Try to think of the last time going into the season that you thought “man, the offensive line is going to be a real beast this year.” 2010? How’d that work out for you? It seems like every year we go into the season hoping we stay healthy, hoping we can piece things together, hoping we can run the ball. Then in those years where we go into the season with high hopes, things fall short. 2005 seems like the last time we did anything worthwhile on the line. One could argue 2007 wound up fine too. Still, that was 5 years ago.
Others seem to echo that we have consistent offensive line problems. UGA fans complain all the time that we under-recruit or can’t develop offensive linemen. So I figured I’d try to see if there is any objective measure of UGA’s ability to recruit and develop talent. I figure we need evaluations of talent by someone other than UGA coaches (Bobo!). While high recruiting rankings may not be a guarantee of college success, others have shown they are very predictive. So using a similar formula, I compared recruiting rankings to another non-UGA evaluator of talent, the NFL draft.
Here’s what I came up with: I simply added up the number of OL recruited and stars each offensive lineman was given by Rivals.com rankings for each year since 2002 until 2008. From that we can calculate a couple of things.
- Total linemen recruited
- Total stars
- Average stars per lineman.
Then I looked at whether those offensive linemen recruits got drafted and assigned a basic valuation depending on where in the draft they went, which I call an “NFL Point. ” A first round pick was given 7 points, a third round pick 5 points, and a seventh round pick 1 point. This gives us several pieces of information:
- Total OL drafted
- Percentage of OL recruits drafted
- Total draft points
- Points per drafted OL
- Points per recruit
Then to find out how a school is doing at developing talent as judged by recruiting rankings compared to NFL points, I divided the average recruiting star ranking of a team by the average NFL point per player recruit. I did this for UGA, Florida, LSU, and Bama. This is called “Stars/AvNFL” on the table. I asked the internets which school has the most OL currently in the NFL, and got Texas. So I included them too.
Here are the results:
|Total OL Recruits||33||24||30||23||23|
|Total NFL Points||24||15||14||12||17|
|Average NFL Point/Player Drafted||4||3.75||4.666667||4||5.666667|
|Average NFL Point per player recruit||0.727273||0.625||0.466667||0.521739||0.73913|
|National Champs 2005-2012||0||2||1||1||2|
(note: I didn’t do Auburn because I hate Auburn. The Bama numbers may not have been the best choice since they were on probation and not at a Saban level during the time in question. He has had only one offensive lineman drafted that he recruited, and that kid was a 2009 recruit (the only 3 year o-lineman I found), so it is not included. The average star ranking for Saban’s OL recruits from 2007-2012 was 3.59. UGA’s during that time has been 3.36. Saban has recruited 22 OL, Richt 28 during that timeframe. It will be interesting to see how the draft results play out. )
From this table, we learn a couple of things:
- UGA recruited a LOT of offensive linemen from 2002- 2008, more than any of these schools
- UGA on average recruited a lower ranked lineman than Florida, LSU, and Texas
- UGA had more offensive linemen get drafted
- UGA had a higher percentage of linemen recruits get drafted than any other school
- UGA’s linemen were middle of the pack in terms of average NFL points per drafted OL
WHAT DOES THE STARS/AV NFL POINT STAT REALLY TELL US?
I have not been quite sure what to make of the Stars/Av NFL points statistic. It would appear that a lower number indicates greater development ability. For instance, if Boise State recruited an average OL with 1 star, and they were drafted at the same rate as a school who recruited 5 star OL, then their Stars/AvNFL would be lower, indicating a superior ability to develop talent as judged by recruiting rankings and NFL scouts. The table indicates UGA was superior by this measure to Florida, LSU, and Texas in developing talent. That seems reasonable given the number of draft picks and NFL points, particularly in light of the fact that our average recruiting ranking was lower than these schools.
The problem with this measure is Bama. Our statistic suggests that Bama, who had 3 draft picks to UGA’s 6, was better at developing talent. Maybe they were. But the way the statistic is calculated seems to give a bias to a school with fewer OL recruits. Say two schools have the same average star ranking of recruits, but one school brings in 30 kids while the other brings in 10 over an 8 year period (talk about hoping to avoid injury!). Both schools start 5 on the line each week. Assuming equal performance on the lines and an equal number of kids drafted during the same period, the school with 10 recruits will obviously have a higher NFLPoint/recruit since each recruit got drafted.
But this brings up a more interesting debate:
Quantity v. Quality
UGA fans cry every year about all the OL we have that get hurt, leave, fail out, or ride around in Mudcat’s car. But this is not all that surprising given the sheer number of kids we bring in. It would appear Richt’s philosophy is to recruit substantial numbers at OL, perhaps at the cost of quality, and see how things play out. Florida has recruited over 25% less linemen than us over time. The question becomes… which is better, quantity or quality? I don’t know the answer to this.
Georgia has recruited more OL than any of the above schools from 2002-2008 (that pace appears to continue to this day). Georgia has produced more NFL OL draft picks than any of these schools from those classes. Georgia outperforms these schools in developing NFL talent compared to recruiting rankings, though the value of this stat is questionable. In short, by these objective criteria one would not think UGA has a problem in developing offensive linemen. But as a UGA fan, it just doesn’t feel right.
Finally, I am not a statistician. I am an attorney who blogs some evenings. I don’t know how to determine statistical variation, or whether anything is statistically significant. Anyone who wants to see my spreadhseet is welcome to see it, just let me know it the comments and I’ll zip it to you somehow. Any suggestions on how to improve this or look at other data is welcome, though I may let you do it yourself.