Hi all, I did an analysis of how QBs' play changes from their rookie to 2nd year and here it is. I'll give some info on my data and then the results.
Data was collected for every QB who was a rookie in 2000 or later (arbitrary year, but I didn't want to go back too far since the game is changing fairly rapidly). I only looked at QBs that played at least 8 games (also arbitrary, but I wanted a floor on games played) in both their rookie and 2nd years, played for the same team both years and had at least 14 pass attempts per game played (14 attempts per game is the requirement for someone to be a statistical category leader according to ESPN, not sure if this is an NFL rule or not). The data sample wound up consisting of 20 players.
Player data was scaled to 2011 totals. To do this I totaled each stat category (Pass yards, pass tds, rush yds, rush tds) for each year among all players (not just those that fell into the above category). I then divided each players' individual stat by that year's total and multiplied by the 2011 total. This is an attempt to compare 'apples to apples' by taking into account the ever increasing amount of passing happening in the nfl. Also it makes sense from a fantasy perspective because you don't really care how many yards your QB gets, you care how many he gets relative to the other QBs.
Once I had the scaled data for the qualified 2nd year QBs I looked at rookie year versus 2nd year passing yards, passing tds, rushing yards and rushing tds. The average change, standard deviation, and average change after eliminating outliers (more the 2 standard devs away from mean) are:
Stat Average StDev AvgNoOutlier
PassYds +38 37 +33
PassTds +0.39 0.40 +0.43
RushYds +0.17 5.5 -0.51
RushTds +0.020 0.13 0.0035
There's a year-to-year increase of about 35 passing yards per game, which is 560 yards over a full 16 games. Out of all 20 QBs, only 2 saw their YPG go down, Matt Ryan and Sam Bradford, but they also had the two highest (scaled) rookie YPGs out of the bunch. The average td per game increase was .4, or 6.4 tds over 16 games. Once again, only two QBs saw their numbers decrease, Sam Bradford and Vince Young; Young is right in the middle for rookie tds, Bradford is 4th highest. The rushing stats don't move ... the average changes are close to zero and totally swamped by the standard deviations.
4 players will fall into this category of QBs this year barring injury: Christian Ponder, Blaine Gabbert, Andy Dalton and Cam Newton. Their numbers from 2011 (per game multiplied by 16) were
Player PassYds PassTds RushYds RushTds
Ponder 2700 19 319 0
Gabbert 2360 13 1.5 0
Dalton 3400 20 150 1
Newton 4050 21 700 14
Notes on Gabbert and Ponder: Ponder played in 11 games, but in the first game he played he only attempted 17 passes and in the last 2 he only attempted 23 total (I assume they shut him down at the end of the year since Min was terrible and not going anywhere but I don't know for sure). Gabbert played in 15 games but only had 6 attempts in his first game and 19 in his last game. If you count those 'partial' games as half games for the scaling, Ponder's and Gabbert's numbers become
Player PassYds PassTds RushYds RushTds
Ponder 3100 22 370 0
Gabbert 2530 13.7 110 0
Using the 'half-games' for Ponder and Gabbert, and assuming all four players play 16 games and improve at the average rate for rookies entering their second year, their 2012 projected stats look like
Player PassYds PassTds RushYds RushTds
Ponder 3660 28.4 370 0
Gabbert 3090 20.1 110 0
Dalton 3960 26.4 150 1
Newton 4610 27.4 700 14
I fully expect Cam's rushing tds to come down a lot. Projections generally have Cam and Dalton regressing in passing yards (somewhat understandable but not totally backed up by the numbers), but the thing I find surprising is that many projections have Dalton and Cam staying at the same number of passing TDs or regressing a little bit. This I don't see supported in the data at all. I don't think its at all unlikely that Newton could have 4000/28 and 700/8, and Dalton 4000/26 and 150/1. The other interesting part of this is Ponder. A lot or projections have him at 3300/19 or thereabouts, which I think is low. He saw limited time last year, and projections are always dangerous, but I don't think 3600/28 is out of reach for him. Tack on ~400 rushing yards and a couple tds and the guy could be fantasy relevant.
