revolutionary tools.  groundbreaking articles.  proven results.

Computing an Optimal Draft with RotoViz Tiers

We talk a lot about high-level draft strategies like Zero RB here at RotoViz, but on draft day, you also need to adapt your strategy to the value available from particular players. For example, if you know about a likely WR2 who will be available in the ninth round, then it makes sense to grab one less early WR.

In this article, we’ll use the computer to demonstrate how you can take advantage of the value identified by the RotoViz tiered rankings in order to construct an optimal draft.

Optimal Draft Methodology

We will look at the first eleven rounds of the draft, when you need to grab your likely starters and key backups. Here, we’re looking at taking 1 QB, 4 RB, 5 WR, and 1 TE during those rounds.

To attach values to the players available, I used the tiers included in the FFDraftPrep tools. We have tiers for each position from six different RotoViz rock stars. For the calculations described below, I also needed estimated fantasy points for each tier. To build those, I took the average of projections for each of the players in the tier. As each estimate is an average from a half dozen or so players, these should be more stable and less error-prone than relying directly on projections. (I tried the experiments described below using projections from our own Dave Caban as well as Mike Clay of ESPN. The results were very similar, showing that the choice of projections used to compute average points of each tier are not especially important.)

With values in hand, all that remains is to try every possible draft order and see which gives the best projected points per the tiers above. There are only 13,860 distinct draft orders, which would be a lot for you or me to consider but is nothing for my trusty laptop.

I had my computer test each one of these draft plans. At each pick, it was only allowed to take players who are very likely to be available at that position per the ADP, which usually means taking players at least a few picks before where they are usually drafted. The result should be a plan that you would likely be able to stick to on draft day.

Overview of Results

Depending on your draft slot, there is a large difference between who is available in the early rounds of the draft, so I tested all twelve starting spots. The following figure shows the distribution of positions taken in each round by the optimal drafts from each starting position.

results

We can see a few trends right away:

  • Every optimal draft took a QB late — in Rounds 9 or 10.
  • Nearly all of optimal drafts (9 of 12) also went late at TE, though a few grabbed a high TE1 in one of the earlier rounds.
  • Every optimal draft took an RB in Round 2. Almost every one (10 of 12) took an RB in Round 1 also, but as we will discuss below, WRs should also be options in Round 1.

Below, we will look through some of the specific players that were used repeatedly across these optimal drafts. As we will see, a number of the rounds have clear, locked-in best picks, while others have a range of reasonable options, leaving us with room to adapt to our own favored draft strategy.

Marqise Lee, WR, Round 11

Every optimal draft took Marqise Lee, usually in the 11th round. He ends up 34th in these rankings at receiver, only a few spots lower than guys like Emmanuel Sanders and Corey Davis, who are taken much, much earlier. Of course, 34th is just a low-end WR3, but that’s incredible value for a bye week fill-in, if nothing else.

Late-Round QB, Round 10

As noted above, every optimal draft took a QB late, usually in Round 10. The drafts took a few different players here: Philip Rivers, Patrick Mahomes, and Marcus Mariota, our QB14, QB15, and QB9, respectively. Mariota and Mahomes, while six spots apart in the rankings, are only projected 3.5 points apart, so there are plenty of good late options here.

Jack Doyle, TE, Round 9

Nine out of 12 optimal drafts took Jack Doyle. He’s our TE9. That’s pretty remarkable value in the ninth round.

Even with Eric Ebron on the roster, Doyle’s role seems secure. Ebron is more likely to take targets away from the wide receivers not named T.Y. Hilton.

Duke Johnson, RB, Round 8

One of Shawn Siegele’s Zero RB targets, Duke Johnson, ends up as a consistent pick in the eighth round. Until his ADP shoots up due to Derrius Guice’s injury, though, Chris Thompson is also an option here.

Will Fuller, WR, Round 7

In his games with Deshaun Watson under center last season, Will Fuller showed us that he has WR1 upside. While both Watson and Fuller are likely to regress this year, both have a long way to drop before they would no longer be values — especially Fuller at his ADP. Fuller is our WR28, a high-end WR3 available in the seventh round.

Those optimal drafts that could not get Fuller (because he might be taken) took Devin Funchess, who lands just below him in our rankings. While I strongly prefer Fuller, Funchess also has high upside.

