Print

QB Stats and Future Performance: Part II

You are viewing free content provided by FantasyGuru.com. Why not consider subscribing today?

by Mike Horn, Staff Writer

Published, 4/4/14

 

Last week I looked at whether or not QB Completion Percentage (Comp%) and Yards Per Attempt (YPA), when taken together, help predict future performance. I concluded that they were a good indicator of the chances that a Top 10 fantasy QB will repeat his Top 10 performance. But also:

 

Two key things this analysis ignores: the running component of fantasy scoring and age. Luck got 20% of his fantasy scoring from rushing yards and TDs last year. Newton got 29% and Foles 15%; the rest were under 10%. One of the things I need to do is examine if the percentages are different for running QBs vs. pocket passers.

 

Also, this ignores the effects of aging: possibly positive for young QB (Foles, Luck, and Newton all were only 24 last year, Stafford was 25) and possibly negative for older players (Peyton was 37, Brees was 34).

 

So I’m back to talk about those factors.

 

Background

 

This summarizes the study parameters from the first article; you can skim this if you remember them.

 

I’ll use the period from 1988-2012 to examine how those stats help predict the next-year performance (that brings us up to 2013) of Top-10 fantasy QBs. I start at 1988 so that I’m only looking at consecutive 16-game seasons and avoid the strike era of 1982-1987.

 

For fantasy scoring, I’m using 25 pass yds = 1 Fantasy Point (FP); 10 rush yards = 1 FP; passing TDs = 4 FP, and rushing TDs = 6 FP. No turnovers are included, and the odd receiving yards/TDs that are accumulated by QBs in this period are scored the same as rushing yards/TDs (no PPR).

 

I limited the QBs to those who had at least 250 passing attempts in one year (I’ll call this the base year or “Y”) and then played the next year (“Y+1”). I’m also going to use fantasy points per game (FPG) rather than total FP to reduce the impact of injuries in Y+1 – while injuries probably affect FPG, they have less impact on that number than on total FP.

 

 

 

The graph shows Comp% on the x-axis and YPA on the y-axis. The gray lines represent the averages for both stats: 59% Comp% on the x-axis and 7.0 YPA on the y-axis. The Top 10 QBs in FPG in the base year are shown as green or red diamonds. The green indicates that they also ranked in the Top 10 in Y+1, while the red means they dropped out. The roman numerals indicate the four quadrants of the graph. Notice that the green diamonds are heavily clustered in the “I quadrant”.

 

Running QBs

 

The average QB in the sample got 10% of his FPG from running the ball. So that’s the basic break point between a “running” QB and a “traditional” pocket quarterback. In order to look to separate the players into more meaningful groups, I arbitrarily group them into 5% increments: 0-5% of their FPG from rushing; 5-10%; 10-15%; and 15+%. If it helps you to put labels on these players, I’d call the categories: immobile, mobile, scramblers, and runners respectively.

 

Over the period of the study, here’s the chart of the Top 10 QBs (FPG) in the base year and how they did in Y+1 with the contribution of their rushing FPG indicated as a percentage of their total FPG:

 

 

 

OK, this is a busy graph. Basically, the bright reds and greens are the pocket passers (less than 10% of their fantasy value from rushing FP). They tend to be clustered in Quadrant I because they had to be good passers to have fantasy value. The paler reds and greens are “Running QBs,” defined here as those who got more than 10% of their FPG from rushing yards/TDs. Since players in Quadrants II, III, and IV are not as good at passing as the Quadrant I QBs, not surprisingly the starting fantasy QBs in those quadrants tended to be more running QB types. But to make sense of the data, a table is probably more useful:

 

 

Overall

<5% FP from Rushing

5-10% FP from Rushing

10-15% FP from Rushing

>15% FP from Rushing

Quadrant

Total Top 10 Base Year

Top 10 Both Years

Total Top 10 Base Year

Top 10 Both Years

Total Top 10 Base Year

Top 10 Both Years

Total Top 10 Base Year

Top 10 Both Years

Total Top 10 Base Year

Top 10 Both Years

I

139

64%

50

64%

39

59%

20

60%

30

73%

II

36

44%

9

33%

8

50%

4

100%

15

33%

III

21

43%

7

29%

5

0%

7

86%

2

50%

IV

34

41%

3

0%

4

50%

5

60%

22

41%

Grand Total

230

56%

69

54%

56

52%

36

69%

69

54%

2013 QBs

 

