A Super Bowl Indicator: Net TD Efficiency

In the last post we ranked NFL offenses according to their touchdown efficiency per plays, yards, and drives. Eliminating field goals and calculating rankings based on efficiency rather than raw scores helped isolate the best and worst offenses.

So for the next step, we’ll add defense. I added up touchdowns allowed by each defense and figured out the same efficiency stats as yesterday. Then, simple enough, I subtracted defense from offense. Positive scores indicate better teams. For example, the New England Patriots make their opponents go 13 extra plays, 64 extra yards, and 2 extra drives to get a touchdown. They are ranked third overall.

First and second, by a solid margin, go to the two Super Bowl teams. They were 1-2 in every category, and outpaced the third place teams significantly in Net Plays per TD and Net Drives per TD. Predictive stats are often too reductive, but if you used this measurement to pick the Super Bowl teams at the end of the regular season — it would have guided you well.

The Eagles, despite their highly ranked offense, come out only 13th in the net rankings. If you didn’t figure it out already, Sean McDermott’s defense was a huge liability in 2010. Also note that while we’ve eliminated some biases, the stat doesn’t account for strength of schedule. Not sure how much that would change things, but keep it in mind nonetheless.

See the full rankings below.

Photo from Getty.

Ranking NFL Offenses by Touchdown Efficiency

Yesterday we saw how offensive touchdowns are much more important for success than field goals. But let’s not stop there.

If touchdowns are important, it’s not enough to just rank teams by the number of touchdowns. Some teams have more opportunities or have worse field position. Instead of just counting overall touchdowns, we can weight them by how many yards, plays, or drives each team needs to get one.

With that in mind, I calculated touchdowns scored per all three factors in the 2010 season. Each measures something slightly different. For example, the Eagles were 8th in Yards per TD, 5th in Plays per TD, and 7th in Drives per TD. Fairly close, but the best showing in Plays per TD probably reflects the huge number of big plays and long touchdown catches last year.

Below you’ll find the full chart of all NFL teams, best offense to least according to the average rank of these three measurements. It likely gives a more accurate picture of the best and worst offenses than any ranking that accounted for field goals or other methods of scoring.

Tomorrow we’ll bring the ranking one final step further.

Photo from Getty.

Friday Figures: Eagles Offensive Play Count Stats

Not a lot of complexity to this post. Just some stats and (hopefully) a few interesting observations. Let’s jump right in.

Ladies and gentlemen, your 2009 Eagles offensive player play counts, broken down by run/pass, courtesy of PFF:

So about 40% of the Eagles’ offensive plays were runs last year. When some of these players were on the field though, one could make a pretty good guess that a run was coming. Take Mr. Eldra Buckley for example. When he was in the game the Eagles were over two times more likely to have Donovan hand the ball off.

The embattled wildcat quarterback, the second/blocking tight end, and the fullback, all also signaled to the defense that a run was on the way. Brian Westbook on the other hand? Not so much.

Let’s take a gander at 2008 for comparison’s sake:

Basically the same team run/pass ratio, but it gives us a few more data points and intriguging spots to look.

• Eldra Buckley is apparently Kyle Eckel part 2. Not good for Eldra’s chances of making this team — the Eagles can easily use Charles Scott for those 30-40 plays.
• Speaking of which, notice how almost all of the guys who heavily skew to rushes tend to get replaced: Eckel, Klecko, Schobel, Baskett, (Alex) Smith. If you can’t be multidimensional in Andy’s offense, you’re going to have a hard time sticking around.
• Hank Baskett is a tight end in wide receiver’s clothes. Meaning Riley Cooper better work on blocking if he wants to supplant Baskett as the #4 guy off the bench.
• Actually there is a pretty clear correlation. The further down the depth chart a player is at wide receiver, the more run blocking he’ll be required to do.
• Brent Celek played almost every offensive snap. What a beast.
• Brian Westbrook became much more pass-centered last year. Might be part of how the coaches tried to shield him from a big workload.
• Andy Reid must feel much more confident in Kevin Kolb now than a year ago. Back in 2008 he was helping Kolb out by calling runs more than half the time. In the two games in 2009, however, Kolb passed about as much as McNabb.

