NFL Best Ball: Roster Construction
Putting together an optimal squad is nuanced and fluid, with several variables always in play. Drafts are dynamic and roster construction acts as scale of balance.
Quarterback (2-3 picks)
In 2019, drafters have the luxury of selecting from what might be the deepest quarterback class in recent memory. Gone are the days of major tier drops, and instead, an already static position is dominated by passers that are nearly impossible to differentiate. Per FantasyData.com, there are only a five points per game separating projected QB1 Patrick Mahomes and QB15 Mitchell Trubisky.
For Playdraft the sweet spot for the position is between two and three to a roster. Per Rotoviz, drafting two quarterbacks has the highest win rate at 8.5%, though it’s barely above the 8.4% win rate for selecting three quarterbacks. With only one of each in the starting lineup, the rule I tend to adhere to in an 18-man format is deciding between two quarterbacks and two tight ends. I will deviate and take three of each if I’m satisfied with my stable of running backs.
What about quarterbacks in the FFPC and Fanball (BB10s)? Conceptually it would make sense for the need for extra signal callers in deeper formats. A quick roster breakdown can be found in this Bestball 101 article.
From the chart below we can see that the overall win rate is incredibly similar on Playdraft with the slightest of edges going to two quarterbacks. In FFPC, rostering three quarterbacks gives fantasy owners quite the boost. This becomes more critical because a trio of signal-callers improve your chance of spike weeks and stacking possibilities. Surprisingly, BB10s have a higher win rate by 0.7% while drafting only two passers. The takeaway here is to understand the minute but critical differences each platform brings.
*All win rate data is taken from the Rotoviz Roster Construction Tool
Waiting to draft the position is the way to go, although the trick is to not wait too long. The two-quarterback approach in both formats is a major boon (above 8% win rate), but an added advantage can be taking both passers with in a similar range.
In BB10s, the win rate jumps to 10.2% if you grab your first quarterback after the fifth round and before the 13th round, followed by another passer after the eighth round (and before the 13th round). In Playdraft, selecting both quarterbacks between rounds 10 and 12 led to a whopping 10.4% win rate.
Running Back (5-6)
Have you ever seen a video of a Black Friday sale with the crowd stampeding into stores? Early on in drafts, the running back position is the same way. So much for Zero RB or going against the grain, it’s a mad dash to grab bell cows early and often. This can lead to surrendering value for perceived scarcity. It’s easy to get caught up in the moment and lose sight of your team construction. So what’s the optimal number of running backs on your roster? Yet again the answer is contingent on your platform of choice. The below chart highlights the optimal build by platform for the 2017-18 seasons.
Below is the win rate change when adding an extra running back. Again, balance is crucial and roster construction acts as the skeleton key.
Depending on draft slots I tend to take five running backs if I hit the position hard early. Six might be more prudent if you are taking multiple pass catchers in the first few rounds. For the most injury-prone position in a shallower format, its key to draft backs late who have standalone value but also the upside to jump tiers.
I don’t highly recommend handcuffs unless there are unique circumstances. For example, if you have the intestinal fortitude to draft Todd Gurley, it’s not a bad idea to pair him with Darrell Henderson a few rounds earlier than his ADP.
Wide Receiver (6-8)
The magic number across formats for wide receivers is right around seven. Unlike tight end and quarterback, pass catchers offer more flexibility both from a value and build perspective, while avoiding the injury concerns associated with running backs. All three formats require three wide receivers in starting lineups, making them the most heavily-rostered position.
The change in in win rate when rostering one more or one less than seven is relatively minimal.
My preferred start out to a draft in MFL10s this year is a stud running back with my first pick, followed by three wide receivers. Pairing a bell cow with formidable receivers early can increase your chances of success significantly, as seen below.
Using the RCE from Rotoviz we can see that the win rate soars to 10.9% when starting out with the above roster configuration.
Tight End (2-3)
Tight End is another single-starter position but unlike quarterback, we are presented with several dilemmas including scarcity, high variance, and increased injury concern. There isn’t a position with a greater tier drop.
On any Best Ball platform, I find myself shooting for dual elite tight ends early. On Playdraft this creates a monster win rate of 11% if you take two before the end of the fourth round.
Here is a breakdown by number of tight ends selected:
We can see that in the FFPC the position is in higher demand given its larger rosters and TE premium format. On Playdraft, it feels risky only grabbing two tight ends but it’s the most viable strategy.