Tuesday night is the auction draft for a 14-team league I play in locally. This got me thinking. For as much as fantasy football pundits like to talk about process, why aren’t there more articles out there about the preparation process for a real-world draft? Sure, there’s a ton of content that falls under the heading “fantasy football draft strategy,” but process is more than just strategy.
Strategy is “Upside-down,” “Zero RB,” “Robust RB,” “2-2-1 RB,” “Late-round QB,” “stars and scrubs,” and so on. Process, on the other hand, is more about what projection systems we use, how we aggregate and weight them, how we adapt/tweak projections to fit our subjective beliefs, how we translate average draft position (ADP) to fit the exact structure of our league and draft, how we use our league’s draft history to inform us about ADP or auction cost, how we adapt to in-draft dynamics, etc. What I felt was missing from the online record was a real-world example of a so-called expert’s fantasy football drafting process; not his or her theoretical strategy. What follows is my attempt to fill said void.
League and Draft Parameters
As I mentioned at the start, this is a 14-team league with an auction draft. Here are a few more important league details:
- It’s redraft; no keepers.
- There are no divisions and only 6 teams make the playoffs. If that isn’t top-heavy enough, the payout structure implies that the maximum winnings equal 46 percent of the total prize pool.1
- In-season, there’s a daily auction from Tuesday to Saturday each week with a free agent acquisition budget (FAAB) of $100.
- Lineups must include one QB, two RBs, three WRs, one TE, one flex, one K , and one D/ST.
- The scoring system for the four offensive skill positions is .04 points per passing yard, 0.2 per rushing attempt,2 0.1 per rushing or receiving yard, 6 per rushing or receiving touchdown, 5 per passing touchdown, 2 per rushing or receiving two-point conversion, and -2 per interception.
And here are our draft constraints:
- Roster size is 18 players.
- Draft budget is $165.
- Once the above starting lineup requirements have been achieved, there are no positional constraints for populating the rest of the roster.
My Draft Process
So hopefully, the above details adequately cover the constraints of this draft. Now, onto my preparation process, which can be reduced to the following steps:
- Given that my Footballguys colleagues, David Dodds, Bob Henry, Maurile Tremblay, and Jason Wood rank among the most accurate fantasy football prognosticators, I grab their projections off of our site.
- Given that “the wisdom of crowds” is crucial to any forecasting system, I average these four sets of stat projections.
- Using the methods exemplified here and applied to fantasy football here, I calculate “true” fantasy-relevant stat rates for each player and multiply them by the average Footballguys opportunity projection (i.e., attempts, routes run, etc.) to produce my own “true” stat projections.
- I calculate the average of Footballguys’ “wisdom of crowds” projections and my own “true” projections.
- I apply the league’s scoring system to each player’s projected stat line, which gives me each player’s projected fantasy points.
- I calculate each player’s draft value via Joe Bryant’s (aka Footballguys grand poobah) principle of value-based drafting (VBD).
At this point, I know each player’s projected fantasy points and how much value said points represent. This is where I would stop in a snaking redraft. But this draft is an auction. As such, my preparation process requires several more steps aimed at pinpointing player cost:
- Using data from the past three years of our drafts, I calculate the average cost for players with a given ranking at a given position. Fo instance, the average WR1 costs $37, WR2 costs $36, WR5 costs $33, WR10 costs $28, WR20 costs $21, and so on.
- I apply these average costs to my VBD-based positional rankings, such that my WR1 is expected to cost $37, my WR2 is expected to cost $36, and so on.
- I grab Footballguys’ ADP data and apply the aforementioned average costs to those rankings, such that ADP’s WR1 is expected to cost $37, ADP’s WR2 is expected to cost $36, and so on.
- To identify each player’s auction value, I subtract his cost according to ADP from his cost according to my rankings. Negative numbers mean good value (e.g., Marcus Mariota is worth $12 per my rankings, but $2 per ADP), while positive numbers mean bad value (e.g., Jamaal Charles is worth $22 per my rankings, but $30 per ADP).
Upon completing the above 10 steps, I end up with my cheat sheet . For your edification, here it is:
But wait, there’s more. Step 11 involves my co-owner and I engaging in a subjective conversation about our “universe” of players. Basically, I send him my cheat sheet, and we then go through it one-by-one to include players we both want and exclude players we both don’t want. In the above link, our player universe is highlighted in green or yellow and the color distinction breaks down like this: Green = players we’re targeting; Yellow = players we’re OK with; Clear = players we won’t end up with under any circumstances.
And finally, Step 12 is adapting to the draft as it happens. When our green-highlighted players come up for bid, our overpayment threshold is 30%3, our yellow-highlighted players are 20%, and our no-highlight players are 0%.
DT : IR :: TL : DR
My process for Tuesday’s auction draft is as follows:
- Winning requires properly adapting one’s strategy to scoring and lineup parameters.
- Rely on the average of smarter people’s projections and my “true” projections.
- In-draft dynamics revolve around predicting everyone else’s general strategies and/or specific picks. Your league’s draft history provides huge clues.
- Combine objective stats with subjective opinions.
Granted, this only happens when the #1 team in terms of points scored also finishes as the #1 playoff seed and then goes on to win the league championship. But that’s what we’re shooting for, right? ↩
I emphasize this because it’s a scoring quirk we usually don’t come across and it has an outsized influence on strategic considerations ↩
Except Odell Beckham, who we’re willing to overpay for by more than 30% ↩