The Stats Room: Zman's QB Projections

The Stats Room: Zman's QB Projections

This article is part of our The Stats Room series.

Creating these simple quarterback projections (The Zman's) ended up being a little harder than expected, but they're done. They are simple, non-context projections on how a non-rookie quarterback is expected to perform. While other factors will contribute to their production, these projections provide a great baseline for a quarterback's ability.

For the projections, I wanted to create a projection system similar to Marcels in which Tom Tango created for baseball players.

As Tom describes it, "... uses as little intelligence as possible. So, that's the allusion to the monkey."

I went with the same approach with just a tad more detail in the weights and regression. The basic procedure was to take a three-year weighted sample, regress the data as needed and then add an aging factor. I did not project playing time in which I'll examine in a future article.

Overall, the final formulas came out as expected, heavy regressed. This aspect is always the hardest for owners to digest. With just 16 regular-season games, a big game or two can really boost a player's value and they will, on average, see their production drop the next season after a breakout season. Finally, for these projections, I used data from 1990 to 2016 from ArmChair Analysis.

Additionally, I found a few attributes which had no year-to-year correlation, like fumbles. For these stats, I just used the 2016 league average value. For all the regressions amounts, I regressed to just the 2016 values. I may change this value to

Creating these simple quarterback projections (The Zman's) ended up being a little harder than expected, but they're done. They are simple, non-context projections on how a non-rookie quarterback is expected to perform. While other factors will contribute to their production, these projections provide a great baseline for a quarterback's ability.

For the projections, I wanted to create a projection system similar to Marcels in which Tom Tango created for baseball players.

As Tom describes it, "... uses as little intelligence as possible. So, that's the allusion to the monkey."

I went with the same approach with just a tad more detail in the weights and regression. The basic procedure was to take a three-year weighted sample, regress the data as needed and then add an aging factor. I did not project playing time in which I'll examine in a future article.

Overall, the final formulas came out as expected, heavy regressed. This aspect is always the hardest for owners to digest. With just 16 regular-season games, a big game or two can really boost a player's value and they will, on average, see their production drop the next season after a breakout season. Finally, for these projections, I used data from 1990 to 2016 from ArmChair Analysis.

Additionally, I found a few attributes which had no year-to-year correlation, like fumbles. For these stats, I just used the 2016 league average value. For all the regressions amounts, I regressed to just the 2016 values. I may change this value to some yearly weighted average at future date.

Besides the regression amounts, I see my projections getting tweaked in other ways. I will like to see if the projections change by using the most recent seasons. Also, how do the quarterbacks expected supporting cast affect the projection?

Here is a quick walk through of the quarterback stats I projected:

Completion percentage: Three-year weighted average with the last season being weighted heavily. One of the least regressed values.

Yards per completion: One-year weighted average with lots of regression.

Touchdowns per completion: Three-year weighted average with last season getting half the weight. Lots and lots of regression. I expect these stats to get the most ire.

Interceptions per pass attempt: Three-year evenly weighted average and with lots of regression.

Fumbles per pass attempt: League average used. Quarterbacks average two fumbles lost per team, so impossible to project with such a small sample.

Rushing attempt per pass attempt: Standard three-year weighted average with the least regression of any stat. I'm not sure I like using a ratio (attempt/attempt) but I am going to use it for now.

Rushing yards per rush attempt: Three-year evenly weighted average with some regression.

Touchdowns per rush attempt: League average used. There was no year-to-year correlation with so few rushing touchdowns.

And now for the projections most people jumped ahead to. If anyone threw a single pass last season, they are included in the list for reference. No rookies are included and I will look at all rookie projections at a later date.

