More Stafford, completion percentage, efficiency, etc.

As I’ve posted, I’m intrigued by the criticism of Matthew Stafford’s game being lacking due to his completion percentage languishing in the mid-50s. Now granted, some of it comes from chatter on rival message boards that can certainly be discounted as little more than another opportunity to take a gratuitous shot at the enemy, but there are plenty of knowledgeable folks out there who think completion percentage is an important factor in evaluation.

My question is why. Completion percentage in and of itself doesn’t tell us much about how much a team scores, which, after all, is the purpose of an offense, nor does it tells us anything about whether a team wins or loses, which, after all, is the point of playing the games. At most, it’s an indicator of how efficient that team’s offense may be – in gaining yardage, in moving the chains, in keeping the other team’s offense off the field – which in turn may shed some light as to the bigger questions. But it’s an incomplete metric, even for that purpose.

What got me thinking about this recently was this post about Oklahoma’s Sam Bradford over at Statistically Speaking. Bradford led the nation last year in passing efficiency, his stats highlighted by a scintillating 69.5 completion percentage. (Stafford, as most of us know, finished last year at 55.7%.) Of most interest to me in Matt’s post, though, was this point:

Last year Bradford’s efficiency helped the Sooners score the 6th most points in Big 12 play despite their 9th place finish in yards.

Now that’s useful. If Oklahoma’s offense is geared towards the short passing game, like many college offenses are these days, then it makes sense that the more efficient its quarterback is throwing the ball, the more likely that should lead to scoring.

But here’s the thing: last season, Georgia was eighth in the SEC in completion percentage and was sixth in passing efficiency. Despite those mediocre showings, the Dawgs still managed to rank higher in the conference in scoring (fifth) than they did in total yardage (sixth), which would indicate some level of efficiency in scoring that perhaps should be attributed to other factors.

And that’s where Phil Steele enters the discussion.

Steele tracks a stat that he claims authorship of: yards per point. It sounds just like what it is, a simple measure of the number of yards a team generates on offense divided by the number of points it scored. As such, it would seem to be a general yardstick to measure scoring efficiency, as in the fewer yards a team needs to gain in order to generate points, the more efficient its offense is performing.

Steele, on page 299 of his 2008 College Football Preview, writes that he went back and calculated the YPP for all teams from 1990 to the present and found that the average YPP over that period is approximately 15.5. Here’s Steele’s list of the ten best YPPs in college football last season.  Note the bolded team, creaky passing game and all, that makes the list:

  • Oklahoma – 10.6
  • Florida – 10.8
  • Boise State – 11.0
  • Kansas – 11.2
  • Navy – 11.3
  • LSU – 11.4
  • UCF – 11.5
  • Virginia Tech – 11.5
  • Georgia – 11.5
  • West Virginia – 11.5

That’s a pretty disparate bunch. But those schools do have a couple of things in common. First of all, every one of them had winning records last year. Not just barely, either – the worst of the group was Navy’s 8-5 mark, and Navy and Florida are the only two on the list that didn’t win in double figures. Second, those schools did a good job of scoring, the ten averaging almost 38.5 points per game. (Georgia was next to last in scoring among the ten, at 32.6 ppg.)

What the ten don’t have in common is high passing completion percentages. Half the schools on that list exceeded 63% last season; the other half fell below 58%. (Georgia was next to last here, as well.)

Judging from Georgia’s numbers, it looks like there’s more than one way to skin the scoring efficiency cat.  What might some of those other factors be?  This blogger offers a few:

  • turnovers
  • special teams
  • reliance on the running game [“because the risk of a turnover is much lower on a running play (only about 1/3 the risk of an interception on a passing play) and the risk of a zero gain is much less.”]
  • third down conversions (check out Groo’s post on this topic, as well)

I can think of another.

So what has Georgia been doing specifically to be offensively efficient (that YPP figure, by the way, is easily the best of the Richt era)?  And in light of what that might be, where does Stafford’s completion percentage fit into the picture of what Georgia needs to do to be successful?  More to come on this…


Filed under College Football, Georgia Football, Stats Geek!

4 responses to “More Stafford, completion percentage, efficiency, etc.

  1. BenG

    YPP speaks to the strength of a team’s defense and special teams as well as its offense, no? A good defense keeps the ball on the other side of the field, even when it’s in the hands of the opposing offense. Likewise, good punt and kick return teams can make the difference between a long field and a short field to start a drive. YPP seems to indicate the strength of the team overall.


  2. It depends on the team, doesn’t it?

    For example, Navy is fifth on that YPP list. The Midshipmen ranked 99th in the country last year in total defense.


  3. BCDawg97

    I’ll put my criticism this way. In a year we certainly have the talent to win the NC, we need a majority of breaks to go our way. Sure the running game and defense will be solid. And of course we can’t control somethings like injuries or the way the ball bounces on a fumble. But Stafford scares me in that we might need a 3rd and 6 at some point to stay alive or win a game and he’s going to throw off his back foot or rocket it in there instead of with some touch (which in my thinking is one reason why his % is low up to this point). Yes he’s learned to throw balls away and had less INTs. Yes he has all the talent in the world. But something just makes me nervous when drops back to pass. This is just my perception based on what we’ve seen for 2 years. I will gladly eat 1000 lbs of crow on Jan 8 in Miami while I cry like a little girl when we win the NC.


  4. Hobnail_Boot

    You know what most of those teams do have in common? Eight of them had at least one prolific player who turned a lot of FGA’s into TD’s.

    Oklahoma – Bradford (36 TD), Murray (13)
    Florida – Tebow (55)
    Boise State – Tharp (30), Johnson (16)
    Kansas – Reesing (33), McAnderson (16)
    Navy – Kaheaku-Enhada (20)
    LSU – No prolific scorers, but one bat-** crazy coach.
    UCF – Smith (29)
    Virginia Tech – No prolific scorers, but great special teams.
    Georgia – Moreno (14)
    West Virginia – White (28), Slaton (17)

    It’s not a surefire way to gauge team success (see: Crabtree, Michael) but it’s something to think about.