Fantasy Football: Stop Drafting Dalvin Cook at His Current Cost
If you feel 100% certain about your stance on anything in fantasy football, it's time to reconsider.
Fantasy football is a game of people making their best guesses about a game that relies on coaches and front offices making their best guesses. There's no certainty involved -- and that's part of what makes it fun.
Recognizing that uncertainty is also one of the best ways to get an edge in the game. Human beings aren't wired for making rational decisions in the face of uncertainty, and there's a lot to be gained by understanding the faults in our decision making. Often there's a big edge in simply doing the opposite of what people seem to be blindly buying into.
As of writing, the fantasy football market is reflecting a ton of confidence in the fact that Dalvin Cook is the top dog in the Minnesota Vikings' backfield, and that Latavius Murray is little more than a minor annoyance.
Per MyFantasyLeague.com, Cook is being drafted at an average draft position (ADP) of 45.2 (that's the tail end of the fourth round in 12-team leagues), making him the 13th running back off the board on average. Murray is going at 98.9 (the start of the ninth round) and as the 27th running back off the board.
If we knew as a 100% fact that Cook would lead the Vikings in touches, that might be reasonable. But no matter how excited you are about Cook's potential in the NFL, there's no way you should be touching him at his current ADP.
Rookie Running Backs in Fantasy Football
Ezekiel Elliott had a monstrous 2016 season. That's fresh in our memories, so it's easy to get really excited about rookie running backs. But letting last year's big story skew our perception of a scenario this year is falling victim to recency bias (check out Dr. Renee Miller's excellent work on cognitive bias in fantasy sports for more in that area). We can't let one story drive our decision making, so let's dig into the numbers and take a closer look at how rookie running backs have fared recently relative to their draft positions.
Over the past five seasons, 49 rookie running backs have an ADP inside the top 250, per MyFantasyLeague.com. (I'm cutting the sample down to 47, removing Ryan Williams and Mikel Leshoure, both of whom suffered preseason injuries and didn't play as rookies.) These remaining 47 backs have had ADP's ranging from as high as the 3rd running back off the board (Zeke) to as low as 60th ( Cameron Artis-Payne and Joseph Randle). They've also posted fantasy finishes all over the board, all the way from RB2 to RB133. If you're looking for one specific example to back your claim up in a sample like this -- you're going to find it, no matter what stance you're taking. But taking this group as a collective, some trends emerge.
The rookies in the sample have gone, on average, as the 41st running back off the board, while posting an average finish of RB49. For context, the ADP gap between RB41 and RB49 last year was the difference between a late 9th-round pick and an early 12th. So, in general, rookie running backs have failed to live up to the hype, and are being overdrafted by about two full rounds. If we narrow things down a bit and ignore the late-round dart throws, backs you take knowing there's a chance you drop them two weeks into the season, things don't change too much either.
Looking only at rookies who were among the top 25 running backs in ADP, we still get a ranking gap of about 7 places, with an average ADP of RB19 and fantasy finish of RB26. Better but still poor value.
Being a high pick in the NFL draft is obviously an encouraging sign, though -- it indicates at the very least that the team has some faith in the player. Backs taken in the first and second rounds of the NFL draft have performed better than the total sample -- with an average finish of RB40 instead of RB49. In a vacuum, this is a good sign for Cook, but when you factor in the behavior of fantasy players, that outlook changes.
That group has an average ADP of RB28 -- so despite that improved production, they are actually even worse values. People get too excited about these big names, which has led to rookie running backs being an exceedingly poor fantasy investment over the past five seasons.
Their teammates, however, didn't have those problems.
Again keeping things in the top 250 of ADP, as we did with the rookie sample, we get 20 veterans whose teams drafted running backs in the top two rounds of the NFL draft.
This sample of backs had an average ADP of RB32, and a an average finish of RB33. If you blindly targeted veteran backs whose teams invested a high draft pick on a rookie, you would have had significantly better results than targeting those rookies.
"But This Time Is different!"
Of course, those are just the overarching trends. There are exceptions on both sides -- rookie backs who outperform their ADP significantly as well as rookie backs who bust spectacularly.
