≡ Menu

On Friday, I asked the question: how many carries would we need to take away from DeMarco Murray in order to drop his YPC average to at or below league average?

Today, I want to look at it from the other side. How many of Trent Richardson’s worst carries would we need to erase to bring his YPC above league average? For this experiment, assume that we are sorting each running back’s carries in ascending order by yards gained. I’ll give you a moment to think about the answer.

[Final Jeopardy Music]

[Keep thinking…]

[Are you ready?]

[Your time is now up. Post your answer in the comments!]

For Trent Richardson, the answer is 27, which represents nearly 17.0% of his carries last year. It’s a sign of how far the former third overall pick has fallen that we need to remove so many carries just to bring him to a hair above league average. But, believe it or not, two other running backs fare even worse in this experiment. Houston’s Alfred Blue had a worse YPC average than Richardson and a few more carries, so Blue actually needs to remove 30 of his carries to bring his YPC above 4.16, the average for all runs last year.

But the most extreme case belongs to New York Giants rookie Andre Williams. I think his failings may have slipped under the radar for some fans, but he was just barely better than the historically bad performance Ray Rice produced in 2013. But here’s the good news, Giants fans: if you remove his worst 34 carries, he was above average!

The table below shows each running back in 2014 who recorded at least 100 carries and averaged fewer yards per rush than league average:

Running BackRshRshYdYPCNum
Andre Williams2167173.3234
Alfred Blue1685293.1530
Trent Richardson1605203.2527
Andre Ellington2016633.326
Matt Asiata1645703.4824
Darren McFadden1565343.4221
Ben Tate1193703.1121
Toby Gerhart1013263.2318
Steven Jackson1897083.7513
Branden Oliver1605823.6413
Knile Davis1344633.4613
Bishop Sankey1535723.7411
Fred Jackson1415253.7211
Matt Forte26610383.910
Joique Bell2238663.8810
Doug Martin1354943.6610
Rashad Jennings1676393.837
Terrance West1716733.946
Alfred Morris26510744.053
Giovani Bernard1686804.053
Ronnie Hillman1064344.092
Chris Ivory1988204.141
Isaiah Crowell1486074.11
Anthony Dixon1054324.111

Leave your thoughts in the comments — what would you do with this data? Should we extend this idea to other positions?

  • James

    I wasn’t sure how many, all I knew is that it was going to be a lot more than for the above average YPC backs!

    I definitely want to see this for other positions, if only to compare the various magnitudes involved.

  • dbqp

    Is it really such a bad thing to have a large number here? It could just mean you didn’t have a lot of negative-yard runs. A player with 10 -5-yard runs will need a lower amount of runs removed than a player with a lot of 1-yard runs. I might be overlooking something but it doesn’t seem like a valuable stat.

    • Richie

      Looking at the names at the top of the list – it looks like a large number here is bad.

      • dbqp

        Sure, but that might be because they all start off with below average YPC and a large amount of carries. Andre Williams actually has to have a lower percentage of his carries taken away than Blue or Richardson.

        My problem with this kind of digging is that a theoretical back with 400 carries of exactly 4 yards each and one five yard carry would need 400 carries removed to bring him above average. But honestly, wouldn’t that be an amazing running back? Consistently getting 4 yards even in short yardage situations is worth a ton.

        I see the point in the other article, because long runs can really skew averages, but the data isn’t symmetrical. It’s not like there’s a bunch of backs with -30-yard runs that are otherwise above average rushers.

        Usually with data like this, I would assume a log transformation would be the natural place to start instead of just adding or removing data points.

  • Richie

    As a fantasy football owner of his, I expected Toby Gerhart to top the list.

    Though as a percentage of total carries, it looks like Gerhart is one of the worst.

  • Dan

    What if, for each RB, you removed the best 5% of his carries and the worst 5% of his carries, and then you looked at his YPC for the remaining carries?

    That might be a better way of looking at low YPC backs, since low YPC is more related to a lack of big runs than the presence of a few very bad runs.

    • dbqp

      That’s the simple way to do it. It would be pretty good.

      I still think it’d be even better to calculate a confidence interval (the average plus minus one standard deviation for instance) by using logarithmic transformation of the raw data. For a running back the most interesting number would probably be the average minus one standard deviation, as that gives you a realistic impression of what you can expect on a given run, with anything above that being a bonus.

      • Dr__P

        yes the CI would be the best way to show if off

    • Dan

      Another way to use the data is to calculate “percent of carries going at least n yards” for various values of n. Then you could make a heatmap for each RB.

      For example, you can calculate “percent of carries going at least 2 yards” – if a RB is much better than average at that then that square of the heatmap gets colored bright green, if he’s bad at it then it is dark red, if he’s in between then it gets colored somewhere along the scale from dark red to bright green. Then repeat for “at least 3 yards”, “at least -1 yards”, “at least 12 yards”, and every other length. Then line up the colored boxes left to right, shortest distance to longest (but probably not equally sized – you want more of the visual space to be taken up by the most common run lengths). A great all-around RB will just be a green line, a bad RB will be all red, a boom-or-bust RB will be red on the left (e.g., fewer than avg carries going at least -1 yards) and green on the right (e.g., more than avg carries going at least 12 yards), and an effective grinder will be green on the left & red on the right (lots of 4-5 yard carries, few 20+ yard carries).

  • Pingback: Running Back Heat Maps()

  • Clint

    Love this!
    A couple things…

    The Browns running backs were victims of very poor o-line play after Alex Mack went down. He was the glue. It just completely fell apart from the center-right tackle.

    Crowell, Tate and West are all better than this.

    Also, T-Rich has had a couple “minor” knee surgeries that the Browns had totally threw under the radar. Pretended they weren’t a big deal. Before ’12 and one before ’13. He has no burst. Doesn’t know how to make a cut and read the o-line anyways. In the end, he looks like a great athlete who has never played running back before.

  • Pingback: Fantasy Football 2015: I Don’t Play Fantasy Football For Money, But Here Are 5 Tips If I Did | The Locker()