strategy

 

 

What is the Right Draft Strategy?

Like many RotoViz writers, I’ve been studying the results of some previous drafts, and thinking about draft strategy. There’s a lot of information that can be gleaned from an exercise like this; much more than I present here. I think the data below is interesting and informative. In particular I’m intrigued by the impact draft slot has on draft strategy.

Draft Strategy

Much of your draft strategy is determined by your league size, roster requirements, and scoring settings. We’ve got a lot of great content that informs you about strategies old and new.

But an underappreciated element to draft strategy is draft position. Drafting late in the first round is very different than drafting early in the first round, as we’ll see.

The Draft Format & Other Considerations

For this exercise I studied all 239 MFL10 leagues that had complete data from 2012, looking only at the drafts of the winning teams. Information about the leagues:

  • 12 team, PPR leagues
  • 16 round snake draft
  • 1 QB, 2 RB, 3 WR, 1 TE, 1 Flex (RB, WR or TE), 1 PK and 1 DEF and 6 bench players

Unlike 2013 MFL10s, these teams were actively managed during the season. That means that owners were free to adjust lineups, drop, add, or trade players in-season, so the team they drafted isn’t necessarily the team with which they won their league. However, the team they drafted IS the team that allowed them to make whatever in-season moves they did make, so it’s still a very important consideration. Remember that the data presented are from the 239 teams that won their league. And I’m only really looking at the first 6 rounds; after that the divergence in picks and positions selected really accelerates- and I want to keep this to a readable length! Finally, this information is illustrative of what happened last year. While I think the lessons apply this (and every) year, this isn’t a predictive analysis.

Draft Position Matters

Here’s something most of us know but many of us haven’t internalized: your draft slot matters. Take a look at this chart. Here’s where the winning teams in my study drafted from.

draftslot.chart

If draft slot had no impact on winning, each draft slot would have the same win percentage (roughly 8%). But that’s not the case. You can see clearly that the first 4 draft spots have a much higher win percentage: just over 50% of all the league winners started their draft with one of the first four “early” picks. Only 25% of the league winners had a “late” first pick (between 1.08 and 1.12.) Let that sink in a little bit. Out of 239 leagues, only 62 were won by a player who picked 1.08 – 1.12. This is similar to something Brian Burke demonstrated at Advanced NFL Stats:

If you were to run a perfectly efficient snake draft, with the best player picked first, the second best player picked second and so on, the resulting teams would project to these point totals…  If you pick in the top half of the first round, you have a massive advantage.

optimized.chart

OK then. So if picking early in and of itself confers an advantage, should you employ the same draft strategy if you have an early or late pick? Intuitively the answer is no, and the data support that supposition. Late drafters required more flexible strategies – and better accuracy – to draft winning teams.

Unique Players by Round

This chart shows the number of unique players taken, by round, for three separate cohorts: “overall”, or all 239 teams, “early”, or the 122 teams with a pick between 1.01 and 1.04, and “late, or the 62 teams with a pick between 1.08 and 1.12.

unique.chart

Look at the early round drafters in round one. Only 8 different players were selected between picks 1.01 and 1.04 (122 total selections). But the late round drafters (62 selections) picked from over 20 different players. After round one, things reversed, with the early round drafters having a broader player universe to select from than the late round pickers.

What does this mean? Teams with an early selection grabbed the biggest, most obvious stud player available in round one, and then were increasingly free to select the “best available” player in later rounds. Lesson: don’t get fancy with an early first round pick. Take the stud RB. They also had a bigger margin for error, as we’ll see later. Conversely, teams with a later pick were forced to search farther abroad for a “first round stud”, and then were more limited in their later round selections, as they tried to catch up to the head start the early drafters enjoyed. In other words, the late drafters had a smaller margin of error- they had to pick from a smaller pool of players in order to win their league. That implies that the late drafters had to make more accurate judgments about which players to draft – or not – than their early pick brethren.

Early Pick, Positions by Round

Here’s how the early drafters distributed their draft picks, by position.

earlypick.chart

Players with an early pick chose a RB almost 75% of the time in the first round. Second round picks were evenly split between RB and WR. After that WRs ruled the day, with TE and QB increasing in rounds 5 and 6 respectively.

Here’s a by-the-pick breakdown of the first four picks.

Draft Picks Round 1 Round 2 Round 3
1.01, 2.12, 3.01 RB 100% WR 54%

RB 39%

QB 7%

WR 61%

RB 36%

QB 3%

1.02, 2.11, 3.02 RB 94%

QB 6%

RB 50%

WR 32%

QB 11%

TE 7%

WR 54%

RB 46%

1.03, 2.10, 3.03 RB 67%

QB 29%

WR 4%

RB 45%

WR 45%

QB 7%

TE 3%

WR 71%

RB 29%

1.04, 2.09, 3.04 RB 39%

QB 33%

WR 28%

RB 56%

WR 27%

QB 10%

TE 7%

WR 64%

RB 30%

QB 6%

Takeaways

With an early pick, taking a RB with the team’s first pick was clearly the best move. Going RB again in the second round was also a winning formula, as was taking advantage of the 2nd/3rd round turn to take two top WRs. No matter which early pick a team had, however, taking a QB early was rarely successful.

