Re: To all (well the 3) Defense is not the Problem Posters
posted at 3/21/2014 2:55 PM EDT
In response to seattlepat70's comment:
In response to prolate0spheroid's comment:
Seattle, I agree that the O did not perform great in either of those games. In 2007, in particular, the O performed particularly badly, mostly because it was vastly overmatched in the trenches. In 2011, I'd say the O was mediocre, largely because the O has mostly been mediocre when you take Gronk out. We don't have enough other weapons. Still, on a points per drive basis, they weren't horrendous.
The issue, though, is whether the defense played well. I think it played okay in 2007, but in 2011, despite the 19 point total, it was bad. Basically, the Giants had just 8 offensive drives (their ninth was a kneel down to end the first half, so I don't count that one). They scored on half of them--two TDs and two FGs. The Pats gave up 2.375 points per drive (and it would have been 2.5 if the Giants had gone for a one-point conversion and made it rather than missing a two-point conversion). Just as bad, the length of the drives was astounding: 4:37 on average, 4:53 if you exclude the late two-minute-drill drive when the Giants had to score fast (which, of course, they did). On average in the NFL in 2011, drives were 2:37, so the Giants' average drives were a full 2 minutes longer than expected--not quite double the average length, but close. Then you have Manning's 75% completion rate (30 completions on 40 passes). These are horrendous defensive statistics. People who ignore that and just look at 19 points really miss what was actually happening in the game. Nineteen points is a bit below average for a game - - - but it's way above average for 8 possessions --- and if you look at everything that happened in that game there's no way to claim the defense played well.
The Giants strategy was to keep the ball out of the Patriots' hands, and our defense did absolutely nothing to stop them from executing their strategy. A "keep away" strategy, if it succeeds, will result in a low number of possessions and low final score. But that low score wasn't really the result of our defense playing well. It was primarily the result of the Giants' offensive strategy working and our defense being helpless to stop it.
I think I agree... giving up 2.375 points per drive, I think is bad.
The word "think" is there because I don't know if what the win/loss probabilities are for different levels of PPD. Analyzing that is a big task I am not willing to spend time on. It feels right, I can say that.
1) ToP is overrated if it does not correlate to points scored. Giving up ToP is not really that abd for a D if they do not give up points. Conversely, controlling ToP is not great for the O either if they do not score. A team that runs more will tend to have better ToP than a team that passes more.
2) ToP is a two way street. The D has a role, but so does the O. IMO, the O has a bigger role, because the O has the ball when the play starts. The D should be reacting to them. The D could guess what the O is doing, and the key is keeping the D guessing and making the D pay dearly for a wrong guess.
ToP itself is not that significant, but average ToP per drive does have a mathematical relationship to the number of drives in the game. It's a sixty minute game, so if average ToP per drive is longer, then the number of drives will be fewer. And the number of drives places a limit on scoring chances, and the fewer scoring chances there are, the lower the point total is likely to be.
I think a defense should strive to give up short drives all the time, short in yards particularly, but also in time. The goal is to get the other team off the field as quickly as possible and allow them to advance the ball as little as possible. While preventing scores is most important, giving up yards correlates well with giving up points. Defensive ToP per drive does not correlate as well with either points scored by your offense or points given up by your defense, but giving the ball back to your offense quickly does give them more time to work with. Most of the best defenses do have low defensive ToP per drive. The exceptions (Seattle last year being a notable one ) tend to be good with turnovers or really good in pass defense and therefore get run at a lot.
On offense, ToP per drive actually strikes me as less important, because scoring quickly is often beneficial, especially early in the game and when you're behind, while scoring slowly is beneficial if you have a good lead late in the game or, in some cases, if you are facing a team with a highly explosive offense and want to limit their scoring opportunities. So whether long ToP in a drive is strategically good or bad on offense depends on situation.
Anyway, no single stat is perfect, but you can't look at point totals and just conclude that either the offense or the defense played well or badly based on final scores alone. There are times when a defense plays badly but the opposing offense plays worse, for instance. And there are circumstances, like number of possessions, that do influence scoring totals.