NFL draft, college stats, drafting quarterbacks

Killer nuts's picture
January 2, 2014 at 3:16p

This article is from a couple years ago and is regarding different players but it makes some interesting points regarding college careers, analyzing QBs and projecting them as NFL players based on college statistics. It can certainly be applied to the likes of Braxton Miller, Teddy Bridgewater,  Johnny Manziel, etc this year. It's also interesting to have the gift of hindsight looking back at the article now three years later.  Here's a table from the article, it is worth a quick read:


Six Big-Name College Quarterbacks

Player    Comp.  Att.       Pct.     Yards   YPA   TD  INT  Rating*
Player A   851   1,354   62.85   11,201  8.27   90   33   100.93
Player B   825   1,232   66.96   8,772    7.12   76   37   95.60
Player C   841   1,383   60.81   10,286  7.44   84   36   93.15
Player D   493   797     61.86   6,625    8.31   52   21   99.04
Player E   564   986     57.20   7,731    7.84   51   33   85.72
Player F   661   985     67.11   9,286    9.43   88   15   120.72

* Using the NFL formula for passer rating, not the NCAA formula

The article discusses these six players and identifies that Player F jumps off the page as the best of the bunch. It is definitely fascinating to see the reveal of who each of these players are. Each one is very recognizable for their college achievements and their respective pro careers.

Warning: this article definitely involves a significant amount of SEC butt-kissing


For the whole article:

Comments Show All Comments

TBDBITL0509's picture

I took a business stats class at Fisher, and I vividly remember a guest speaker who was very famous for predictive modeling outside of the stock market (plenty of folks work on that exclusively). His firm is top tier in the industry. 
He showed us plenty of models and how good they were, talking about how he got to the right predictive inputs. The results were impressive. Someone in the class asked if he ever failed at getting a good model.
He told us the NFL once hired his firm to build a model that would predict college football player success. His team took the job and spent a ton of time doing research and charting player measurables and stats. To date, the NFL draft is the only industry where he couldn't predict future success.

4thandinches's picture

Stats don't measure moxie!

I wasn't born a Buckeye but I became one as fast as I could. 

Statutoryglory's picture

Stats also don't measure whether the coaching staff properly utilizes the player i.e. Metcalf up the Middle.  

CptBuckeye24's picture

One thing predictive models and statistics cannot and do not account for is: field vision and mentally toughness/aptitude to run an NFL offense. 
You need the vision and command of the playbook to see the field.  Timing is so important and being able to anticipate and put the ball in a spot before it develops.  Processing that information differentiates many QBs.  This is something Brandon Weeden could not handle, even though he has a big arm and some of the prototypical traits. 

4thandinches's picture

CFB and the NFL are two completely different games. It's like comparing apples to oranges. 

I wasn't born a Buckeye but I became one as fast as I could.