But how confident can we be in him?
We do know this about him:
However, we still only have less than a year’s worth of experience to go by from the young running back. Many are starting to compare him to Adrian Peterson already. It’s too early in his career to make that comparison definitively, but there is one question we should be able to answer by now:
Is he better than an average NFL running back?
Since he’s being drafted as the #1 overall running back in fantasy football, the overwhelming answer from most people is likely “YES!” But if we look at the data objectively, you might walk away surprised.
Let’s try to answer this question by looking at every single carry made by Gurley last year.
Gurley had an average YPC of 4.83 last year on 229 carries. However, average yards per carry hides a lot of information. Gurley had several long runs that helped balance out many shorter runs.
To visualize his full body of work, see below for a histogram of all of his carries in 2015, along with a marker that shows the NFL average of 4.13 yards per carry.
Data from Pro Football Reference.
Looking at it that way makes the answer not so clear.
For a summary, here is the same data grouped in larger buckets:
|Range||Percent of Carries|
|0 Yards or Less||19.7%|
|1 - 5 Yards||58.1%|
Looking at the distribution of Gurley’s runs, 77.7% of his runs were 5 yards or less. Since his overall average was 4.83, that means the vast majority of his runs were below his own average.
Now, it’s normal to have a lot of long runs as a running back. It’s clearly not a bad thing.
And we’re not going to go down the path of “What if you remove this long run? Look what it does to his average!” That’s not what we’re analyzing here.
All we want to find out is this: Based on his actual distribution of runs, how likely is it that he is better than an average NFL running back?
We’ll use a traditional statistical test to look at whether or not Gurley is better than the average running back. The traditional metric used and taught everywhere is a t-test.
The null hypothesis is that Todd Gurley is no different than an average running back. If we were a statistician, we would state up front that we want to have something like 95% confidence in our result.
When we perform the one-sided t-test, we would expect the results to calculate at least a t-statistic of about 1.66 (PDF).
When we run all of Todd Gurley’s carries through a t-test calculator, we get these results:
|Hypothesized mean (h):||4.13|
|Sample mean (x):||4.83|
|Sample standard deviation:||9.280|
|One-tailed probability P(h > x):||0.873|
His t-statistic was only 1.141, which is less than 1.66. That means we do not have 95% certainty that Todd Gurley is better than an average NFL running back!
The result of that calculation is that there is an 87.3% probability that Todd Gurley’s carries came from a running back better than the average NFL running back. But there is a 12.7% chance they came from an average NFL running back.
That’s still solid evidence that shows that Gurley is pretty good. But it is surprising that it fell short of a more traditional 95% or even 90% certainty level.
The result is that there is an 87.3% chance that Todd Gurley is better than an average NFL running back. It’s an indication that he’s a good running back, but a statistician would claim that there is no concrete evidence (yet) that Todd Gurley is anything but an average NFL running back.
Now you at least have the full data to help you decide for or against drafting Todd Gurley in the first round of your drafts this year. Keep this in mind along side all of the other information you have such as game tape, news reports, and Hard Knocks episodes.
Never Reach For a Player Again!
Join the 18,000 others who stay up to date!
Get FREE articles now!
Did you enjoy this article?