StrategyJune 16, 2009

Post to Twitter

Breaking Down the Top 10 Running Backs (Non-PPR)

By Matt Grant

With OTAs underway and training camp just around the corner, it’s about time you start thinking about your yearly redraft rankings — if you haven’t already, shame on you. Lucky for you, I have started the information gathering process while attempting to spot some values and dangers in the top 10 (based on ADP) of this year’s running back class. This will be part one of a three-part series — the first two parts will set up the conditions we will use to analyze this year’s potential studs and duds, and those players will appear in part three. Let’s get started.


ADP is obviously a great tool in predicting the success of the upcoming year’s running backs. If you don’t use it, then shame on you again. To illustrate this, let’s look at what would happen if we just drafted the top 10 of the previous year and compare that to the top 10 that ADP produces.

This data set is from 2000-2008, using the top 10 from a previous year to determine who to take in the top 10 in the current year, and finding the players that repeat in the top 10 in consecutive years. So we are looking at who repeated in 2008 from 2007 (10 opportunities), 2007 from 2006 (10 opportunities), an so on, down to 2001 from 2000. Out of these 80 opportunities, there are 30 repeats. We would only find ourselves in the top 10 37.5% of the time if we draft solely on who finished in the top 10 the previous year.

Obviously, this isn’t a good strategy.How much better is it when we consult the collective knowledge of the fantasy football community and use ADP? Again, we use the same method but instead use the ADP of the current year’s running backs and find that out of the top 10 backs in each year, we successfully predict 39 out of 80 for a 48.75% success rate. If we break it down further and look at players with a top five ADP, those players have a 60% chance of appearing in the top 10, while players with an ADP ranking sixth through tenth only had a chance of 37.5%. Note how much more successful a top five ranking is compared to the next five.

This is quite a big difference considering how weak the accuracy of the prediction is to begin with. There is a lot of noise in the system of fantasy football, a lot of things that are unpredictable and dependent on each other. But there is still value in the information produced by running backs from the previous year — perhaps ADP doesn’t correctly reflect the weight of these factors. Now we will look at how 400+ touches and averaging under 4.0 yards per carry may help us in being more efficient with our predictions.

Let’s start by taking a look at how running backs that had over 400 touches (playoffs included) with an ADP in the top 10 fare the following season. We do this by looking at who is drafted in the top 10 of the running back position for 2001, finding out who touched the rock 400+ times from the previous season, then see who actually lives up to this billing in the upcoming season. This procedure is repeated up to 2008 and we will have an idea of how successful this group of backs is.

Doing this produces 23 occurrences where a running back gets 400+ touches the previous season and has an ADP in the top 10 the next season. 10 of these occurrences were successful predictions (43.5%). In 17 of these occurrences the running back is ranked in the top five, with the running back ranked sixth through tenth in the other six occurences. For the 17 top-five occurrences, nine of them were successful top 10 predictions for the upcoming season (52.9%). For positions 6 through 10, only one of the six predictions produced a top 10 running back the following season (16.7%).

There are a few points to make about these findings. Overall, 400+ touch running backs are only slightly less successful as a predictor of future success than the ADP top 10 as a whole (48.75% for ADP top 10 and 43.5% for 400+ touch running backs in the ADP top 10). The same is true for the top five, where the effect is a little larger and where the large majority of this sample is based (60.0% down to 52.9%). In positions 6-10, we find that the success rate is very low, but this sample is rather small (37.5% to 16.7%).

An interesting fact to consider in this sample is that LaDainian Tomlinson accounted for five of the nine successful predictions. Tomlinson is truly unique in this case, having 400+ touch seasons in three consecutive years, and later in back-to-back years, all the while remaining in the top 10. His freakish ability to stay healthy and recover may make this sample seem more successful than it really is. If Tomlinson is removed from the sample, then there are 18 samples, five predicting a top 10 player (27.7%), 12 of which are top-five players and four being accurate (33.3%). This is obviously a much more bleak outlook for 400+ touch rushers. Removing outliers like Tomlinson in this case can damage the integrity of the remaining sample and its results, but because Tomlinson has such a significant effect on the sample (five out of the 23 occurrences, all of them successful), the estimate is probably more accurate without him in it.

Okay, now what about players with under 4.0 yards per carry? I think most fantasy football players would assume that a player with under 4.0 yards per carry would be less likely to find himself in the top 10 than a player with a more respectable average. Let’s investigate this using the same methodology as we did with running backs with over 400 touches.

This produces 15 occurrences of a running back having a yards per carry average under 4.0 and being in the next year’s ADP top 10. In only two of these occurrences was the player drafted in the top five, with the rest being selected sixth through tenth. Across the sample as a whole, seven out of 15 of these occurrences appeared in the top 10 of that year’s ADP ranking (46.67%). This is very close to the accuracy of the top 10 as a whole (48.75%).

There is a very interesting element to this finding however. As I stated before, 13 of these 15 players were drafted in the bottom half of the ADP top 10. Six of those 13 running backs made it into next year’s top 10 (46.2%). Compare this to the success rate of the 6-10 ADP players we noted earlier of 37.5%, and we have a very interesting contradiction. In terms of ADP, players with under 4.0 yards per carry that appear in the 6 to 10 range, may actually be undervalued in comparison to the group as a whole.

The next part of this article will look at a few more variables of top 10 running backs, including yards per carry from 4.0-4.5, 4.5-5.0, and 5.0+. If you have other factors that you believe are worth exploring, leave me a comment at the link provided below to the Cafe forums.

Matt is a graduate student at the University of Michigan (GO BLUE). He spends entirely too much time researching fantasy football and his wife reminds him of this on a regular basis. You can catch up with Matt in the Cafe's forums where he posts under the name of mattUTD20.
Rate this article: DreadfulNot goodFairGoodVery good (15 votes, average: 2.93 out of 5)
Loading ... Loading ...

Questions or comments for Matt? Post them in the Cafe Forums!

Want to write for the Cafe? Check out the Cafe's Pencil & Paper section!

Post to Twitter

Related Cafe Articles

• Other articles by Matt Grant

No related articles.