In the months leading up to the fantasy football season we spend a great deal of time discussing value. It makes sense; we all want to get the biggest bang for our buck and spend as little as possible.
Whether we’re dropping a couple of dollars on a sandwich or buying a new car we want to get the best deal that we can. The less we spend, the more we have to contribute to future purchases. Value is good. Getting more for less is what we’re constantly trying to do as consumers.
When considering value from a fantasy football perspective we are essentially comparing a player’s ADP to our expectations of his likely outcomes in the coming season. In many ways, fantasy drafters are consumers. Prices are dictated by the market and are heavily influenced by supply and demand.
As such, it’s easy to let our general goals as consumers, including seeking out value, permeate into our tendencies as fantasy drafters. However, this can be a slippery slope leading us towards making less than optimal decisions.
What We Know About ADP
There’s no denying that, at least on some level, we need to consider value when assessing fantasy football players. We can use ADP data as a way to confirm our projections or identify players that we may need to re-evaluate. Perhaps we are overly bullish on a sophomore running back or too fearful of a team’s offensive situation.
It’s always a good idea to price check our prognostications against the market, as it will play a large role in the flow of drafts dictating when players will be available. These are all important considerations. Having said that, value alone is not a reason to draft a player.
Though ADP provides a nice structure in which we can identify market values of fantasy assets, it’s not the most efficient market. As Ryan Rouillard noted in Does Average Draft Position Imply an Efficient Market?, ADP is significantly influenced by a player’s performance in the prior season. Yet, the relationship between prior year performance and current year performance is not strong. As such, our observable market is inefficient at predicting a player’s performance in a given season.
This is one of the reasons I am hesitant to draft a player simply because he is perceived to be a good value. I like to view an individual player as a range of outcomes, encapsulating risk, upside, situation, opportunity, growth, regression, etc. However, ADP data is largely ‘projecting’ a player based on prior year performance and may not be considering these important factors.
The 2015 Season
I calculated the absolute difference between ADP and season ending rank for the first 36 RBs and 48 WRs selected in 2015 drafts. Additionally, I stratified each position into groups of 12 and counted the number of players in each bucket that posted an ADP vs. rank differential of 10 or less.
|RB ADP||AVG Diff||< 10|
|1 – 12||22.1||5|
|13 – 24||21.2||5|
|25 – 36||17.8||4|
|WR ADP||AVG Diff||< 10|
|1 – 12||13.4||8|
|13 – 24||15.3||7|
|25 – 36||37.4||3|
|37 – 48||46.5||4|
The results seem to confirm Rouillard’s research on the accuracy of ADP in the most recent NFL season.
Regardless of the market’s inefficiencies in projecting/valuing players, there’s an even more important factor that we need to consider. When purchasing goods in real life, we are able to carry forward our savings and use them to our benefit in the future.
However, in a fantasy draft, even if we get the ‘best deal ever’ we are left with no future benefit to carry forward. For this reason, the best ‘value’ during the draft is not necessarily the best option for your squad.
At the end of the day, we are hoping to assemble the best team possible and though finding value may assist us, it in no way guarantees an optimal fantasy football roster or provides us the utility that we may assume it does.
ADP vs. Expected Production
If we make decisions based upon value alone, we are ignoring roster construction, opportunity cost and failing to see the bigger picture. Plus, just because a player is under-valued, it doesn’t mean that he’s going to provide your team with more production than one that is over-valued.
For example, let’s consider John Brown and Ameer Abdullah in a 12-team PPR league. Brown is currently going as WR34 while Abdullah is the 31st RB selected. Both are being drafted somewhere around 77 overall in the seventh round of drafts. Brown has recently been identified as an over-valued receiver and Abdullah has been labeled as an under-valued running back. Yet, if we look at each player in the Sim Score app, we’ll notice that Brown has a more favorable projection in all cases.
Of course, there’s much more to building a winning fantasy team than just chasing points. Though Brown may provide your team with more, you need to consider a multitude of factors. For instance, the level of talent at each position that would be available in coming rounds, what you would be giving up by selecting either, the current construction of your roster at that point in the draft and an abundance of other variables before making your selection.
In addition — and this is a very significant caveat — we need to recognize our predictive fallibility. If we are drafting based on value alone, we need to be supremely confident in our ability to accurately project players. In the case of one such as Abdullah, I’m not sure that we have enough historical data or understanding of how the Detroit offense will function in 2016 to firmly label his price tag as over- or under-priced. As such, drafting him solely based on value is likely not the soundest strategy.
The takeaway here is that there are many things we need to consider during a process as fluid as the draft. Value is a useful consideration both during and in the weeks leading up to the draft, but it should only be an input used in a larger decision-making process. In isolation, value is not a good reason to go for or pass on a player.
No matter how great of a value a player may appear to be, the draft capital we spent to acquire him is a sunk cost. We can’t use any ‘savings’ to strengthen our team in the subsequent rounds. I’m not saying that you need to ignore value completely, but I do think that the emphasis placed on value during the pre-season needs to be significantly reduced when actually drafting players.