PLAYERAGECOMP%YARDS/COMPTD/COMPINT/PAFUM/PARA/PARY/RATD/RA
A.Dalton 30 65.8% 11.6 0.071 0.019 0 10.8% 3 0.043
A.Luck 28 63.4% 11.9 0.077 0.022 0 10.7% 4.8 0.043
A.Rodgers 34 65.4% 11.3 0.079 0.015 0 10.0% 5.4 0.043
A.Smith 33 66.6% 11.2 0.070 0.016 0 12.0% 4.6 0.043
B.Bortles 25 60.1% 11.1 0.072 0.024 0 9.5% 6 0.043
B.Gabbert 28 62.7% 11.5 0.073 0.021 0 12.6% 4.6 0.043
B.Hoyer 32 63.5% 11.5 0.072 0.019 0 5.9% 2.2 0.043
B.Hundley 24 61.8% 11.7 0.074 0.022 0 8.7% 3.7 0.043
B.Osweiler 27 61.6% 10.9 0.071 0.021 0 8.1% 3.3 0.043
B.Petty 26 61.9% 11.6 0.073 0.022 0 7.1% 3.8 0.043
B.Roethlisberger 35 66.1% 11.7 0.074 0.022 0 5.1% 1.8 0.043
C.Cook 24 60.7% 11.6 0.074 0.023 0 7.0% 3.8 0.043
C.Daniel 31 64.6% 11.8 0.074 0.019 0 9.0% 3.2 0.043
C.Henne 32 64.0% 11.8 0.074 0.019 0 7.9% 3.9 0.043
C.Jones 25 63.1% 11.8 0.074 0.021 0 8.0% 3.6 0.043
C.Kaepernick 30 62.3% 11.6 0.073 0.017 0 16.7% 5.8 0.043
C.Keenum 29 63.2% 11.5 0.071 0.020 0 8.1% 2.7 0.043
C.Kessler 24 63.8% 11.5 0.073 0.020 0 7.4% 3.1 0.043
C.Newton 28 59.0% 12.2 0.075 0.021 0 19.5% 4.4 0.043
C.Palmer 38 62.4% 11.6 0.074 0.022 0 4.8% 2.4 0.043
C.Wentz 25 62.9% 10.8 0.070 0.021 0 8.1% 3.3 0.043
C.Whitehurst 35 62.7% 11.8 0.073 0.020 0 8.5% 3.7 0.043
D.Anderson 34 65.2% 11.8 0.073 0.021 0 9.0% 2.7 0.043
D.Brees 38 68.1% 11.3 0.074 0.023 0 4.8% 1.8 0.043
D.Carr 26 63.0% 11.4 0.074 0.018 0 6.7% 3 0.043
D.Fales 27 63.7% 11.8 0.074 0.020 0 8.1% 3.7 0.043
D.Prescott 24 65.5% 11.7 0.074 0.018 0 10.8% 4.4 0.043
D.Stanton 33 60.6% 11.7 0.073 0.021 0 10.1% 2 0.043
E.Manning 36 63.8% 11.2 0.074 0.022 0 4.3% 2.2 0.043
E.Manuel 27 62.3% 11.8 0.073 0.020 0 11.2% 3.5 0.043
G.Smith 27 63.2% 11.8 0.073 0.022 0 10.5% 4.2 0.043
J.Brissett 25 63.2% 11.8 0.073 0.020 0 10.5% 4.3 0.043
J.Cutler 34 64.6% 12 0.073 0.023 0 7.5% 4.5 0.043
J.Flacco 32 65.5% 10.7 0.070 0.020 0 5.3% 2.3 0.043
J.Garoppolo 26 64.3% 11.7 0.074 0.019 0 10.8% 2.1 0.043