Jarvis Landry, WR, Round 4

Moving forward in the draft, we have a few more picks that were nearly universal in the optimal drafts. The first of these is Jarvis Landry in the fourth round. He’s our WR14. With Corey Coleman now on the Bills, Josh Gordon AWOL, and Antonio Callaway facing possible suspension, Landry looks set up to see high volume once again.

Landry finished as the No. 4 WR overall in PPR last season and put up double digit points every week. He is a very safe WR2 in our rankings but could easily end up as a WR1 again.

Stefon Diggs, WR, Round 3

The receiver Shawn Siegele says most reminds him of Antonio Brown is the most frequent third round pick in the optimal drafts. Stefon Diggs lands as WR10 in our rankings.

While Adam Thielen stole the show last season, it’s important to keep in mind that Thielen, now 28, did so in the prime of his career, while Diggs at just 24 is still improving. (Thielen was my most-owned WR in dynasty going into last season, but I have now sold all shares, picking up shares of Diggs wherever I can.)

If Diggs is taken, then Amari Cooper, our WR13, should be available. Nine of the 12 optimal drafts took one of those two receivers in the third round.

Jerick McKinnon, RB, Round 2

Eight of the 12 optimal drafts took Jerick McKinnon in the second round. McKinnon looks slated for the Devonta Freeman role in Kyle Shanahan’s high-powered offense. However, compared to Freeman, McKinnon has two advantages: he is a ridiculous athlete and Shanahan is still his coach. I don’t expect McKinnon to see 20 touches a game, but he should have incredible upside with a Freeman-like 15 to 18 touch workload, particularly one that is heavy on receptions.

McKinnon lands in the same tier for us as Christian McCaffrey, who is likely to go earlier in the draft. They are our RB11 and RB12, respectively.

The Rest

The key picks above lock in all but three rounds: first, fifth, and sixth. Between those three, we need two more RBs and one more WR. If we can find enough RB values in Rounds 5 and 6, then a WR in the first round would be reasonable.

In the fifth round, the optimal draft usually takes Ronald Jones or Royce Freeman, who land as RB19 and RB23, respectively, in these rankings. Given Jones’s early struggles, though, compared to Freeman’s early success, the latter might be the safer bet. Both are fully capable of putting up RB2 numbers this season (especially later in the season, once they’ve proven themselves to coaches).

In the sixth round, the most common picks are Tevin Coleman and Rex Burkhead, our RB22 and RB25. Coleman remains undervalued due to being the second RB on the depth chart. This looks like the classic Zero RB situation: Coleman has plenty of value even with Freeman healthy, while an injury to Freeman (e.g., another concussion) launches him into a potential RB1. The other option is Burkhead, who I may like even more than Coleman. He’s being undervalued out of fear of New England running back rotation, but he has potential three-down use in his range of possibilities, he gets goal-line work, and, with Julian Edelman suspended for four games, there are extra short-range targets looking for a home.

If you draft early in the first round (where an RB pick seems required) or want to take a safe, bell-cow like Melvin Gordon later in the round, then you’ll need another receiver in the fifth or sixth round. The optimal drafts like Sammy Watkins in the sixth round. He’s our WR24. While he disappointed last season, tape watchers like Greg Cosell tell us that Watkins had no problem getting open, he just wasn’t placed by coaches in a position to get targets. The Chiefs, on the other hand, signed him to a $48 million deal with $30 million guaranteed. If Watkins inherits 76 targets from Albert Wilson’s role and takes a dozen or so from Tyreek Hill, who struggled at times trying to be the team’s WR1, Watkins can easily deliver WR2 upside.

Putting It All Together

Here is a complete example of an optimal draft from the 1.08 spot. It grabs Sammy Watkins late, leaving it free to take an RB, Melvin Gordon (our RB7), in the first round.

opt-rv-1.08

As we just discussed, however, DeAndre Hopkins or Odell Beckham would also be workable into this draft: simply swap Watkins for Coleman in the sixth round.

recent and related...

in case you missed it...

5 Fantasy Football Takeaways from Thursday Night

We’re kicking off week two semi-NFL football, and fantasy football Twitter is full of reminders that it’s only preseason. Twitter timelines are dominated with warnings of what we should and should not buy into for the coming fantasy football season – here are our preseason takeaways from the Thursday night

Read More
Connect
Support

rotovizmain@gmail.com

Sign-up today for our free Premium Email subscription!

© 2019 RotoViz. All rights Reserved.