I had some time to play with some numbers earlier this week, so I decided to look at measures of offensive efficiency. I looked at number of plays, total yards, and touchdowns scored from the stat box. Plays and yards is obvious, but I looked at touchdowns only because scoring a touchdown is the most important thing an offense can do.
I then looked at two ratios. The first was yards per play, because it’s a measure of how well a team moves the ball. The other ratio is plays/TD, which tells the character of an offense, whether it is a quick strike team or slower, methodical, and less efficient team. Now, to each category.
As you might expect, the number of plays is the worst predictor of success there is. Of the 70 games involving I-A teams last weekend, only 39 winners, or 55.71% ran more plays. The top ten teams in number of plays run are:
1. Memphis, 96 – loss
1. Nebraska, 96 – win
3. New Mexico, 91 – loss
4. Penn State, 89 – win
5. Houston, 88 – loss
5. Wyoming, 88 – win
7. Missouri, 87 – win
8. Tulsa, 86 – win
9. Arizona State, 85 – win
9. Stanford, 85 – loss
9. Toledo, 85 – loss
That’s technically 11 teams, with a ratio of 6 winners to 5 losers. Of the top 20, 11 are winners and 9 are losers. This definitively shows that number of plays does not predict a winner.
Intuitively, outgaining an opponent will generally lead to victory. Mathematically? Well, for one week anyway this is true, with 57 teams, or 81.43% of teams, that outgained their opponents winning. I have a feeling that the abundance of good teams playing terrible teams may bias these stats towards victory for the bigger gainers, but over the course of the season any anomalies should even out. The notable losers:
14. Houston, 545 yards
26. Michigan, 479 yards
31. Memphis, 467 yards
38. Illinois, 435 yards
39. Minnesota, 434 yards
These teams are the only ones with more than 400 yards of offense that lost. Michigan is a classic case study on why you need more than offense to win a game. Despite outgaining Appalachian State 479-387, it was interceptions, blocked field goals, and missed 2 point conversions prevented the Wolverines from avoiding humiliation. Racking up a lot of yards is generally a good sign and an indicator that your team is likely to win.
Intuitively, scoring touchdowns is the most important thing you can do, and if you score more touchdowns that your opponents, you should win. After all, it takes at least 3 scores by other means to overcome the effect of a touchdown and PAT. Now, this category could be biased some because it doesn’t look at defensive or special teams touchdowns, which are comparatively rare. That is why teams that excel in defense scoring (like the 2000 Baltimore Ravens) tend to be extremely tough to beat.
As it turns out, 61 of 70 winning teams scored more offensive touchdowns than their opponents. That’s a 87.14% win rate for scoring more touchdowns. So, if you get in the endzone more on offense, you’re likely going to win.
The highest up team to lose in the national stats was Illinois, who tied for 24th place with 5 TDs scored. No other team with 5 or more touchdowns lost, but it should be noted again that Illinois’ opponent Missouri also scored 5 TDs. Several teams scored 4 TDs and lost.
Yards per Play
This is the first real measure of efficiency. Gaining more yards per play means your offense can move the ball in larger chunks than your opponent can. If you lose the yards per play battle but win the game, it’s not necessarily a good sign because on a play by play basis, your defense got outplayed. If you win yards per play but lose the game, that’s obviously bad because you lost the game because you couldn’t capitalize on your superior ball movement.
For the weekend, teams with higher yards per play won 62 games of 70 for 88.57% of the contests. Louisville won this category for the weekend with a staggering 10.23 yards per play. In other words, Louisville averaged a first down every play. Gator fans rejoice, because Florida came in second at 9.44 yards per play. Oklahoma, who won the weekend in total yards at 668, finished third at 9.15 yards per play. Rutgers at 8.53, and Hawaii at 8.36 round out the top 5. The rest of the top ten were: West Virginia (8.09), UCLA (7.90), Michigan State (7.60), Cincinnati (7.41), and Purdue (7.39).
