and stuffed runs.
Let me give some quick anecdotal examples of how this applies before I continue to the numbers:
If a running back is always brought in for 1-yard to go type situations and *always* converts, he has done a perfect job, even if he only has a 1 yard per carry average and 50 total yards on the year. 50 yards on 50 carries looks terrible... but if you're a 1-yard specialist and you have a 100% success rate at achieving the first down, you've actually added tremendous value to your team, so looking at yards and yards per attempt aren't helpful in analyzing this player's production. (This would get you somewhere in the 8.50 range, which you'll know more about what that means by the end of this post).
From another angle, a player who gets tackled for a loss on first down every drive, but manages to bust a 50 yard run every game will probably even out and look like they have a decent yards per attempt. But is a single play that gets your team into field goal range going to make up for all the 3 and outs you help set up by getting stuffed every other drive?
OK, as a second disclaimer, I'm not going to reveal all of the numbers that are going into this right now, or the formula behind it. But I will reiterate that this weighs two parts: 1/3xYPA and 2/3xEfficiency.
The Efficiency is my sum of points evaluated based on fumbles, loss of yardage, no gains, below average gains, average gains, above average gains, first downs, and touchdowns.
So, through these first 2 games of 2013 so far, we have the following production:
Loss of yardage: 3
No gain: 6
Below average gain: 5
Average gain: 3
Above average gain: 3
First down: 3
Loss of yardage: 3
No gain: 4
Below average gain: 8
Average gain: 3
Above average gain: 6
First down: 3
And just so we can see how the team is looking overall:
Ravens Team Total
Loss of yardage: 6
No gain: 10
Below average gain: 13
Average gain: 6
Above average gain: 10
First down: 6
That last one is my final evaluated score... I'm calling it Value, but if you want to think of it as a rushing-version of Adjusted Yards Per Attempt, that's good too. It's more complex than the passing version of AYPA, but it's a similar concept all the same.
The most I'll say about it right now is that anything above 8.33 is nearly perfection, the kind of performance that means you can put the team on your back and produce first downs and touchdowns on virtually every drive. In reality, the possible scores range from -59.5 to 59.5.
Anything in the 3 or lower range is not good and indicates a serious lack of contribution and trouble helping to sustain drives.
Somewhere in between 3 and 7 is where we'll find the vast majority of players.
Right now, league average rushing Value is 5.05 (includes QBs, but with kneeldown plays removed).
As you can see from the results above, Bernard Pierce is helping this team a lot more than Ray Rice, despite their traditional stat lines looking about the same.
Overall though, our team rushing Value is significantly below league average right now.
Currently, the league leader is LeSean McCoy with a Value of 7.82. This total is unlikely to stand... For comparison's sake, last year's leader was Adrian Peterson with a 6.04. That gives us all a little confidence I think that these small sample sizes in a 2-game stretch to start the season probably aren't going to be indicative of what we'll see throughout the course of an entire season.
Edited by callahan09, 18 September 2013 - 01:29 PM.