My view of interceptions

A few leftover thoughts from week 4.

I get why everyone is upset. Dallas lost 2 very winnable games and those games could be the difference between 8-8 and 10-6 or 10-6 and 12-4. I’m probably a little more relaxed because I really didn’t think that Dallas was good enough to succeed in the playoffs this year so this all seems like gravy. However, Dallas has played so well that those 2 games may turn out to be a huge missed opportunity (e.g. the difference between playing at home and playing on the road in the playoffs) … so maybe I should be a little more exercised.

The jumping to conclusions mat

I thought the first 3 weeks taught us not to rush to snap judgments. How optimistic I was.

Wk 1: Romo’s a goat

Wk 2: Romo’s a hero

Wk 3: Romo’s incredible, I’ll never doubt him again

Wk 4: Romo’s a goat! Why did I ever like him!

I mean, Tom Brady would never throw 3 interceptions and give away a game like that, right? Romo sucks! Of course in week 3 Tom Brady threw four interceptions and gave away the Buffalo game.

I've stopped worrying about interceptions. Interceptions are an inevitable part of football. Tom Brady can’t avoid 4 pick games. Peyton Manning can’t avoid 4 pick games. Brees can’t. Rogers can’t. Rivers can’t. Neither can Tony Romo. Passing is not optional. NFL teams cannot win with without passing. Interceptions are also unpredictable. It would be foolhardy for me to try and draw conclusions on a game to game basis from a variable that’s highly unpredictable.

In the Detroit game Dallas had everything you’d want. They were playing with a big lead (lower likelihood of interceptions). Dallas had good run / pass balance. Dallas had manageable down and distance. And Romo still threw 3 picks.

To the extent I draw conclusions from interceptions I look at the long term data. I look at interception rates for the season. See if Romo’s interception rate for the season is higher than Brady, Brees, Rivers, Rogers. And btw, I write this not to be argumentative. I write this to help with your sanity. You’ll be a lot happier and weather the up and downs a lot better if you don't worry about every blip in a tremendously unpredictable event.

Long term perspective

Here’s what I wrote after week 3

remember that Washington muffed a gimme FG that easily could have tipped this game to the Skins. Absent that muffed FG Dallas very likely would have lost this game and you’d be hearing a very different fairy tale, one about how Dallas lost because Garrett hasn’t fixed all the old problems: penalties, mistimed snaps, scoring in the red zone, WRs running wrong routes, boneheaded plays by Tashard Choice to end halves, etc.

Right on cue we get a close game that Dallas loses. As I said above, taking a longer term perspective helps me weather the ups and downs. After week 3 the long term perspective said Garrett hadn’t discovered the secret to winning close games, after Detroit it says Romo isn’t a goat.

For his career Romo’s interception rate right at 3%. Here are some other top QBs. Rogers: 1.94%, Brady: 2.2%, Eli: 3.3%, Brees: 2.3%, Big Ben: 3.1%, Rivers: 2.4%, Peyton: 2.7%.

Having looked at other elite QBs I agree there is room for improvement. Brady and Rodgers have 2% career interception rates while Romo’s career interception rate is 3%. So I think a 3% rate can be improved.

However, 3% isn’t an atrocious figure. With Big Ben and Eli clocking in at 3.1% and 3.3% respectively it’s clearly possible to win a Super Bowl with a 3% interception rate (and Romo is better than those QBs in other measures like YPA and completion %).

KC Joyner makes a similar point in an interview with Raf

KC: Ben Roethlisberger was one of the worst decision makers in the league for years. The Steelers would still win because of their defense and because they have an explosive offense … so they could absorb Roethlisberger's bad plays.

I’m also going to reference one of my favorite pieces of research. The research looked at the statistics for QB’s who changed team and which statistics stayed the same and which statistics changed. The idea was to isolate which statistics are dependent on the QB and which statistics are dependent on the team (remember there are team interactions, the action of the receivers can cause an interception). Here’s what the study found.

It's pretty clear which performance category is least consistent from year to year and probably belongs to the individual quarterback the least. It's probably the one the general public uses to judge a quarterback the most--interceptions. You rarely hear after a game about how the receivers caused the interceptions, or the bad luck of it all, or the game context dictated the interceptions.

The lesson is that teammates have a material impact. If you think Dallas should be gunning for a 2% interception rate its takes a team effort. For example, Matt Cassell had a 2.1% interception rate his one year running the NE offense (similar to Brady’s). However, Cassell has a 2.65% rate in KC in the NFL’s most run heavy offense. This year Cassell’s interception rate is 4.5%.

Are Romo’s teammates having an effect? Here are 2 notes that suggest they might be.

5 of 7 interceptions last season were all tipped off the hands of receivers as well.

