Draft Strategy

3 Structural-Based TE Draft Strategies for MFL10s

Last year I wrote my Two-TE1 thesis for best-ball leagues, which had four main points to it.

  1. TEs between Rounds 4 and 9 (TE2 to TE12) were projected for similar points to their WR counterparts.
  2. This freed up the ability to draft 3-DEF, or 3-QB or an extra WR/RB while locking down the TE position.
  3. It was robust to different assumptions, as shown by two different sets of Monte Carlo simulations.
  4. It worked with a multitude of structural-based drafting strategies, including Zero RB, robust RB, and more.

So how am I approaching the TE position this year? Slightly different, but really, not super different from last year. Each of the three strategies hinges around a structural-based draft strategy. To see why structural-based drafting is important for 95 percent of us, I suggest you read the intro to Josh Hermsmeyer’s structural drafting in dynasty article. Let’s start with a look at some projections, then I’ll present the strategies that arise from it.

Projections

Like last year, I’ll start with the projections from our RotoViz Staff composite projections for 2017.

Staff Projections 2017 Phase 1

In this graph, what’s interesting is the orange curve and the green curve overlap from ADP 20-ish until about ADP 50-ish. That’s right where the first through fourth TEs are being taken — Rob Gronkowski (ADP 21), Travis Kelce (ADP 32), Jordan Reed (ADP 39), Greg Olsen (ADP 54). What that suggests is you can forgo a WR in this range, and pick up one of these TEs to easily fill the singleton TE slot with a player that is projected to perform just as well as the average WR taken at similar ADP. The other thing of note: New England, Kansas City, Washington, and Carolina each have different bye weeks, so if you take two of these TEs, you’ll have the bye weeks covered.

Now, to ensure this isn’t just RotoViz bias, I went ahead and grabbed PPR projections from Fantasy Football Analytics, which takes projections from other sites like ESPN, CBS, Yahoo!, Fantasy Sharks, and FF Today, and produced a weighted average based off each site’s historic accuracy.

FFA.net Proj 2017

Big picture — these projections don’t look too different from our projections, validating our staff projections. A subtle difference here is that the TE position stays closer to the WR position through about pick 120, which is right around where Jack Doyle is going. When I average our projections with Fantasy Football Analytics, giving us an average weight of the six components, the drop-off starts at Eric Ebron, who happens to be the TE12 by ADP.

The other thing you should notice in both graphs is that the TE position drops away from the WR curve and toward the RB curve pretty steeply, ultimately becoming just about level with the RB curve in both projections slightly after pick 150. We probably want to stay away from TEs in this range, which goes from Austin Hooper at TE17 to Dwayne Allen at TE 24.1

So how does this all add up into drafting strategies at the TE position?

My Three Structural-Based TE Draft Strategies for MFL10s

The Two-TE1 Strategy

My preferred strategy is to stick with the basic Two-TE1 strategy from last year, given the composite projections show that the average TE and the average WR in the ADP 20 to 120 range are similar. I’d lean toward taking at least one TE on the earlier side so you can feel comfortable about sticking with just two TEs. Mike Beers notes in his TE decision tree that if you take a top-six TE you can almost certainly stick with two TEs.

TE Decision Tree

I’m going to amend this to a top-four TE. However, I also think you can stick with two TEs if you grab two top-12 TEs that have different bye weeks. I’d prefer one top-four TE if you approach this, especially given the RotoViz Staff projections show a bit bigger drop-off at TE after the first four TEs than the Fantasy Football Analytics projections do, but I can understand just grabbing two top-12 TEs. In other words, the Two-TE1 strategy is intact this year, and this year it also includes Rob Gronkowski at his lower ADP relative to last year.

The Beefed Up Two-TE1 Strategy with Zero RB

This is simply the Two-TE1 strategy beefed up to pick two of the top four TEs. The thing that’s good about approaching a draft with this strategy in mind is that you can always shift to the basic Two-TE1 strategy, or the third strategy which I’ll talk about in a minute (although I prefer not taking a top-four TE with the other TE strategy…you’ll see why).

The caveat here is you’ll only want to use this strategy in a Zero RB setting, rather than with any strategy like robust RB or WRx5. That’s because you’re picking two TEs so early that you’re eliminating any structural-based strategy other than Zero RB and instead will essentially just be picking the best player (with roster construction considerations in mind) for the rest of the draft if you don’t use Zero RB.

The Three-TE Strategy

The Three-TE strategy I’m recommending is somewhat specific. In general terms, you take a top-12 TE and two later TEs, but I think there’s more to it than that.

Looking at Mike Beers’ decision tree, he recommends three TEs if you pick a TE7 or worse as your first TE, but preferably a top-10 TE. I think that’s a little specific and narrow to go from TE7 to TE10 as your first TE and then take three, so I’ll use the projections to kind of tier it out a bit differently. These are the ground rules I’m laying out for my structural-based approach to TEs if you want to go 3-TE:

  1. For your first TE, pick a top-12 TE.
  2. Preferably make your first TE in the TE5-12 range rather than the TE1-4 range.
  3. Pick two more later TEs.
  4. I prefer picking one in the TE13-16 range and one in the TE25+ range given how the TE17-24 range drops down toward the RB curve in both graphs.

This allows a couple things. First, you can safely roll out any structural-based strategy you want. Robust RB, WRx5, Zero RB are all in play. Second, it’s only a slight modification to Mike Beers’ decision tree, where we’re picking our first TE between TE5 and TE12 instead of TE7 to TE10. Third, you can grab an advantage at the non-onesie positions. Getting multiple advantages at a position is helpful, because you can only get one advantage at the TE position (and a second if the TE is an advantage in the flex…more in that in a future article).

If you go Zero RB, you can essentially turn it into a WRx4 or WRx5 start, and grab two of your RBs in that ADP 150ish range where the TE curve dips down toward the RB curve. Joe Williams and Marlon Mack are two names I’m particularly fond of in that range, but there are a lot of other nice RB options there as well.

Conclusion

Structural drafting is one way to gain a long-term edge in MFL10s (and in most fantasy football formats) even if you aren’t a top-tier player evaluator. These three structured TE draft strategies can be used in conjunction with other structural-based strategies, like Zero RB, WRx5, or robust RB to put yourself in position to gain an edge in MFL10s this year.

Adopting a portfolio of the three different TE approaches is a somewhat “game theory optimal” (GTO) approach, because each strategy has a certain probability of being the winning one. If you want to be exploitative, and think a particular one of these approaches is best (maybe based off how you foresee the other positions playing out — for example if RBs are systematically being overvalued, you could choose the beefed up Two-TE1 approach) you could choose to go all-in or nearly all-in on one of these strategies. It’s a higher risk-higher reward approach, so it comes down to your risk tolerance and your evaluation of where the market lies at each position.

Regardless, a structural approach is one that helps you avoid “pick the best players,” because as much as we all like to think we’re great at player evaluation, really we can’t all be top-five percent at it.

Subscribe for a constant stream of league-beating articles available only with a Premium Pass.

  1. Jason Witten is a notable exception in both projections at ADP TE18, and is one of the best overall values according to our staff projections.  (back)
By RotoDoc | @RotoDoc | Archive

Comments   Add comment

  1. I think the flatness portions of the argument are equally applicable to non-bestball leagues. I took Eifert and Ertz with my 7th and 8th picks in a recent draft and felt pretty great about it all things considered.

Discuss this article on the RotoViz Forums