Why is it that, if injuries, suspensions, and poor performance force second-string and even third-string running backs, wide receivers, and tight ends into action, we see the opportunity to roster undervalued players — for instance Mike Gillislee and Jacquizz Rodgers in Week 7 — but when a backup quarterback starts a game we think that the world is going to end.
Because quite simply nothing affects an entire offense more than a QB change.
With that in mind, I want to explore this question: What position does a QB change hurt the most, and what does that mean for DFS?
What Does The Data Say?
For the purpose of this study, I looked at 2016-17 (through week 6) as well as the entire 2015-16 season. Teams were excluded who satisfied one of the following.
– Their QB played all 16 games.
– Their QB situation looks like this with no clear rhyme or reason.
– Also, situations like Tom Brady’s suspension this year are inverted. So Weeks 1-4 without him are lumped in with the ‘replacement QB’ category.
I recorded the Plus/Minus on FanDuel for each player, broken down by RB1, WR1, WR2, TE, kicker, and defense/special teams for games played with the starting QB and the replacement QB. The RB1 was chosen as the player with the most projected points and the WR1 and WR2 were differentiated the exact same way in each game.
By the way, a player’s Plus/Minus is his actual points minus his expected points. Here’s an example: In a game with Ben Roethlisberger in which Antonio Brown scores 25 points and his expectation is 20 points, he has +5 Plus/Minus.
The following chart displays the average FD Plus/Minus values for each position in each QB situation:
In a sample of 214 games, WR1 was by far the position most impacted negatively by a QB switch, moving from +2.66 to +0.01 points above salary-based expectation.
WR2 is also severely impacted, losing 62.1 percent of its Plus/Minus value in the QB switch.
And TE is also impacted, losing 1.69 (or 44.5 percent) of its Plus/Minus. TE relies heavily on game flow and most positively correlates with home favorites with a large spread. Not only is a team with a replacement QB much less likely to be favorite, but it’s also less likely to be in scoring position as often.
DST is also massively impacted by a QB change for a lot of the reasons TE is. Defensive points are heavily reliant on turnovers and defensive touchdowns. If a team is playing from behind, it’s much more likely to take chances to get back into the game. If it’s ahead, it will likely be conservative with the ball, and the opposing DST’s opportunities to create turnovers and score TDs will be limited. If a team with a replacement QB struggles to move the ball and put up points, the DST simply might not be in a situation to accumulate as many points.
Why Do Replacement QBs Destroy WR1s?
Intuitively, it makes a ton of sense. Poor quarterback play negatively impacts the offense as a whole, but a WR1 is likely the player getting the most targets in said offense. Targets are now coming from a lesser talent when a replacement QB is throwing the ball.
Also, the salary is something we must consider. Past performance is a lot of what goes into a salary, and it is reasonable to assume that salaries are not corrected as aggressively as they should be after a QB change. Also, in some cases QB changes occur after salaries have been released. For example, Antonio Brown’s salary in Week 7 is $9300 on FanDuel, down $500 from last week with the loss of Ben Roethlisberger.
But Brown is still $9300!
This is a drastic example — and it’s possible that FD determined Brown’s salary before news of Roethlisberger broke — but it’s clear that even a player of Brown’s caliber is at risk of being severely limited by a switch to Landry Jones. And this situation isn’t really unique. It happens every year.
So it’s easy to wrap your head around the fact that quality of targets directly impacts WR1 scoring and points per dollar. It’s also probably pretty easy to see how a QB switch can have a huge impact on an offense’s ability to stay on the field and generate red zone opportunities, which is why we see Plus/Minus values go down across the board.
It Can’t Be that Bad, Right?
Let’s stick with the example of Brown and the Steelers offense: We’ve seen this before and it’s not pretty:
Without Roethlisberger, arguably the best wide receiver in the league has been very average and definitely not worth a top-tier DFS salary. Brown may in fact still have a nice floor, but his ceiling is absolutely lowered due to the limitations of his new QB.
And What About RB?
Although RB1 is not affected nearly as much as WR1, the QB switch still makes a negative impact. Let’s take a look at Le’Veon Bell’s splits with and without Roethlisberger:
Interestingly enough, Bell’s overall production is comparable.
But let’s look at the details. He has more carries, which we’d expect. Teams tend to lean more heavily on their running game with a replacement QB. In this situation, a RB might have fewer opportunities to score and might face more stacked boxes, but the increase in volume could compensate somewhat for the loss of efficiency and goal-line opportunity.
Bell also has fewer targets, which we’d probably expect: If the replacement QB throws the ball less than the starting QB normally does, it makes sense that the RB as a receiver would see fewer targets.
And it would also make sense that on a per-target basis the RB would be less productive — and that’s what we see with Bell. Not only does he see fewer targets, but he also does less with his targets, accumulating fewer yards and touchdowns per target. And that’s not because Bell is less of a player. It’s because the guy throwing him the ball is less of a passer.
When a starting QB goes down, hope for the best but plan for the worst . . . and probably fade the passing game.
Nothing affects an entire offense more than a QB change, and a lot of times pricing is not adjusted aggressively enough to make for an attractive situation.