-Brandon Worley

And the Cowboys Nation recap of the Detriot game

Last week's meltdown has obscured a disturbing fact. In the Washington game, Dallas didn't have a number one or even a number two quality receiver target…

Against Detroit, Robinson played the Miles Austin role. He did a remarkable job, but after he hurt his foot making a lunging 2nd-and-6 catch, Tony Romo did not complete a single pass to a wide receiver over the last 22 minutes of the game. Robinson was limping. Dez Bryant was aching. Ogletree was wherever he was.

All the downfield completions in that span were to Jason Witten. He's a great tight end, but you can't sustain a passing attack around a tight end

Now that doesn’t prove that Romo isn’t at fault, but it’s not surprising. It certainly makes the interceptions in the Detroit game a little more understandable to me. We’re working with small numbers here. 10-18 interceptions a season so 5 or 6 mistakes by teammates has a huge impact.

Personally, I hadn’t realized Robinson hurt his foot. We know that Laurent Robinson contributed to the 2nd pick. And the third pick was a forced throw to Witten. Isn’t it a little more understandable that Romo was forcing the ball to Witten?

So part of the explanation may be Garrett’s offense (fair criticism). Part of the explanation may be the roster instability (when Romo had stability at WR in 2009 with Austin breaking out and you saw his interception rate go down). Part of the explanation is the receivers.

My conclusion is that while there is room for some improvement it’s not a fatal problem even if there were was no improvement. However I think that with offensive stability (OL gains experience and WRs get healthy) you probably will see some improvement.

Romo's interceptions: there is no trend

Here are Romo's interceptions for his career.

0 0 0 3 1 0 0 1 0 2
2 1 2 1 1 0 1 1 5 1
0 1 1 1 1 1 0 3 1 1
1 1 1 1 1 0 2 0 1 3
0 2 1 0 3 0 1 0 0 0
1 1 1 0 0 0 0 1 1 0
2 0 3 2 0 1 0 1 3

I posted these and said that it was a brainteaser. I said the brainteaser was figuring out what the pattern was. The joke was that would be a devilishly hard teaser because there is no pattern. There is no trend. Romo is getting no better or no worse.

One last note

ESPN’s new QBR rating found the Eli Manning was significantly undervalued. Why?

Now here's where things get interesting. Manning took just 16 sacks in 2010, even as injuries scrambled his offensive line, while Rivers was sacked 38 times. That difference was almost enough by itself to make up for the interceptions gap between the two QBs. Manning also cost his team fewer points by fumbling, ran for more crucial yards and induced more valuable penalties by opponents than Rivers; in each of these categories, Manning was above average and Rivers below average. Add it all up and Manning added 10.15 points to the Giants per 100 plays last season, while Rivers added 9.78 points per 100 plays to the Chargers.

These results don't prove that Manning is a better quarterback than Rivers, or even that he was significantly better in 2010. (The overall difference between them was very small). But they do demonstrate two important things:

1. Play-by-play analysis and data that isn't in traditional stat lines can make a huge difference in comparing players, in this case closing a gap of 16.5 points in passer rating.

2. Of all the things that don't show up in box scores, ball-discipline skills are the most important.

I wouldn't have guessed this. I thought that rushing yards would be the most important factor that QBR counts and passer rating doesn't. But from 2008-10, the average starting quarterback added just 0.56 points per 100 plays to his team by rushing or scrambling, and just a handful of QBs made a real difference with their legs.

In contrast, the combined cost of sacks, fumbles and penalties is much greater: an average of 4.91 points per 100 plays for starting QBs. And QBs' range of performance is much bigger, too, from Peyton Manning (-0.81) at the top of the list to JaMarcus Russell (-8.52) at the bottom.

Peyton Manning's total here is unbelievable, and is one more sign of why he's so irreplaceable: The guy essentially never fumbled or took an avoidable sack, induced far more penalties than he committed and lost a grand total of three or four points per season to ball-control errors.

The interesting information here is the importance of ball-control errors.

And also from Pro Football Reference

While sacks can be the fault of the offensive line or the accomplishment of the defensive player, the evidence is pretty clear that the quarterback is at least as responsible for his team's sack rate as other passing performance measures that we readily attribute primarily to the quarterback.

I focus and worry about the things that Romo can control. There’s always a lot of discussion about interceptions because they have a large impact on outcomes. However QBs cannot entirely control whether passes are intercepted because of teammate interactions.

Meanwhile factors that the QB does control that have an impact on the game are almost completely ignored. For example, I haven’t seen anyone mention that after all the snap difficulties in Washington Romo induced 3 offsides penalties against the Lions. Sacks, fumbles, penalties: these are factors that a QB can control and therefore fans should rightly be upset if their QB is performing poorly. For me, I like to worry less about the unpredictable factors and more about the things the QB does control. Helps preserve my sanity.

Another user-created commentary provided by a BTB reader.