For the first 10 weeks of this 14th season of Fantasy Football for Winners, everything I wrote tied back to theoretical draft and roster management decisions: seemingly undervalued and overvalued players, how injuries might impact multiple backups, and a lot of data analytics covering historical probabilities of player production.
Yesterday I walked through how I put these theories into practice. When I share one mock draft each summer, I know I'm setting myself up for pushback. There should always be pushback. We don't all draft using the same strategies. We don't all agree on player values. I fire an arrow at what I *believe* is close to the bullseye, and some of you are thinking, "um, you fired an arrow into Old Man John's farm. That's not a bullseye. It's a corn stalk."
Since this page's primary purpose has always been to help people win their leagues, if what I'm writing isn't helping, then my thinking and my execution need to shift. So these comments push me to re-evaluate my thinking and my execution. Am I making sound decisions? And can I back them up with evidence?
For example, in yesterday's mock, I drafted TreVeyon Henderson in the fifth round and then made sure to get teammate Rhamondre Stevenson in the sixth. Some of you justifiably believe Stevenson would be a wasted pick in that scenario: if we have one co-starting running back, why take the other? It's a fair point, and there's no objective right answer.
For context, here are the players my nine opponents and two auto-drafters took after I got Stevenson and before it reached my next turn in the seventh. These players are in order: Dak Prescott, Kyle Pitts, Parker Washington, Christian Watson, Jonathon Brooks, George Kittle, Tony Pollard, and Jaylen Warren. There are many smart opinions about these eight guys, some of whom could be fantastic. There are also (in my opinion) huge risks to taking some of them, while others have realistic ceilings that would make them streamable at best.
Something that people don't talk about enough are net benefits and net losses. Suppose you're entering your championship week. You and your opponent each just lost your best starting RB. You're both thin at running back. The only starting RB on waivers is your elite running back's handcuff. If that handcuff is projected to score 12+ points and has great upside, do you pick him up and start him? Or do you let your opponent take him and start him against you?
Anytime this scenario comes up during the season--where a capable handcuff replaces a starter--managers are clamoring to get him. Championship week? Of course he's invaluable. Taking and starting him yields a net benefit for you and a net loss for your opponent. It's a win-win. Letting your opponent take and start him (while you scrounge around for a "lesser" running back) is a net benefit for your opponent and a net loss for you.
In yesterday's mock, I could have taken Dak, even though I knew I could get an historically higher-upside QB (Patrick Mahomes) a couple rounds later. A good case could be made for Pitts or Kittle, though as discussed on a podcast episode I did about drafting TEs, the gap between Pitts/Kittle and my favorite bargain TE (T.J. Hockenson, who I took about 60 picks later) probably won't be that big. Since 2012, preseason TEs 5-9 have averaged 136.5 points, which is only half a point higher than TEs 10-14 and barely one point better than TEs 15-19. (I could do a whole column just on this, because it's shocking and actionable.)
How about Washington or Watson? Not a chance, because this is a common dead zone for wide receivers, and both of those guys are in a three-headed WR situation on their respective teams. Since 2012, the five WRs taken around this spot and going forward (WRs 30-34) have averaged 143.1 fantasy points. The next five (WRs 35-39) have averaged 162.0 points. The next five (WRs 40-44) have averaged 148.0. So unless you believe that Washington or Watson will buck expectations and be a weekly fantasy starter, statistically they're more likely to be worse than the WRs drafted a round or two later.
As for the running backs (Brooks, Pollard, and Warren), they're all potentially "fine." Like Stevenson, they're in relatively crowded backfields. Unlike Stevenson, Brooks has barely played his first two seasons and is in a low-tier offense, Pollard will turn 30 next offseason and is in a relatively low-tier offense. Warren realistically might end up being the #2 behind Rico Dowdle and plays in a less TD-friendly offense than New England.
Here's the risk if we don't take Stevenson: an opponent gets him. Now it becomes a zero-sum game if he or Henderson gets hurt. Since 2012, top-30 preseason RBs have missed an average of 3.2 games per season. There is a reasonable probability that Henderson and/or Stevenson will start four or five games this year with the other sidelined. If Henderson's hurt, do you want your opponent starting Stevenson against you, while you roll out (for example) Christian Watson instead of Michael Wilson (my pick the following round)? Or Dak instead of Mahomes (my pick two rounds later)?
My drafting approach is based partly on these net-benefit / net-loss scenarios, backed by statistical probabilities. I wouldn't recommend this approach for, say, Brooks and Chuba Hubbard, or for any other backfield tandem that hasn't demonstrated they can consistently produce on their own *and* produce alongside their co-#1. The Patriots were #6 in carries last season. After Henderson had established himself, he and Stevenson both had double-digit points in three of the final five games they played together. So yes, they can dominate separately and also be streamable or better while playing together.
People in this community frequently have shared that they struggle with middle-round picks. That's pretty normal, and the statistical evidence backs this up: too often we settle for fringe starters in the middle rounds. We feel resigned to take the best available starting RB, even when that RB is unlikely to crack the top 25. We follow the flow of adding a top 30-35 WR, even though he's unlikely (based on historical data) to average double-digit points. Or we lunge for the TE6, when historically players ranked around there are barely better than TEs you can get five, six, seven rounds later.
So when drafting, consider not only how a particular pick benefits you, but also how ignoring a player might benefit an opponent. Injuries are part of the game, and sometimes making our roster more "injury-proof" will yield more wins than grabbing someone who might average 10.5 points per game, when an 8.5-point guy with comparable or even higher upside is easy to find with your final pick.
My job isn't to convince you, obviously. Your job is to decide whether what you're doing already works beautifully. If it doesn't, then there's always time to figure out how to hit the bullseye *and* apologize to Old Man John for attacking his corn.
---