Offseason Stat-Plosion: The Team Efficiency Index

punt block

It’s 2013, and statistics are ubiquitous in college football: yards per play, 3rd down conversion rate, points per game, defensive efficiency, time of possession, attractiveness of fake girlfriends, turnover margin, blah blah blah blah.   We’ve got all kinds of “objective” statistical methods to “measure” how “good” a football team “is.”  This clearly ain’t Knute Rockne’s football.

The problem with all these objective measures is: football is not objective!  You cannot compare stats, scores, indexes, or any other numbers to predict how one team will do against another team!  We see this time and time again; the pundits quote stats, use those stats to come to conclusions or predictions, and then get it completely wrong.  (I am not at all referring to Lou Holtz.)

You don’t even want to know what kind of stats THIS guy is into.

No, friends: football is the most subjective of games: how do teams respond to an opponent score?  Hostile environments?  Inclement weather?  Unfortunate turnovers?  Shootouts?  Defensive battles?  Having their plays called by Jeff Tedford?  (One of those factors has limited applicability.)

The task of measuring “goodness” becomes extraordinarily difficult once we consider those subjective factors.

Some people are able to accomplish BOTH primary purposes of football.

So, what’s the purest measure of how “good” a team is?  It’s not time of possession, not margin of victory, not points per game.  Think simpler! What’s the primary purpose of football? SCORE MORE POINTS THAN THE OTHER GUY.  (Oh, and beat ‘SC.)

 

“Now, Sean,” some of you might say, “isn’t that just measuring margin of victory?”  No, not really.  You and I know that different teams, conferences, and regions have different theories on how to play the game, and different levels of points are scored as a result: Oregon’s point-a-minute-win-by-a-lot game versus Notre Dame’s ground-and-pound-win-by-a-few is a good example.  (Even as a tremendous Oregon homer, I admit that there’s an art to slugging out that 10-7 victory.  I just would rather poke needles in my eyes than watch it.)

A snippet of the action from the thrilling 7-3 Georgia victory over Kentucky last season.

“So, Sean,” you say, “what’s your genius way to compare teams based on ‘scoring more points than the other guy?'”  Firstly, stop asking so many questions.  Secondly, it’s a little thing I like to call THE TEAM EFFICIENCY INDEX.  The most basic explanation of the Index is the ratio of points scored to points allowed, or (Total Points Scored / Total Points Allowed) = TEAM EFFICIENCY INDEX (or TIE).

This ratio carried over an entire season demonstrates the responsiveness of a team: can a team score after scored upon?  Respond to other adversities?  Maintain a lead?  Build a lead?  This, as explained above, is different than straight-up margin of victory.  A 14-7 victory can be just as impressive an outing as a 34-17 victory.  In the close game, your offense needs to play field position, and your defense needs to get every stop possible.  In the “shootout,” your offense has to score and score some more, and your defense needs to get a stop approximately two out of every three possessions.  But both game plans are designed to do the same thing: maintain the points ratio!

I admit that the TIE is not perfect.  It does not adequately reward teams for winning good close games, gives more weight to blowouts over cupcakes, and has no regard for strength of schedule.  But since we still have five months until college football kicks off, we can use some fodder for debate, can’t we?

And if you REALLY have a problem with it, go ahead and make your own index.

You too can create a football statistics index! Huzzah!

Without further ado, here is the TEAM EFFICIENCY INDEX for the AP Top 25 from the finish of last season, reordered to reflect each team’s TIE:

Team

(AP), (Record)

Points Scored

Points Allowed

Team Efficiency Index

Fewest Points Allowed

1. Alabama

(1), (13-1)

542

153

3.54

2. Florida State

(10), (12-2)

550

206

2.67

Most Points Scored

3. Oregon

(2), (12-1)

645

281

2.29

 

4. Utah State

(16), (11-2)

454

200

2.27

 

5. Texas A&M

(5, tie), (11-2)

578

283

2.04

Fewest Points Scored

6. Notre Dame

(4), (12-1)

335

166

2.02

 

7. Northern Illinois (22), (12-2)

540

278

1.94

 

8. Georgia

(5, tie), (12-2)

529

275

1.92

 

9. Boise State

(18), (11-2)

393

205

1.92

 

10. Florida

(9), (11-2)

344

188

1.83

 

11. Kansas State

(12), (11-2)

505

288

1.75

12. South Carolina

(8), (11-2)

410

237

1.73

 

13. LSU

(14), (10-3)

387

228

1.70

14. Clemson

(11), (11-2)

533

323

1.65

15. Ohio State

(3), (12-0)

446

274

1.63

16. San Jose State

(21), (11-2)

452

277

1.63

17. Stanford

(7), (12-2)

390

241

1.62

18. Vanderbilt

(23), (9-4)

390

243

1.60

19. Oregon State

(20), (9-4)

423

268

1.58

20. Oklahoma

(15), 10-3)

496

331

1.50

21. Michigan

(24), (8-5)

388

258

1.50

22. Northwestern

(17), (10-3)

412

293

1.41

 

23. Louisville

(13), (11-2)

405

309

1.31

Most Points Allowed

24. Nebraska

(25), (10-4)

487

386

1.26

 

25. Texas

(19), (9-4)

464

380

1.22

 

So, conclusions?