J.Goff 23 59.5% 11.3 0.073 0.023 0 6.9% 3.3 0.043
J.McCown 38 60.6% 11.8 0.073 0.025 0 7.2% 4.3 0.043
J.Webb 31 64.5% 11.8 0.074 0.019 0 8.2% 3.6 0.043
J.Winston 23 60.3% 11.7 0.074 0.025 0 9.6% 3.5 0.043
K.Clemens 34 64.2% 11.8 0.074 0.020 0 8.5% 3.3 0.043
K.Cousins 29 67.5% 11.9 0.072 0.019 0 6.1% 2.7 0.043
K.Hogan 25 62.7% 11.6 0.073 0.021 0 9.3% 6 0.043
L.Jones 28 63.6% 11.6 0.074 0.021 0 8.0% 2.3 0.043
M.Barkley 27 62.7% 11.9 0.073 0.024 0 6.9% 2.8 0.043
M.Cassel 35 62.7% 11.6 0.073 0.023 0 8.2% 3.6 0.043
M.Glennon 28 64.2% 11.7 0.074 0.020 0 7.2% 4 0.043
M.Mariota 24 61.9% 12 0.075 0.021 0 10.7% 5.9 0.043
M.McGloin 28 64.2% 11.7 0.074 0.020 0 8.0% 3.2 0.043
M.Moore 33 65.8% 11.8 0.075 0.020 0 7.4% 2.9 0.043
M.Ryan 32 68.3% 12.4 0.075 0.018 0 6.6% 3.1 0.043
M.Sanchez 31 64.6% 11.7 0.073 0.023 0 9.0% 2.9 0.043
M.Schaub 36 63.7% 11.8 0.073 0.024 0 7.9% 3.3 0.043
M.Stafford 29 65.8% 11.4 0.072 0.018 0 7.4% 3.7 0.043
N.Foles 28 62.4% 11.7 0.072 0.021 0 6.6% 2.7 0.043
P.Lynch 23 61.7% 11.6 0.073 0.022 0 9.0% 3.3 0.043
P.Rivers 36 63.9% 12.1 0.075 0.025 0 4.6% 2.7 0.043
R.Fitzpatrick 35 61.1% 11.7 0.074 0.025 0 10.3% 3.9 0.043
R.Griffin 27 63.6% 11.5 0.071 0.020 0 12.7% 4.8 0.043
R.Mallett 29 62.3% 11.7 0.073 0.020 0 6.8% 2.5 0.043
R.Tannehill 29 65.9% 11.6 0.073 0.019 0 8.2% 4.5 0.043
R.Wilson 29 65.9% 11.8 0.073 0.017 0 16.7% 5.2 0.043
S.Bradford 30 69.0% 10.7 0.070 0.017 0 5.5% 2.6 0.043
S.Hill 37 62.7% 11.8 0.073 0.023 0 8.1% 2.4 0.043
S.Mannion 25 63.3% 11.7 0.074 0.021 0 8.0% 3.6 0.043
S.Tolzien 30 64.4% 11.6 0.073 0.020 0 9.3% 2.8 0.043
T.Boykin 24 63.2% 11.7 0.074 0.022 0 9.5% 2.9 0.043
T.Brady 40 64.3% 11.9 0.077 0.020 0 6.7% 2 0.043
T.Romo 37 65.7% 11.7 0.076 0.024 0 6.7% 3 0.043
T.Savage 27 63.8% 11.6 0.072 0.019 0 8.6% 2.8 0.043
T.Siemian 25 61.3% 11.7 0.072 0.020 0 7.2% 2.7 0.043
T.Taylor 28 63.8% 11.4 0.073 0.017 0 19.6% 5.3 0.043