Thoughts on my methodology / conclusions? Where do you guys have these four sophomore QBs?
Data was collected for every QB who was a rookie in 2000 or later (arbitrary year, but I didn't want to go back too far since the game is changing fairly rapidly). I only looked at QBs that played at least 8 games (also arbitrary, but I wanted a floor on games played) in both their rookie and 2nd years, played for the same team both years and had at least 14 pass attempts per game played (14 attempts per game is the requirement for someone to be a statistical category leader according to ESPN, not sure if this is an NFL rule or not). The data sample wound up consisting of 20 players.
Player data was scaled to 2011 totals. To do this I totaled each stat category (Pass yards, pass tds, rush yds, rush tds) for each year among all players (not just those that fell into the above category). I then divided each players' individual stat by that year's total and multiplied by the 2011 total. This is an attempt to compare 'apples to apples' by taking into account the ever increasing amount of passing happening in the nfl. Also it makes sense from a fantasy perspective because you don't really care how many yards your QB gets, you care how many he gets relative to the other QBs.
Once I had the scaled data for the qualified 2nd year QBs I looked at rookie year versus 2nd year passing yards, passing tds, rushing yards and rushing tds. The average change, standard deviation, and average change after eliminating outliers (more the 2 standard devs away from mean) are:
Stat Average StDev AvgNoOutlier
PassYds +38 37 +33
PassTds +0.39 0.40 +0.43
RushYds +0.17 5.5 -0.51
RushTds +0.020 0.13 0.0035
There's a year-to-year increase of about 35 passing yards per game, which is 560 yards over a full 16 games. Out of all 20 QBs, only 2 saw their YPG go down, Matt Ryan and Sam Bradford, but they also had the two highest (scaled) rookie YPGs out of the bunch. The average td per game increase was .4, or 6.4 tds over 16 games. Once again, only two QBs saw their numbers decrease, Sam Bradford and Vince Young; Young is right in the middle for rookie tds, Bradford is 4th highest. The rushing stats don't move ... the average changes are close to zero and totally swamped by the standard deviations.
4 players will fall into this category of QBs this year barring injury: Christian Ponder, Blaine Gabbert, Andy Dalton and Cam Newton. Their numbers from 2011 (per game multiplied by 16) were
Player PassYds PassTds RushYds RushTds
Ponder 2700 19 319 0
Gabbert 2360 13 1.5 0
Dalton 3400 20 150 1
Newton 4050 21 700 14
Notes on Gabbert and Ponder: Ponder played in 11 games, but in the first game he played he only attempted 17 passes and in the last 2 he only attempted 23 total (I assume they shut him down at the end of the year since Min was terrible and not going anywhere but I don't know for sure). Gabbert played in 15 games but only had 6 attempts in his first game and 19 in his last game. If you count those 'partial' games as half games for the scaling, Ponder's and Gabbert's numbers become
Player PassYds PassTds RushYds RushTds
Ponder 3100 22 370 0
Gabbert 2530 13.7 110 0
Using the 'half-games' for Ponder and Gabbert, and assuming all four players play 16 games and improve at the average rate for rookies entering their second year, their 2012 projected stats look like
Player PassYds PassTds RushYds RushTds
Ponder 3660 28.4 370 0
Gabbert 3090 20.1 110 0
Dalton 3960 26.4 150 1
Newton 4610 27.4 700 14
I fully expect Cam's rushing tds to come down a lot. Projections generally have Cam and Dalton regressing in passing yards (somewhat understandable but not totally backed up by the numbers), but the thing I find surprising is that many projections have Dalton and Cam staying at the same number of passing TDs or regressing a little bit. This I don't see supported in the data at all. I don't think its at all unlikely that Newton could have 4000/28 and 700/8, and Dalton 4000/26 and 150/1. The other interesting part of this is Ponder. A lot or projections have him at 3300/19 or thereabouts, which I think is low. He saw limited time last year, and projections are always dangerous, but I don't think 3600/28 is out of reach for him. Tack on ~400 rushing yards and a couple tds and the guy could be fantasy relevant.
Thoughts on my methodology / conclusions? Where do you guys have these four sophomore QBs?