 

Manning (I)
Rivers (I)
Romo (I)

Brees (I)
Rodgers (I)
Dalton (I)
Stafford (II)

 

Foles (I)
Newton (I)
Luck (III)

 

The two columns under the “Overall” heading summarize the data presented in the first article, by quadrant and then overall. To simplify things, I’m just showing the number of QBs who finished in the Top 10 in the base year for each quadrant, so you can see the sample size, and then the percentage of those players who repeated their Top 10 finish in the next year.

 

The next four pairs of columns break QBs into categories based on the percentage of their FPG that came from running the ball. For example, the “<5% FP from Rushing” means that a very small part of their fantasy production came from running. These are the classic pocket passers – if you look at the bottom row, it lists the Top 10 QBs from 2013 who fit into this category (the Roman numeral after their name tells you what quadrant they belonged to in 2013). Peyton Manning (Quadrant I last year) is the prototypical QB in this category. On the far right (>15% FP from Rushing) are the running QBs.

 

The colors in the chart highlight what I think are the key numbers. First, the overall average of 56% Top 10 QBs who repeat in Y+1 is indicated in yellow. That is the baseline number; what I want to do is find categories that greatly exceed (or fall below) that number. In green I’ve marked the places where the percentage of repeating in the Top 10 is substantially above 56% and the sample size is 30 or more players.

 

Basically, it’s almost all the Quadrant I QBs. We can argue whether the QBs with 5-10% of their scoring from rushing should be highlighted: is 59% substantially better than 56%? Maybe not. Also, the Quad I QBs in the 10-15% category aren’t shown in green because the sample size is only 20. Here I chose to highlight the total for this category – this seems to be a pretty successful group in Y+1 as a whole although none of the sub-categories by quadrant are very big. This might matter more now if there were a 2013 example in this group, but there is not.

 

Mostly, these numbers are positive for the eight QBs in 2013 who finished in the Top 10 and Quadrant I. It’s less so for Matt Stafford and Andrew Luck. Stafford is in a small sub-category (Quad II, 5-10% of his FPG from running the ball), but even his overall running FP category is basically the same as the overall average. Luck is in an almost unique spot – only two other QBs in 25 years had Top 10 fantasy numbers with a high percentage of points from rushing in the Quadrant III passing group. The repeat percentage there is essentially meaningless.

 

What I think is meaningful from a stat perspective is that the more unique a player’s talent is, the harder it is to project future performance with stats because we don’t have enough stats to make a projection with any kind of confidence. What this data does highlight is how much of Luck’s value came from running the ball (20%) and that overall, those QBs who make the Top 10 because they run a lot but also aren’t in the top group (Quad I) for passing tend not to repeat in the Top 10. If Luck repeats, it will probably be because he improves his passing production, which I think we can easily see happening.

 

The bottom line from this: generally, Top-10 QBs who get a lot of FP from running (over 15% of their numbers), do not do better in the next season than the more traditional QBs who rely more on their passing (<10% FP from rushing). QBs who are a hybrid (10-15% FP from running) have the best next-year success rate, at least if they were in the Top 10 to begin with. While this may just be a statistical fluke, it may be something significant to watch in future years.

 

Age

 

For this part of the study, I broke the QBs in the overall sample into 4 categories: age 21-25, 26-30, 31-35, and 35+. This is a completely arbitrary way of categorizing players by age.