To collect the expected stats for a week or season, an owner just needs to get an estimated number of pass attempts. I will go ahead and create some full-season estimates by using the average number of passes thrown per team in 2016 (572) for each of the quarterbacks. Additionally, I will rank them by fantasy value using the following values:

Passing yard: 0.04 points

Passing touchdown: 4 points

Rushing yard: 0.1 points

Rushing touchdown: 6 points

Turnover: -2 points

PLAYERATTEMPTSPASS YDSPASS TDINTFUMRUSH YDSRUSH TDFANTASY POINTS
R.Wilson 571 4432 27.5 9.6 3.6 499 4.1 336
T.Taylor 571 4170 26.6 9.5 3.6 592 4.8 335
C.Kaepernick 571 4129 25.9 9.9 3.6 554 4.1 322
C.Newton 571 4111 25.3 11.8 3.6 491 4.8 313
A.Rodgers 571 4233 29.3 8.5 3.6 309 2.5 308
M.Ryan 571 4840 29.1 10.5 3.6 117 1.6 303
D.Prescott 571 4377 27.5 10.4 3.6 272 2.7 300
A.Smith 571 4245 26.5 9.3 3.6 315 3.0 299
A.Luck 571 4318 27.9 12.5 3.6 293 2.6 297
M.Mariota 571 4233 26.7 12.2 3.6 358 2.6 296
R.Griffin 571 4165 25.8 11.4 3.6 347 3.1 293
B.Gabbert 571 4098 26.0 11.8 3.6 327 3.1 288
R.Tannehill 571 4362 27.3 11.0 3.6 209 2.0 288
A.Dalton 571 4360 26.8 10.7 3.6 185 2.7 287
J.Brissett 571 4241 26.3 11.3 3.6 256 2.6 286
K.Hogan 571 4161 26.3 12.2 3.6 321 2.3 286
G.Smith 571 4270 26.4 12.5 3.6 251 2.6 285
K.Cousins 571 4576 27.9 11.1 3.6 94 1.5 284
C.Daniel 571 4343 27.2 10.7 3.6 164 2.2 283
J.Webb 571 4335 27.2 11.0 3.6 170 2.0 282
M.Moore 571 4427 28.0 11.6 3.6 123 1.8 282
J.Cutler 571 4412 26.8 12.9 3.6 193 1.8 281
J.Garoppolo 571 4311 27.1 10.8 3.6 129 2.7 281
D.Anderson 571 4399 27.3 12.1 3.6 136 2.2 281
E.Manuel 571 4188 26.0 11.4 3.6 221 2.8 280
C.Henne 571 4299 26.9 10.9 3.6 174 2.0 280
K.Clemens 571 4315 27.1 11.5 3.6 161 2.1 279
M.Stafford 571 4283 27.0 10.0 3.6 156 1.8 279
S.Tolzien 571 4281 27.0 11.3 3.6 147 2.3 278
D.Fales 571 4275 26.8 11.2 3.6 170 2.0 278
T.Romo (retired) 571 4410 28.4 13.5 3.6 113 1.6 277
M.Glennon 571 4273 27.1 11.3 3.6 166 1.8 277
M.Sanchez 571 4318 27.0 12.9 3.6 148 2.2 276
T.Brady 571 4376 28.1 11.4 3.6 78 1.6 275
M.McGloin 571 4271 26.9 11.5 3.6 144 2.0 275
C.Whitehurst 571 4228 26.3 11.5 3.6 178 2.1 274
C.Jones 571 4260 26.5 12.0 3.6 166 2.0 274
T.Boykin 571 4233 26.6 12.3 3.6 155 2.3 273
S.Mannion 571 4239 26.6 12.0 3.6 166 2.0 273
T.Savage 571 4220 26.4 10.8 3.6 139 2.1 272
M.Schaub 571 4286 26.7 13.5 3.6 148 1.9 271
D.Brees 571 4400 28.7 12.9 3.6 50 1.2 270
R.Fitzpatrick 571 4095 25.7 14.0 3.6 229 2.5 269
B.Roethlisberger 571 4416 27.9 12.4 3.6 52 1.3 269
B.Hundley 571 4145 26.0 12.4 3.6 183 2.1 269
S.Bradford 571 4231 27.5 9.7 3.6 81 1.4 269
C.Kessler 571 4192 26.4 11.4 3.6 132 1.8 268
L.Jones 571 4215 26.8 12.2 3.6 107 2.0 267
M.Cassel 571 4154 26.1 13.4 3.6 172 2.0 266
P.Lynch 571 4080 25.8 12.3 3.6 168 2.2 265
P.Rivers 571 4407 27.4 14.0 3.6 70 1.1 265
D.Carr 571 4087 26.6 10.0 3.6 115 1.6 264
C.Keenum 571 4166 25.8 11.5 3.6 123 2.0 264
B.Bortles 571 3825 24.7 13.7 3.6 324 2.3 264
S.Hill 571 4228 26.1 12.9 3.6 112 2.0 263
M.Barkley 571 4265 26.2 14.0 3.6 109 1.7 261
J.Winston 571 4040 25.5 14.2 3.6 192 2.4 261
B.Petty 571 4092 25.8 12.7 3.6 156 1.8 260
R.Mallett 571 4174 25.8 11.4 3.6 98 1.7 260
N.Foles 571 4180 25.8 11.9 3.6 102 1.6 260
D.Stanton 571 4045 25.4 12.1 3.6 117 2.5 258
B.Hoyer 571 4168 26.0 11.1 3.6 74 1.4 257
T.Siemian 571 4100 25.3 11.5 3.6 112 1.8 257
J.McCown 571 4094 25.2 14.6 3.6 177 1.8 257
C.Cook 571 4024 25.5 13.0 3.6 153 1.7 255
C.Wentz 571 3895 25.1 11.8 3.6 155 2.0 253
C.Palmer 571 4137 26.4 12.8 3.6 66 1.2 252
J.Flacco 571 4013 26.0 11.4 3.6 71 1.3 249
E.Manning 571 4067 26.9 12.7 3.6 54 1.0 249
B.Osweiler 571 3833 25.0 12.0 3.6 152 2.0 249
J.Goff 571 3836 24.6 13.3 3.6 131 1.7 241

