Is Your Fantasy Football Team Better Off Streaming QBs?
I Have a Stream. Streaming of You. Streamcatcher. I could go on all day with “stream” puns. While puns are a joke, the idea of streaming quarterbacks is not.
What is Streaming?
For those who aren’t quite sure what we mean, the concept of streaming is that instead of owning one particular player, in this case, a quarterback, instead, you dedicate a roster spot to a rotating cast where the actual player changes week-to-week based on matchups. In simplest terms, you pick up a new QB just about every week. This allows you to spend all of your draft capital on WRs and RBs with the idea that the advantage you gain at those positions will make up for the hit you take at QB.
How Many Points to Target?
Let’s start by taking a look at the top 12 QBs from 2016, using FantasyPros NFL QB statistics. We will use per game averages and exclude Week 17 so as not to mislead people into thinking Tom Brady was anything other than his usual elite self (he finished outside the top 12 in total points due to his suspension). Aaron Rodgers was 2016’s best fantasy passer with 23.1 points per game. Cam Newton was the last of the QB1s at 17.6 ppg. So, the question is whether you can get roughly 17.5-23 fantasy points from employing a new QB each week?
Identifying Streaming Targets
Next, we need to look at QB fantasy points scored based on opponent. Cleveland was the most generous to opposing QBs, allowing 19.3 fantasy points per game (fppg). San Francisco allowed the 12th most fantasy points to opposing QBs at 17.4 fppg. Based on this information alone, we can reasonably deduce that there was a decided advantage to having one of the five best QBs in 2016. Aaron Rodgers, Matt Ryan, Tom Brady, Drew Brees, and Andrew Luck all averaged more fppg than the average QB facing the Browns. But the No. 6 QB, Kirk Cousins, was equally as good as whatever QB was facing the Browns. Let’s dig a little deeper into the Browns. For the sake of brevity, I will limit the weekly analysis to the Browns, but you have the tools to go deeper into each team’s schedule should you choose to do so.
The most important aspect in analyzing the effectiveness of streaming is whether it was even possible to stream. If the bulk of the Browns’ opponents were top 12 QBs, it’s reasonable to assume you wouldn’t have been able to stream even if you wanted to. Acknowledging that this is admittedly an inexact science and that there are typically more than 12 QBs owned at any given point in your league, for the purposes of this analysis, we will focus on how many times the Browns faced a QB that finished outside the top 12. Using the aforementioned criteria, the Browns faced 10 non top 12 QBs. Even adjusting for what QBs I expect were owned throughout the season, I count 10 times you could reasonably expect to be able to stream against the Browns. So we have 10 streaming opportunities against the worst pass defense in the NFL in 2016. There are nine other defenses that could’ve been exploited to create a QB1. Based on the streaming opportunities against the Browns alone, we can reasonably infer that finding a weekly QB against one of the 10-to-12 worst defenses in the league is something you can consistently do.
Building a 2017 QB1
Let’s take a look at the 2017 schedule and see if we can build a QB1.
In the interest of realism, I am going to limit our player pool to QBs I believe will be largely unowned this season. Here’s my list: Blake Bortles, Carson Wentz, Carson Palmer, Alex Smith, Ryan Tannehill, Joe Flacco, Sam Bradford, Josh McCown, Jared Goff, Brian Hoyer, Broncos QB, Texans QB, Bears QB. My “Frankenstein’s Quarterback” list will include the player, opponent, and in parenthesis will be that team’s 2016 fppg allowed.