A Second Look at Defensive Playmakers

One of the keys to the 2010 Eagles season will be the performance of the defense. Last year the unit was, at best, inconsistent and this offseason we’ve seen a good deal of turnover. Sheldon Brown is gone. So is Darren Howard, Chris Clemons, Will Witherspoon, Sean Jones. Stewart Bradley returns, along with a whole host of draft picks and other new additions such as Darryl Tapp and Ernie Sims.

But it is often difficult to isolate individual player performance from team performance. How do we know whether the guys the Eagles jettisoned were truly underperforming? Hopefully this post will offer one possible way to gain some insight into that, simply by crossing two streams (oh no!) of data: one from Advanced NFL Stats and the other from Pro Football Focus.

Let’s start with Brian Burke’s Advanced NFL Stats measurement of +EPA as a measurement for defensive players. As he explains here, +EPA measures positive performance: “the value of every sack, interception, pass defense, forced fumble or recovery, and every tackle or assist that results in a setback for the offense.” What the stat doesn’t do is show you when the player did something wrong — when he cost the team. Thus, one would expect risk-taking players such as Asante Samuel to score disproportionately well in this system. His positive achievements regarding turnovers are huge, while his negative tackling ability never gets factored in.

Still, even with this caveat, it is an interesting way to look at players as playmakers. Here are the Eagles’ numbers from last year as put up by Burke, ordered by +EPA per the number of games they played in.

Obviously bland +EPA is of little use to determine who’s a better player. It’s far too dependent on the amount a player is on the field. At least Burke’s +EPA/G differentiates by number of games played to factor out some of the bias from players who played constantly and without injury. However, when I looked at this for the first time, I realized that games played was only a half step toward the greater goal of figuring out who caused the most postive plays, per play.

This is where Pro Football Reference’s play charting data — where they count how many times every player is on the field — comes in handy. Here’s for the Eagles. By looking at +EPA by play, rather than by game, you can get a much more “accurate” reading of playmakers. Below you can find the +EPA per time the player was on the field defensively (+EPA/Play), as well as a number that shows how much +EPA he would have accumulated had he played every defensive snap in the average Eagles’ game (~65).

Observations from this data (feel free to post your own in the comments):

• The fewer times a player was on the field, but active in the game, the more unreliable his numbers are going to be. Looking at you Tracy White and Quintin Demps.
• On the other hand, those who racked up huge play totals as well as +EPA/Avg. G (Trent Cole, Asante Samuel) clearly form the backbone of the playmaking contingent of the team.
• Darryl Tapp’s numbers with the Seahawks: 39.8 +EPA total, 704 plays, for a +EPA/Avg. G of 3.67 — placing him right near the top of the list.
• Ernie Sims’s from Detroit: 14.8 +EPA total, 490 plays, for a +EPA/Avg. G of 1.96 — positively Bunkley-ian.
• The scheme also affects opportunities for playmaking. For example, all four of the main defensive tackles fall in a row toward the bottom, indicating that the Eagles have been using the position less for playmaking, and more for solid building blocks of run defense. At the opposite end of the spectrum is Darren Howard, who often played as a DT in pass-rushing situations, giving him many more opportunities to make plays.
• But the fact that Howard was able to cause that pressure when he was in the game is important. While we can’t see some of the negative plays he may have allowed, Howard was clearly a better pass-rushing option than Chris Clemons, who made barely more +EPA per play than Mike Patterson. One has to wonder if Clemons for Tapp was a steal and/or if the Eagles will miss Howard’s rushing ability on third down.
• Other players who seemed to provide little in the area of playmaking: Macho Harris, Sean Jones, Chris Gocong, Will Witherspoon.

It’s worth mentioning one more time that this is far from the final answer at evaluating defensive performance. However, by combining +EPA with play counts, we gain one more interesting nugget for individual evaluation.