A. The Ducks’ numbers are (surprise!) really, really good.  The Ducks scored 67 more points than the second-best scoring team in the AP Top 25: Texas A&M.  This contributed to their #3 overall TIE of 2.29, and took the phrase, “The best defense is a good offense,” to a whole new level.

B. Alabama completely dominated people. Though it pains me to say it, scoring more than three-and-a-half times as often as your opponents is damn impressive.  Damn you, Saban!

C. Ohio State and Stanford were the most overvalued by the AP rankings (3rd, 7th) compared to the Team Efficiency Index (15th, 17th).  This may be explained by the fact that both teams won very close games that voters loved to watch.  Also, both teams played VERY clutch down the stretch: Ohio State winning at Camp Randall and beating Michigan to go 12-0; Stanford beating Oregon State, Oregon (single tear), UCLA (twice), and Wisconsin en route to a Rose Bowl title.

D. Utah State, Northern Illinois, and Boise State were the most overvalued by the Team Efficiency Index (4th, 7th, 9th) compared to the AP rankings (16th, 22nd, 18th).  This may be most simply explained by the conferences that they play in, as success in the WAC, MAC, and Mountain West is not considered on the same plane as success in the “power” conferences.”  But their TIE is competitive with anyone in the country.

E. Some Top 25 teams (Louisville, Nebraska, Texas) BARELY outscored their opponents.  Scoring 31%, 26%, or 22% more than your opponents is not exactly dominant.  But these teams got the wins they needed.

Thanks for walking with me along the path to statistical enlightenment.  Tell me what I got wrong, where my analysis misses, and what sticks out to you in this Team Efficiency Index system in the comments.  After all, we can use the distraction and debate during the long offseason!

Talk numbers all you want, but this is pure poetry. Credit Kevin Cline and Kaly Harward.

Print Friendly
Sean Goodbody

Sean Goodbody

With a high school defensive-coordinator-turned-offensive-coordinator-turned-head-coach as a father, Sean Goodbody has always had an interest in the X's and O's of football. He played two years with FCS University of Pennsylvania as a fullback, but having grown up in a football family Sean has spent much of his life studying the game--reviewing game film, designing offensive and defensive schemes, and game-planning upcoming opponents. Sean has coached running backs, option quarterbacks, linebackers, defensive linemen, and safeties for his dad's high school program. He has been a rabid Duck fan since meeting his significant other Maeve (an Oregon grad). Residing in Grand Junction, CO, Sean and Maeve both work as attorneys while cheering on their beloved Ducks from the Rockies.

  • NativeDuck

    Sean,

    You were right early on that football stats are subjective. This “TIE” is no different.

    No matter what “stat” you come up with it will be subjective. No two teams play the exact same schedule (and I do mean exact, same day, time, game plan, etc). The defenses and offenses played vary – even against the same opponent (I know, this is game planning). Until there is some way to normalize the data every stat will be subjective.

    To normalize the raw data you need some OBJECTIVE way to take ALL variables (a missing or dinged up player, motivation, etc) and put everything into a common unit (ie: mg/ml/mi, mg/ml/degree-F, or yds/play/standard-defense/…). The problem is you can’t get to any common unit that does not have variables which are glossed over (ie: yds/play – does not account for the defense in an objective way).

    • SeanG

      Totally agree with you. The TIE is subjective too, despite my (hopefully) clever defense of them as objective. That’s why it’s fun to argue which stats are relevant, and which are not. In this case, I thought it would be fun to compare the absolute ratios of points for each teams, to reject PPG and PPG-allowed as a useful tool. Like I said; it’s the offseason, we can use the debate topics! Thanks for reading.

  • I know people love to do the stats thing. But ultimately in my mind any list that ends up putting teams like Utah State/N. Illinois up in the same league with the Oregon’s and Alabama’s of the world is ridiculous.

    N. Illinois in a BCS last year was a complete joke. I think FSU should be fined for only beating them by 3 TD’s.
    I will give BSU some credit, when they get into those big games Peterson has them ready and they have competed.

    Heck, remember just a couple years ago when we played Nevada the commentators were referring to it as a tune up game. Several weeks later Nevada and BSU are playing and they were calling Nevada BSU’s big test.

    I know I’ll get little argument here, but if we had a football playoff you could let all the 11-1 teams like Utah State and N. Illinois compete and ultimately we’d know who was truly the best. Works in every other sport!

  • guest

    You need to work on your acronymns, or YNTWOAY. Get it? But I like your style.

    • SeanG

      TVMAIRATF. (Thanks very much and I really appreciate the feedback.)

      • guest

        Team Efficiency Index is TEI. Now do you get it ;)

  • Sjb

    What about “average TIE per game”?