The following is an analysis of QBs' fantasy value based purely on data compiled across the last four seasons (2015-2018), featuring average draft positions (ADPs) for each year's top 50 preseason quarterbacks, as well as fantasy scoring for every quarterback (roughly 70-75 per year). ADPs are based on numbers compiled right before the start of the regular season.
First, the "QB Fantasy Points" graph is self-explanatory. This shows how many points each top-50 ADP QB has scored in the past four seasons. The highest scorer, Pat Mahomes, is visible above the 400 line. Keep this chart in mind as we go through the rest--specifically QB1s are not consistently and convincingly more reliable than most QB2s.
The next set of results is titled "QB Average Scoring Rank Based on Preseason Positional ADP." The graph consists of preseason positional ADP across the X axis and average scoring ranking along the Y. The number all the way to the left, "20," signifies that over the past four seasons, the top-rated preseason QB has averaged the 20th most fantasy points at season's end. For context, Andrew Luck was the preseason #1 in 2015 and finished 28th, while Aaron Rodgers was #1 in 2017 and finished 29th. Both played in only seven games due to injuries. 2017 #1 Cam Newton was 17th best, top-rated Rodgers finished 6th in 2018. Average those four finishes together, and you get 20.
What jumps out at me with this graph is, based on 12-team league tiers, QB1s are averaging roughly the same end-of-season ranking as QB2s. The average finish for a top-12 QB is 14th. The average finish for a top 13-24 QB is 17th. Based on 10-team tiers, the top 10 QBs average 13th, and the top 11-20 QBs average 16th.
To be clear, I'm not saying "Don't draft elite QBs." But the numbers are fascinating, in that they reinforce a belief I and many of you have--that streaming a couple of high-upside QB2s could yield as many or more points as starting an elite QB every week. This conclusion is compounded by the even more shocking discovery that the top 5 QBs, on average, score the 11th most fantasy points. The same is true for QBs 12-16. Yes, injuries play a role in leveling these numbers. But injuries are hardly the whole story. And while we're talking about only four seasons, the numbers are interesting enough to merit more research in the coming seasons.
Next up are four charts called "Expected Performers and Surprise Performers," separated by the top 50 QB scorers, the top 25, the top 10, and the top 5. The X axis signifies the top 50 fantasy scorers. Starting with the top 50 chart, the green line above the "1" on the left reflects the fact that each season's #1 fantasy QB owned a top-50 preseason ADP. Of course, right? Because it would be highly unusual for a back-up or third-string preseason QB to finish anywhere close to #1. The first small blue line on this chart appears above the "14," which reflects Case Keenum's astonishingly strong 2017 campaign, when he finished 14th despite owning a preseason QB-55 ADP.
A few things jump out at me. In the top 50 chart, only two non-top-50 preseason QBs have earned streaming value in the past four years. In other words, surprises like Keenum are highly unusual, though in two-QB leagues many non-top-50 options have offered streaming potential--some hope for two-QB-leaguers who miss out on an all-important #3 QB. The top-25 chart shows how often a QB with an ADP above 25 ends up as a top-25 fantasy scorer: 19% of the time, to be exact. And based on the top-10 chart, a full 50% of the time a QB with an ADP above 10 finished in the top 10. Some of those QBs had ADPs in the 11-14 range, so not that big a deal. But many were huge ADP shocks, like Alex Smith in 2017. And of course the top-5 chart looks brutal, and strengthens the notion that selecting two high-upside, back-end QB1s or high-QB2s in the middle rounds could be a better play than burning an early pick on a supposedly elite QB1.
This data isn't intended to be actionable in isolation. But it's thought-provoking, at least to me. "Sure-thing" QBs don't always produce, and middle-third QBs are not always merely streamers. And in fact, sometimes a good chunk of the latter group performs better than a good chunk of the former group.
This data should motivate us to do more research and take more fact-based risks. Preseason ADPs are based on educated guesses, and often are based on the previous season's numbers. That's no way to devise a draft strategy.
Tomorrow I'll offer some RB analysis . . .
