When To Draft QBs In Best Ball (Updated)
The fantasy football landscape is always changing, but perhaps nothing crazier than the QBs being priced back up again after years of late-round QB success. Is this recent success noise? Or is this a trend worth following? Underdog Fantasy's Hayden Winks combines data, theory, and 2023 projections for his When To Draft QBs In Best Ball feature. The WR and RB versions of this column are here.
QB Price Tags Over Time
These charts from above require some nuance to properly use them. One major aspect is when QBs have been drafted on Underdog Fantasy.
2020: There were two "elite" QB1s. Lamar Jackson and Patrick Mahomes were on the Round 2/3 border, while the QB3-QB6 were drafted in Round 6-7.
2021: There was one "elite" QB1. Mahomes was drafted in Round 3/4, while QB2-QB6 were drafted in Rounds 5-6. In general, QB were slightly higher than in 2020.
2022: There was one "elite" QB1. Josh Allen was drafted in Round 3, while the QB2-QB6 where drafted in Round 4-6. In general, QB prices were slightly higher than in 2021.
2023: There are four "elite" QBs. Mahomes, Jalen Hurts, Allen, and Jackson are drafted in Rounds 2-3, while QB5-QB8 are drafted in Rounds 4-7. In general, QB prices are definitely higher than at any point.
What's most interesting with 2023 ADP is that it's not just the super elite QBs getting priced up. The QB5-QB8 range is also up more than one full round, as is the QB16-25 range. If most QB pricing is up, it takes away from the opportunity cost of the truly elite QBs moving up because we all have to draft QBs at some point in the draft. Right now, the only QB prices that are down relative to 2022 prices are QB9 (Deshaun Watson), QB10 (Anthony Richardson), QB11 (Dak Prescott), and QB14 (Daniel Jones).
Are Elite QB1s Worth It?
Half PPR best ball leans towards QBs. Removing a half point per reception compared to PPR hurts all positions except for QB most importantly, but the late round QB strategy is also slightly diminished because there aren't waivers, which we use to stream the best weekly matchups among the available QB2s.
The other main thing to note is how different today's NFL is. Before 2018, there were 12 NFL seasons where a QB averaged more than 23.2 fantasy points per game. Ever. Since 2019, there are four different QBs who've averaged at least that many points per game over multiple seasons. The biggest differences are teams are passing more while playing faster, and more importantly, the best passing QBs also happen to run. Since rushing production > passing production in fantasy, there are easier upside cases now than before. Taking old data and applying it to this environment isn't sound in my opinion, at least while these elite dual threats are squarely in their primes.
... ... and don't get me started about the potential of all the big athletic QBs copying the 2022 Eagles short-yardage packages, elevating their own fantasy stats while decreasing the RB, WR, and TE numbers in general ... ...
We've also been more predictive with our preseason QB rankings and their actual finishes in recent seasons. I think this is because the new elite QBs have been elite QBs for a couple seasons now. We're also properly pricing in rushing production. No longer are the days of Lamar Jackson types in Round 15. Jalen Hurts before his eruption last year was the QB6. Anthony Richardson without playing a single NFL snap is the QB10 this year. To me, it'll be pretty hard to find late-round QBs who have the ceilings of the elites now that we're pretty good at making projections at this position.
That brings me to weekly ceilings, as this is best ball at the end of the day. I wrote about spike weeks here, but to summarize, the current elite QBs have accounted for 69% of the spike weeks since 2021. That number could climb in 2023 now that we're removing Tom Brady (and likely Aaron Rodgers) types from this sample and adding more spike weeks from emerging dual threats like Justin Fields and Hurts.
I feel good about the stability of our elite QB projections and their ability to soak up the spike weeks, so the remaining question to solve for is opportunity cost of drafting a QB over the other positions. Let's break that down over time.
2020: Lamar (7% advance) regressed hard. Mahomes (22%) was good. Dak (12%) broke his ankle. Kyler (24%) was good. Russ (34%) was very good. Deshaun (30%) was very good. And Josh (37%) was the best. In total, the top-7 QBs in ADP averaged a 24% advance rate, compared to the standard 17% across the site.
2021: Mahomes (14% advance) was fine. Allen (22%) was good. Lamar (18%) was injured but was pacing well. Kyler (14%) was injured and didn't run as much. Dak (14%) was fine but didn't run as much. Russ (11%) was injured and didn't run as much. Herbert (29%) was a stud. In total, the top-7 QBs in ADP averaged a 17% advance rate, which was right in line with the standard rate. Q11 in ADP Tom Brady finished with the best advance rate (31%) with QB13 Matthew Stafford in third (26%) after outlier passing TD years.
2022: Allen (26%) was good. Herbert (12%) was bad and injured. Mahomes (31%) won the MVP. Lamar (17%) had spike weeks but was injured. Kyler (13%) was mid and injured. Hurts (34%) was the best QB pick. Burrow (21%) was good. That's a combined 22% advance rate.
So far, elite to mid-range QB1s have been average to good in all three best ball mania seasons. That's why their ADPs are (rightfully) up in 2023.
When Zero QB Tanks
For the late round QB folks.