Late Picks, by Position

Now let’s take a look at what the late drafters did.

latepick.chart

Wow. That’s very different than what the early drafters did. Only 40% took a RB in the first round – and 30% took a QB! Late drafters chose RBs in the second round at about the same rate as early pickers- but were much more likely to take a TE than early drafters. In round three the late drafters achieve the RB/WR symmetry that the early drafters got in round two, and then make a huge swing to WRs in round 4, focusing more heavily on WRs through round 6, compared to their early drafting counterparts.

Clearly, with a late pick, going RB-RB was not the only way to go. Let’s take a closer look at how late drafters spent their picks. There were 25 distinct methods used for the late pickers’ first three selections. The immediate implication is that the late pickers had to be more flexible and adaptive than their early picker counterparts. The other takeaway is that none of these strategies was nearly as successful as the early pick RB-RB strategy. Specifically, going RB-RB with a late first pick only won 8 of 239 leagues.

Here’s how the late pick strategies broke down.

Strategy Times Used Success Rate vs. Other Late Pick Strategies
6 QB first strategies 21 0.338
10 RB first strategies 26 0.419
5 TE first strategies 8 0.129
5 WR first strategies 7 0.113

With a late first round pick, the best strategy was still to select a RB, but choosing a QB was not far behind. Choosing a WR or TE first was not a good idea! Collectively, taking a TE or WR first with a late pick won only 15 times out of 239 leagues.

Here’s a look at all of the late pick strategies that won more than once. These are how the picks for rounds 1, 2, and 3 were spent. Q=QB, R=RB, W=WR, T=TE.

Strategy # Times
RRW 8
QRW 7
QRR 4
QTR 4
RWR 4
RWR 3
RWW 3
TRW 3
QTW 2
QWR 2
QWW 2
RQR 2
RTW 2
TRR 2
WRW 2
WWW 2

Success Rate of Picks

Point of clarification: Do not look at these tables and think “well, if I’m a late picker, I can just do whatever I want”. It would be more accurate to say “if I’m a late picker, it’s crucial that I get the good value for each pick”. Let’s define “value” as getting a starter (top 24 WR or RB, top 12 QB or TE). That’s really simplistic, but it illustrates the point.

Late drafters only chose from 39 unique players over the first three rounds of the draft. Only six of them (Fred Jackson, Percy Harvin, Greg Jennings, MJD, Ryan Mathews, and Darren McFadden) failed this simple value screen. Late drafters therefore hit on 84% of their first three draft picks.

Early drafters however, selected 66 unique players over the first three rounds, far more than the late drafters did. Twelve of them (Vick, J.Stewart, Jennings, Harvin, MJD, McGahee, Helu, Hernandez, Gates, Finley, Antonio Brown, Bowe) failed this value screen. So the early drafters hit on 82% of their first three picks. Despite hitting 100% on their 1st round picks, they only hit on 79% of their second and third round picks.

In other words, the late drafters that won their leagues were better evaluators. Or at least they did a better job of sticking to high value selections: it’s possible the early drafters were irrationally buoyed by getting a great player with their first pick, and then got a little too careless or aggressive with their later picks.

Takeaways

If you’ve got an early pick, congratulations: enjoy our RB-RB (or RB-RB-RB) content. But remember to stay focused, and don’t get too risky in the middle rounds.

If you’ve got a late draft pick, you need to be more flexible in your draft strategy. Sticking to a position-by-round formula really doesn’t help. It’s less important to get any given combination of positions (though RB first is still a good idea) than it is to get the most value possible with each selection.

This is where RotoViz has your back. Our Apps help you consider each player’s potential production in a fresh way, as does our content. In particular you’ll want to be familiar with these pieces. Each will help you squeeze the most out of each draft pick:

Bonus Nugget

We really shouldn’t need to be reminded of this, but apparently we do. Don’t take kickers earlier than necessary. First, consider that of 239 winning teams only 217 even drafted a kicker. In other words, they drafted an extra position player, then dropped a player (or traded two players for one) before the season began. This gave them an extra shot at getting a productive player in their lineup. Also consider that only 26% of the winning teams took a kicker before the final two rounds of the draft.

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

1 comments
joshyjosh49
joshyjosh49

Great stuff Charles. I'm really enjoying my Rotoviz subscript. I'm down to only two leagues these days (the oldest ones), one is an auction, other is snake. While I think auction is superior (the subject of this article sort of proves that) I just can't quit the snake. I think the challenge of winning from various draft slots over the years is part of the appeal. Your article really can't be overstated...draft slot matters, and a bad draft slot can really doom you since it leaves a smaller margin of error. It's just the reality of it. And probably the main reason I would recommend to anyone to try an auction draft if they haven't already.