Week 1: Bears QB vs ATL (2nd), Palmer at DET (5th), Goff vs IND (7th), Bradford vs NO (11th)
Week 2: Flacco vs CLE (1st), Bortles vs TEN (3rd), Palmer at IND (7th), McCown at OAK (9th)
Week 3: McCwn vs MIA (4th), Tannehill at NYJ (6th), Hoyer at LAR (10th), Goff vs SF (12th)
Week 4: Texans QB vs TEN (3rd), Bradford vs DET (5th), Bears QB at GB (8th), Broncos QB vs OAK (9th), Tannehill vs NO (11th)
Week 5: McCown at CLE (1st), Tannehill vs TEN (3rd), Hoyer at IND (7th), Flacco at OAK (9th)
Week 6: Texans QB vs CLE (1st), Tannehill at ATL (2nd), Bradford vs GB (8th), Bortles vs LAR (10th)
Week 7: Browns QB vs TEN (3rd), McCown at MIA (4th), Tannehill vs NYJ (6th), Bortles at IND (7th), Palmer at LAR (10th)
Week 8: Bradford at CLE (1st), McCown vs ATL (2nd), Flacco vs MIA (4th), Bears QB at NO (11th), Wentz vs SF (12th)
Week 9: Flacco at TEN (3rd), Texans QB vs IND (7th), Tannehill vs OAK (9th), Palmer at SF (12th)
Week 10: Browns QB at DET (5th), Bears QB vs GB (8th), Texans QB at LAR (10th)
Week 11: Bortles at CLE (1st), Bears QB vs DET (5th), Flacco at GB (8th), Bradford vs LAR (1oth)
Week 12: Bradford at DET (5th), Broncos QB at OAK (9th), Goff vs NO (11th)
Week 13: Texans QB at TEN (3rd), Broncos QB at MIA (4th), Smith at NYJ (6th), Bortles vs IND (7th), Bears QB vs SF (12th)
Week 14: Palmer vs TEN (3rd), Broncos QB vs NYJ (6th), Browns QB vs GB (8th), Smith vs OAK (9th), Wentz at LAR (10th), Texans QB vs SF (12th)
Week 15: Flacco at CLE (1st), Hoyer vs TEN (3rd), Bears QB at DET (5th), Broncos QB at IND (7th), McCown at NO (11th)
Week 16: Bears QB vs CLE (1st), Goff at TEN (3rd), Smith vs MIA (4th), Flacco at IND (7th), Bradford at GB (8th), Wentz vs OAK (9th), Bortles at SF (12th)
It goes without saying that the 12 worst defenses in 2016 will certainly not be the same 12 in 2017. But that’s okay. Just consider this a rudimentary blueprint of a more concrete plan that is malleable and can and will change during the season. More so than attempting to adhere to some form of this schedule, focus on the fact that even when some of these defenses improve and some of these QBs end up being rostered consistently, odds are you will be able to find this year’s newly poor pass defenses to exploit as the season goes on.
Are you Better off Streaming?
Maybe. If you can secure one of Rodgers, Brady, Brees, Ryan or Luck, then statistically you are better off with one of them. If you can secure a top 12 QB in one of the last few rounds, the results will likely be similar. QBs 10-12 average about 18.0 fppg last year, which is about what you can reasonably expect to get if you do a really good job streaming against the bottom 10 defenses, but aren’t able to get that QB facing the two or three worst defenses each week.
I think there’s a “Goldilocks Zone” for streaming if you will. It exists from the time the fifth best QB comes off the board until about the 11th or 12th round. That is when your draft is poised to extract the most value out of employing the streaming QB strategy. During those rounds, you can still grab WRs and RBs very capable of helping your team while the differences between the QBs isn’t all that much. Last year, the difference between the QB9 and the QB15 was one point per game. The QB6 to QB8 were also only separated by 0.5 fppg. The drop off from Andrew Luck (QB5) to Kirk Cousins (QB6) was 1.4 fppg. It’s not all that surprising that this year’s current QB5 by ADP is going 20 selections higher than this year’s current QB6. But this year’s QBs 10-12 are all currently going somewhere in the first half of the 10th round. The QB13 doesn’t go until the 14th round. If last year is any indication, the QB13 is not four rounds worse than the QB10. The QB13 is almost definitely not seven rounds worse than the QB6 or QB7. Rather than burn a late single-digit pick on a QB that only gives you a marginal advantage, continue taking WRs and RBs with two backup plans: a) take a top 12 QB late because they happened to fall, or b) stream.