The rushing yards really push up the top four, but none will see the 571 attempts, especially Kaepernick. The name that really sticks out is Rodgers coming in at No. 5. The industry has him as a nearly unanimous pick as the top quarterback. Here's where the consensus and my projections disagree.

Rodgers' 2015 60.7 percent completion rate has my projected rate lower than might be expected. Only 571 pass attempts are assumed after throwing 610 last season. If he hits 610 again, his touchdowns increase by a couple.

Touchdown pass regression is pretty strong. From my data set, nine quarterbacks threw 38 to 42 touchdown passes in a season. The next season they averaged 22.6 touchdowns (22 median) with an average drop of 16.6 (median 20). While some quarterbacks got hurt leading to the drop, none improved and the smallest drop was Tom Brady in 2011 to 2012 going from 39 to 35. A 35 touchdown total is probably a little ambitious. In one magazine, I saw his total at 41. No sane person can predict a jump.

Finally, I went ahead and ran projections for previous seasons so people can compare the projections.

That's it for now. For my next article, I will either run wide receiver projections or try to come up with a projected pass attempt number (which will adjust once the season starts). For now, look at the data and let me know where my calculations seem off. It will be interesting to see how they hold up compared to other projections this season.

Want to Read More?
Subscribe to RotoWire to see the full article.

We reserve some of our best content for our paid subscribers. Plus, if you choose to subscribe you can discuss this article with the author and the rest of the RotoWire community.

Get Instant Access To This Article Get Access To This Article
RotoWire Community
Join Our Subscriber-Only NFL Chat
Chat with our writers and other RotoWire NFL fans for all the pre-game info and in-game banter.
Join The Discussion
ABOUT THE AUTHOR
Jeff Zimmerman
Jeff is a former RotoWire contributor. He wrote analytics-focused baseball and football articles for RotoWire. He is a three-time FSWA award winner, including the Football Writer of the Year and Best Football Print Article awards in 2016. The 2017 Tout Wars Mixed Auction champion and 2016 Tout Wars Head-to-Head champ, Zimmerman also contributes to FanGraphs.com, BaseballHQ and Baseball America.
NFL Staff Picks: Week 11 Winners
NFL Staff Picks: Week 11 Winners
Guillotine League Strategy: Late Season Strength of Schedule Downgrades
Guillotine League Strategy: Late Season Strength of Schedule Downgrades
Survivor: Week 11 Strategy & Picks
Survivor: Week 11 Strategy & Picks
NFL Waiver Wire: Week 11 Deep Dive
NFL Waiver Wire: Week 11 Deep Dive