2020: It started to tank in Round 7-8 and became very clearly sub-optimal in Round 11.
2021: It really tanked in Round 7-8.
2022: It really tanked in Round 4-6.
When To Draft QB2 On Elite QB1 Teams
I have a full column on this here, but the early data is that two QBs before Round 7 is doing too much, unless you're trying to get really unique with builds. Instead, the opportunity cost for a QB2 is minimized in this Round 9-10 range and beyond. Let's track it by year:
2020: QB2 teams were at or nearly optimal from Round 9 to Round 18.
2021: QB2 teams were at or nearly optimal from Round 4 to Round 18. We're dealing with very small samples before Round 7, but QB2 teams were definitely optimal after Round 9 on average. This was a year where 2-QB builds were better than 3-QB builds.
2022: QB2 teams paced below QB1 teams until Round 12, though the gap was small from Round 4 and on.
It looks like when we hit on a QB, it's best to wait until this Round 9-12 range, but we haven't sacrificed that many points with 2-QB builds as early as Round 6-7 throughout all three seasons of best ball mania. That's a pretty big surprise versus the narratives I hear.
When To A Draft QB3
When done properly, I think this may be a little underrated, for game theory reasons. In these best ball tournaments, especially as it becomes harder to advance in each playoff round, the ownership in Week 17 finals will continue to consolidate around the players who popped off in the week or two weeks prior. To get unique, having multiple QB options (the position that scores the most points) can add leverage, especially when we're team stacking.
For example, we ride our Herbert/Chargers team stacks in the regular season. They have a decent enough game in Week 15 to advance us to the next round. Then he has a massive Week 16 game where it'll be hard to advance without them. When we go into the Week 17 finals, Herbert is owned on 40% of teams. At that point, we'll want multiple other team stacks at low ownership to vault us up boards in Week 17. This happened with George Kittle (39% owned in Week 17) last year after rampaging in the week's prior, who then finished as the TE12 on the week in the finals. Meanwhile, Dalton Schultz (4% owned in Week 17) after some duds finished as the TE1 on the week.
2020: QB3 teams became optimal or nearly optimal from Round 14-18. There was no practical difference in average scores between 2- and 3-QB teams.
2021: QB3 teams closed the gap on QB2 teams in Round 11, but QB3 teams were never optimal at any point on average. In fact, QB2 teams stayed optimal from Round 8 and on.
2022: Once again, QB3 teams closed the gap on QB2 teams in Round 12, but never eclipsed them at any point on average.
I'm sure if I did case studies, the best times to add a QB3 are when skipping the elite QB1s. I did a uniqueness column and found the double-tapping QBs with ADPs very similar to each other was less-drafted, so perhaps going with 3 QBs in this Round 9-13 range is a fun strategy for galaxy brainers out there.
2023 QB Strategy
So let's take all this information and apply it to this year's player pool. In my opinion, it'll be very hard to find season-long and even relative weekly ceilings from QBs outside of the top-10 (and there are more questions around QB9 Deshaun Watson and QB10 Anthony Richardson than the top-8 very clearly). In previous years, it's been MVP candidates Tom Brady, Aaron Rodgers, and Matthew Stafford with an aggressive play caller hanging on the QB1/2 border. This year, it's only QBs who have never eclipsed mediocrity.
Dak Prescott: He's had very good stretches, but he's running less post-ankle and now has a likely downgrade in play caller.
Tua Tagovailoa: He big games in 2022, but that's a small sample and defenses played Miami differently later in the season, which hurt his production. There's also elevated concussion risk. Alas, there's upside for sure.
Kirk Cousins: Kirk Cousins.
Daniel Jones: He was low-end QB1 fantasy viable on outlier rushing production in 2022. He's been very mid aside from that.
Geno Smith: One year of low-end QB1 production.
Aaron Rodgers: Aging and playing on a new team with a good defense.
Jared Goff: Only 17.1 fantasy points per game when offense was rolling.
Russell Wilson: Theoretical bounce back appeal, but the rushing is toast.
Derek Carr: Derek Carr.
Jordan Love: Didn't even bet on himself in contract negotiations.
Matthew Stafford: Aging with worse roster, but some upside.
Kyler Murray: Rehab risk and unlikely to run as much, plus no Nuk.
With the elites now dropping into Round 3-5, I'll be buying as many of them as possible. Most will be paired with the QB3 tier QBs (Stafford, Stroud, Young, Pickett, Brock, Mac, etc.), and I'll even mix in some 3-QB builds really late in drafts. I'll also mix in an elite QB1 with this Dak-Tua-DJ-Kirk-Geno tier if stacked with their pass-catchers. But I won't be leaving too many drafts without a top-10 QB. So far, my highest drafted QBs are:
Richardson: 18%
Stroud: 18%
QB5 Fields: 15%
Stafford: 15%
QB7 Herbert: 13%
Love: 13%
QB3 Allen: 13%
QB2 Hurts: 10%
QB4 Lamar: 10%
Watson: 10%
QB1 Mahomes: 8%
QB6 Burrow: 8%
Mac: 8%
That's 77% of my drafts with a top-7 QB. Wish me luck.
And go read Stacking Elite Rushing QBs in Best Ball.