When should quarterbacks retire? The answer is...it depends based on roster structure and draft needs.
Alright, thanks for your time everybody!
Oh wait, that's not good media. I should actually do some analysis...
So I wanted to do was essentially look at how quarterbacks have performed once reaching regression and figure out at what point they are no longer viable options to start. What I did was look at the careers of these 24 quarterbacks which is, I think, every quarterback in league history with either 2000 career attempts or 5 seasons as a starter (sorry if I missed anyone):
Mat Akselsen
Ryan Applehort
Franklin Armstrong
Cooter Bigsby
Avon Blocksdale
Mike Boss
King Bronko
Joilet Christ
Easton Cole
Andrea Falconi
Kevin Fitzpatrick
Stan Francisco
Childish Gambino
Corvo Havran
Rose Jenkins
Borkus Maximus III
Micycle McCormick
Wolfie McDummy
Logan Noble
Chris Orosz
Brad Pennington
Andrew Reese
Gus T.T. Showbiz
Dan Wright
I then plotted their performance in each season of their career against their..well I don't know what we call it, 'player age'?
For the standard of 'performance' here, I used Adjusted Yards per Attempt, which is (Yards + 20 * Touchdowns - 45 * Interceptions) / Attempts. Basically, it's like QB Rating, but probably better, and definitely easier to calculate. Also, I wanted to account for fluctuations in league trends over time, so rather than using raw numbers, I adjusted for league average. Basically it's (Player AY/A / League AY/A) * 100 (the same idea as OPS+, if you're familiar with the baseball stat). So if the result is an 'AY/A+' of 110, the player was 10% better than that season's league average.
Anyway, y'all just want to see some graphs?
![[Image: 120a15e695ed139e5a6df74640c55ccb.png]](https://i.paste.pics/120a15e695ed139e5a6df74640c55ccb.png)
Nothing too shocking here. Quarterbacks are not very good in their first 2 seasons, which is why many stay in the DSFL early in their career. Then at the end of careers we see a steep drop off due to regression. The somewhat interesting information here is slight decline from season 3 to season 6, which doesn't make sense, but can probably be attributed to small sample size. The other interesting part is that season 8 has been the strongest season of quarterback careers, which stands out because regression starts after season 7...
That one above was the average AY/A+, below includes all individual data points, with a trendline that shows more along the lines of what I was expecting.
![[Image: ac33427e1ed2df019bd7a53af44a68cd.png]](https://i.paste.pics/ac33427e1ed2df019bd7a53af44a68cd.png)
Obviously there are some outliers there, but between the two graphs, I think the conclusion is not too hard to see here. Quarterbacks struggle in their first 2 seasons, and start to fall below average at season 9. Season 10 is really the cut off for viablity, as a true rookie QB would generally perform just as well.
Alright, thanks for your time everybody!
Oh wait, that's not good media. I should actually do some analysis...
So I wanted to do was essentially look at how quarterbacks have performed once reaching regression and figure out at what point they are no longer viable options to start. What I did was look at the careers of these 24 quarterbacks which is, I think, every quarterback in league history with either 2000 career attempts or 5 seasons as a starter (sorry if I missed anyone):
Mat Akselsen
Ryan Applehort
Franklin Armstrong
Cooter Bigsby
Avon Blocksdale
Mike Boss
King Bronko
Joilet Christ
Easton Cole
Andrea Falconi
Kevin Fitzpatrick
Stan Francisco
Childish Gambino
Corvo Havran
Rose Jenkins
Borkus Maximus III
Micycle McCormick
Wolfie McDummy
Logan Noble
Chris Orosz
Brad Pennington
Andrew Reese
Gus T.T. Showbiz
Dan Wright
I then plotted their performance in each season of their career against their..well I don't know what we call it, 'player age'?
For the standard of 'performance' here, I used Adjusted Yards per Attempt, which is (Yards + 20 * Touchdowns - 45 * Interceptions) / Attempts. Basically, it's like QB Rating, but probably better, and definitely easier to calculate. Also, I wanted to account for fluctuations in league trends over time, so rather than using raw numbers, I adjusted for league average. Basically it's (Player AY/A / League AY/A) * 100 (the same idea as OPS+, if you're familiar with the baseball stat). So if the result is an 'AY/A+' of 110, the player was 10% better than that season's league average.
Anyway, y'all just want to see some graphs?
![[Image: 120a15e695ed139e5a6df74640c55ccb.png]](https://i.paste.pics/120a15e695ed139e5a6df74640c55ccb.png)
Nothing too shocking here. Quarterbacks are not very good in their first 2 seasons, which is why many stay in the DSFL early in their career. Then at the end of careers we see a steep drop off due to regression. The somewhat interesting information here is slight decline from season 3 to season 6, which doesn't make sense, but can probably be attributed to small sample size. The other interesting part is that season 8 has been the strongest season of quarterback careers, which stands out because regression starts after season 7...
That one above was the average AY/A+, below includes all individual data points, with a trendline that shows more along the lines of what I was expecting.
![[Image: ac33427e1ed2df019bd7a53af44a68cd.png]](https://i.paste.pics/ac33427e1ed2df019bd7a53af44a68cd.png)
Obviously there are some outliers there, but between the two graphs, I think the conclusion is not too hard to see here. Quarterbacks struggle in their first 2 seasons, and start to fall below average at season 9. Season 10 is really the cut off for viablity, as a true rookie QB would generally perform just as well.
![[Image: image0.png?width=822&height=426]](https://media.discordapp.net/attachments/830854638353645609/830854703713746954/image0.png?width=822&height=426)
![[Image: purchase_sig.png]](https://cdn.discordapp.com/attachments/319945672310784010/629840905351725069/purchase_sig.png)