---
Sign up for Pre-Draft Top 300 Rankings: http://www.fantasyfootballforwinners.com/
Follow on Twitter: https://twitter.com/_FF4W
First, the "QB Fantasy Points" graph is self-explanatory. This shows how many points each top-50 ADP QB has scored in the past four seasons. The highest scorer, Pat Mahomes, is visible above the 400 line. Keep this chart in mind as we go through the rest--specifically QB1s are not consistently and convincingly more reliable than most QB2s.
The next set of results is titled "QB Average Scoring Rank Based on Preseason Positional ADP." The graph consists of preseason positional ADP across the X axis and average scoring ranking along the Y. The number all the way to the left, "20," signifies that over the past four seasons, the top-rated preseason QB has averaged the 20th most fantasy points at season's end. For context, Andrew Luck was the preseason #1 in 2015 and finished 28th, while Aaron Rodgers was #1 in 2017 and finished 29th. Both played in only seven games due to injuries. 2017 #1 Cam Newton was 17th best, top-rated Rodgers finished 6th in 2018. Average those four finishes together, and you get 20.
What jumps out at me with this graph is, based on 12-team league tiers, QB1s are averaging roughly the same end-of-season ranking as QB2s. The average finish for a top-12 QB is 14th. The average finish for a top 13-24 QB is 17th. Based on 10-team tiers, the top 10 QBs average 13th, and the top 11-20 QBs average 16th.
To be clear, I'm not saying "Don't draft elite QBs." But the numbers are fascinating, in that they reinforce a belief I and many of you have--that streaming a couple of high-upside QB2s could yield as many or more points as starting an elite QB every week. This conclusion is compounded by the even more shocking discovery that the top 5 QBs, on average, score the 11th most fantasy points. The same is true for QBs 12-16. Yes, injuries play a role in leveling these numbers. But injuries are hardly the whole story. And while we're talking about only four seasons, the numbers are interesting enough to merit more research in the coming seasons.
Next up are four charts called "Expected Performers and Surprise Performers," separated by the top 50 QB scorers, the top 25, the top 10, and the top 5. The X axis signifies the top 50 fantasy scorers. Starting with the top 50 chart, the green line above the "1" on the left reflects the fact that each season's #1 fantasy QB owned a top-50 preseason ADP. Of course, right? Because it would be highly unusual for a back-up or third-string preseason QB to finish anywhere close to #1. The first small blue line on this chart appears above the "14," which reflects Case Keenum's astonishingly strong 2017 campaign, when he finished 14th despite owning a preseason QB-55 ADP.
A few things jump out at me. In the top 50 chart, only two non-top-50 preseason QBs have earned streaming value in the past four years. In other words, surprises like Keenum are highly unusual, though in two-QB leagues many non-top-50 options have offered streaming potential--some hope for two-QB-leaguers who miss out on an all-important #3 QB. The top-25 chart shows how often a QB with an ADP above 25 ends up as a top-25 fantasy scorer: 19% of the time, to be exact. And based on the top-10 chart, a full 50% of the time a QB with an ADP above 10 finished in the top 10. Some of those QBs had ADPs in the 11-14 range, so not that big a deal. But many were huge ADP shocks, like Alex Smith in 2017. And of course the top-5 chart looks brutal, and strengthens the notion that selecting two high-upside, back-end QB1s or high-QB2s in the middle rounds could be a better play than burning an early pick on a supposedly elite QB1.
This data isn't intended to be actionable in isolation. But it's thought-provoking, at least to me. "Sure-thing" QBs don't always produce, and middle-third QBs are not always merely streamers. And in fact, sometimes a good chunk of the latter group performs better than a good chunk of the former group.
This data should motivate us to do more research and take more fact-based risks. Preseason ADPs are based on educated guesses, and often are based on the previous season's numbers. That's no way to devise a draft strategy.
Tomorrow I'll offer some RB analysis . . .
---
Sign up for Pre-Draft Top 300 Rankings: http://www.fantasyfootballforwinners.com/
Follow on Twitter: https://twitter.com/_FF4W