In general, betting against a trend without a very strong case for an exception is going to be a losing bet. But Cook's ADP (and my Twitter mentions when I suggested Murray as the back to own) show that people believe Cook is one of those exceptions.
The belief seems to be that "talent" will win out, and that Cook is already definitely a better NFL running back than Murray. I have a huge issue with using a vague, nebulous "talent" designation in fantasy football, but let's dig a little bit deeper into this.
Here at numberFire, we have a metric called Net Expected Points (NEP), which essentially measures the expected points a player generates on a given play (you can read more about NEP in our glossary). Murray posted 0.04 Rushing NEP per carry in 2016, which ranked 21st among the 69 backs who recorded at least 50 carries on the season.
It's not often that a team selects a rookie in the top two rounds to pair with a starter that posted a positive Rushing NEP per carry in the previous season. (Negative Rushing NEP per play marks aren't uncommon because of how inefficient running the ball is.) We're bordering more on "interesting anecdote" than "trend" here with a four-player sample, but the veterans had an average fantasy finish one spot better than their ADP (and three of the four outperformed their draft spots), while only one of those rookies finished inside even the top 100 fantasy backs, with none cracking the top 50.
Again, we can't read too much into a tiny sample like this, but rookies have not had an easy time outperforming productive veterans. In fact, no rookie running back that finished RB13 (where Cook is being drafted) or better was competing for touches with a back who had posted a Rushing NEP per carry better than -0.04 in the previous season.
While I'm sure NEP doesn't completely encompass your personal definition of "talent," it does help capture "productive NFL running back" -- which is a lot more relevant to fantasy points than "talent" if the talent doesn't come with the productivity and/or volume.
Additionally, Murray has been a very strong fantasy producer. In 2015, he was drafted as the RB18 and outperformed that with a finish of RB10. This bumped his ADP to RB15 last season, and he again outperformed that with an RB13 finish.
Again, it may not be "talent," but a proven ability to be a top fantasy producer certainly can't hurt his odds of producing well in the future.
Opportunity is what gets you fantasy points. Efficiency is fun to chase, but your focus should be on whether a player is getting the ball, not what they do when they do when it happens.
Per a study done by 4for4's Chris Raybon on what stats correlate with fantasy production, it was bulk stats (yards per game and touchdowns per game) that had the strongest correlation with fantasy scoring -- not surprising considering those are the stats that provide fantasy points. But stats that reflect opportunity had a significantly higher correlation than efficiency stats -- with touches per game having a 0.71 correlation (in half-point PPR scoring), while yards per touch (0.20) was one of the least correlated stats he looked at.
Additionally, as 4for4's TJ Hernandez outlined, chasing running back efficiency is a fool's errand. He found that running back's yards per carry have almost no year-over-year correlation (r = 0.11), while fantasy points per game were one of the more highly-correlated stats (r = 0.58). As he concludes: "A running back's previous-year efficiency is a poor predictor of his future efficiency -- don't put much stock into yards per touch and yards per carry."
So who is going to get the ever-valuable touches in Minnesota? That's where we get into a guessing game.
Personally, I believe it's Murray. Per Spotrac, his contract gives him $8.55 million in practical guarantees -- the 11th-most among running backs (it was the 10th-most at the time of signing). That's not exactly a contract that screams "backup."
Make the Best Guess
The beauty of this situation, though, is that while you can make an educated guess on the breakdown in the backfield, you don't have to be able to accurately project the share of touches to find value because of how far apart their ADPs are. Even if you think that it's likely Cook gets more touches than Murray -- there is still huge uncertainty. Uncertainty that, historically, has not been properly priced into draft position, benefiting players like Murray.
Just about everything has to break perfectly for Cook to be worth drafting at his ADP. Fantasy players' analysis that he is a special talent (something that even NFL teams regularly fail horribly at identifying) needs to be true. He then needs that talent to earn him plenty of opportunity, and he needs to become a total outlier in terms of thriving in his situation.
Murray's the unsexy pick. He's probably not a guy whose highlight tapes you've spent the offseason watching, and he's not someone being talked about in a positive light very often. But history is on his side here, and at a steep discount, he's the Vikings' back to own this season.