 

 

 

Another busy graph. In this case, the bright reds and greens are younger QBs (30 or fewer years old), and the paler reds and greens are veteran QBs (those over 30). I don’t see any strong generalizations, although you might. Let’s go straight to the table:

 

 

Overall

Age 21-25

Age 26-30

Age 31-35

Age 35+

Quadrant

Total Top 10 Base Year

Top 10 Both Years

Total Top 10 Base Year

Top 10 Both Years

Total Top 10 Base Year

Top 10 Both Years

Total Top 10 Base Year

Top 10 Both Years

Total Top 10 Base Year

Top 10 Both Years

I

139

64%

16

75%

66

67%

47

60%

10

50%

II

36

44%

8

63%

20

35%

6

33%

2

100%

III

21

43%

6

83%

8

38%

4

0%

3

33%

IV

34

41%

17

59%

12

25%

3

0%

2

50%

Grand Total

230

56%

47

68%

106

54%

60

50%

17

53%

2013 QBs

 

 

Foles (I)
Newton (I)
Stafford (II)
Luck (III)

Rodgers (I)
Dalton (I)

Rivers (I)
Romo (I)
Brees (I)

Manning (I)

 

This table is structured much like the previous one. Just as in the first table, the two columns under the “Overall” heading summarize the data with the number of QBs who finished in the Top 10 in the base year for each quadrant and then the percentage of those players who repeated their Top 10 finish in the next year. Obviously, these numbers are exactly the same in both tables.

 

Now, the next four pairs of columns break QBs into age categories, and the bottom row again lists the Top 10 QBs from 2013 who fit into these categories.

 

Finally, I’m using the colors to again highlight what I think are the key numbers. The same baseline average of 56% Top 10 QBs who repeat in Y+1 is indicated in yellow. And green marks the places where the percentage of repeating in the Top 10 is substantially above 56%, and the sample size is 30 or more players.

 

The first thing that stands out is that young (21-25) QBs who have already made the Top 10 are very likely to repeat. The individual quadrant sizes are too small to be meaningful, much as I’d like that 83% repeat rate for Quad III young guys to be even more positive for Luck than the overall percentage. Still, the bottom line in this age group is the first stat in this series that strongly favors another Top 10 showing for him. It’s also a good number for Stafford – it’s hard to remember how young he still is. Nick Foles and Cam Newton had already been looking good, but this is another point in their favor.

 

While the next older age group overall performs in line with the total population of QBs I’m looking at, the Quadrant I QBs in this category have also been very successful in Y+1. I don’t think this really changes my opinion of Aaron Rodgers, but it does raise my estimation of Andy Dalton a bit.

 

Notice that about half of all Top 10 fantasy QBs from 1988 to 2012 were Quad I players between 26 and 35. It’s also worth noting that the younger group of those players repeated more often than the older group. Drew Brees, Tony Romo, and Philip Rivers are in a positive age group/quadrant category. But it’s not as strong year-on-year as the QBs aged 21-25 (overall) or those between 26-30 years old in Quad I. So that’s a slight caution.

 

The player who looks like the biggest risk here is Peyton Manning. Old QBs don’t look as reliable, although the sample size itself makes that assessment unreliable. And the number “17” may actually overstate the sample size. Only eight QBs of age 35 managed Top-10 fantasy finishes from 1988 to 2012. It’s a pretty strong group, with mostly Hall-of-Fame-quality players and a couple of late bloomers: Brett Favre, Doug Flutie, John Elway, Kurt Warner, Peyton Manning, Rich Gannon, Steve Young, Warren Moon. Obviously, Manning fits in that group – heck, he actually is in that group, repeating in 2013 after ranking high in 2012. Yet even that group got old and couldn’t keep its fantasy performance in the Top 10 on a regular basis. As I said earlier with Luck, unique players – like one of the best ever at QB in Peyton – are difficult to project with stats because of the difficulty in finding a relevant sample. I’m not sure this sample of older QBs is relevant, but I think of it as capturing the risk inherent in Peyton’s neck. Or, if you prefer, in making the point that a 38-year-old Peyton is a riskier fantasy play than a 27-year-old Peyton.

 

To wrap this up, I want to stress two points. First, be careful in looking at how I’ve carved up the data to not focus too much on the small sample numbers. Second, I wouldn’t put too much weight on any one number. What I want to do next is pull together the data on QBs from the SOS Article as well as the two sets of studies on Comp% and YPA – plus any studies on future QB performance I might get to – to provide a more comprehensive look at what might happen in 2014. I also want to see if Comp% and YPA have any predictive value for QBs who didn’t make the Top 10.

 

You're going to score 1,968 points this week!

Back to the top