The highest team in the national stats to lose was Michigan, at 6.22 yards per play, the 29th highest total for the weekend. Houston followed closely at 30th with 6.19 yards per play. Michigan and Houston keep showing up as having had good offensive statistics but still having lost. Once again, this illustrates why it takes more than just offense to win a game. Interestingly enough, Michigan and Houston’s opponent, Oregon, play this weekend.
Plays per TD
This is like the AB per HR stat in baseball. The gaudier this is, the better you are at scoring touchdowns quickly. You must score lots of touchdowns to do well in this category, and that’s the point since scoring more touchdowns means a better shot at winning. As it turns out, this is the best indicator of winning at 69 of the 70 winners, or 98.57%, this weekend had a better plays per TD ratio. The only team with a better ratio that lost? Poor old Illinois again, who clocked in at 33 overall and who lost to 40th place Missouri. Only 4 teams in the top 50 in the country for plays per TD lost, and the three other than Illinois lost to teams that had better ratios.
In his heyday, Barry Bonds was hitting homers at a rate of one every 8 or 9 at bats. The top teams for last weekend unsurprisingly were Louisville (6.40) and Oklahoma (6.64) who scored 10 and 11 touchdowns, respectively. The top ten:
1. Louisville, 6.40 (64 plays, 10 TDs)
2. Oklahoma, 6.64 (73, 11)
3. West Virginia, 7.44 (67, 9)
4. Hawaii, 7.67 (69, 9)
5. Florida, 7.71 (54, 7)
6. Cincinnati, 9.22 (83, 9)
7. Purdue, 9.43 (66, 7)
8. Washington, 10.33 (62, 6)
9. Boise State, 10.38 (83, 8 )
10. Indiana, 10.43 (73, 7)
Anyone who had Indiana showing up in the top ten, give yourself a pat on the back. Florida once again shows up as one of the most efficient offenses, which makes sense because the offense with Tim Tebow under center was nigh unstoppable and garbage time with all the freshman got lightninged out. Go Gators.
The results of this past weekend indicate that the best predictor of winning from an offensive standpoint is having a higher plays per TD ratio than your opponent. This makes sense when you take a step back because it means not only was your offense better, it means your defense did a better job of keeping the other team out of the endzone. So, in a way it covers both offense and defense and it definitely takes both to win a game. It ignores special teams score, but that’s fine since they are rare.
It also indicates that teams should go for it on 4th and short in opponents’ territory rather than settle for a field goal since all of a drive’s plays are basically wasted if you don’t score a touchdown because the offense failed to do the most important thing it can do. Sure there will be times when kicking a field goal is worth it, but as I said before you need three field goals just to offset one touchdown and PAT.
Page 2’s Gregg Easterbrook has worked with Accuscore and run simulations and found that not punting in certain situations adds about a point to a team’s score per game. In the NFL, that is a lot more dramatic than in college where more points tend to be scored. What my numbers show so far is that scoring more touchdowns will get you a winning percentage around .871, and that will get you in the college football hall of fame.
Focusing on driving yards per TD down may have an unintended side effect of also driving down your time of possession and wearing out your defense. Well, a proper mix of running with your passing and not running tons of sideline plays will keep the clock moving regardless. There’s also the correlation/causation question: does having a low plays/TD ratio mean you have a good offense, or is a side effect of having a good offense found in having a low plays/TD ratio? Well, I tend to think the former because having a big play-oriented offense is a choice, controlling field position so you have fewer yards to go to score is a choice, and consciously choosing to go after touchdowns is a choice.
Winning the plays per TD stat though will get you a winning percentage of .986, which is completely unprecedented. I urge coaches to go for TDs, since that’s what will win you games. If you have a reliable deep threat, use it early and often because that will drive your yards per play up and your plays per TD down. And for goodness sakes, make sure your defense understands this too, and does whatever it can to keep teams out of the endzone. This includes never, ever playing prevent defense. It may seem obvious, but even trading one touchdown for two field goals over every three possessions for each team will get you a loss every time.