Pick6TerenceNewman
Benched
- Messages
- 2,665
- Reaction score
- 0
Five Downs With Football Outsiders: Miles Austin is the NFL's most effective WR this season
8:00 AM Fri, Oct 03, 2008 | Permalink | Yahoo! Buzz
Tim MacMahon http://www.***BANNED-URL***/blogs/images/email-icon.jpg E-mail http://www.***BANNED-URL***/blogs/images/email-icon.jpg News tips
http://cowboysblog.***BANNED-URL***/NS_21CowsI-thumb-250x169.jpg
Bill Barnwell, a contributing editor at footballoutsiders.com and one of the minds behind the must-read Pro Football Prospectus, will answer five Cowboys questions each week of the season using the FO data and methodologies. Here's the Week 5 edition:
1. Miles Austin ranks No. 1 among receivers right now in your DVOA formula. Can you explain what that means and what, if any, conclusions can be drawn from Austin's ranking?
What it means is that Austin, relative to every other receiver in the league, has done the most with the chances thrown in his direction.That performance is adjusted relative to opponent, down, distance, and situation; it doesn't take into account differences in coverage or the quarterback or offensive line he plays with.
Does that make it a flawed stat? Yes and no. Obviously, no one's saying that Miles Austin's the best wide receiver in the league, and they shouldn't. It's a much more useful stat, though, than saying "Miles Austin has 171 yards". Antonio Bryant has 225 yards, but he's been thrown 37 passes; Austin's only been thrown 12. Austin's been far more effective, and DVOA bears that out.
A good comparison to Austin might be Patrick Crayton, even though they're very different stylistically. In 2006, Crayton had the highest DVOA in the league as the Cowboys' third wide receiver. It didn't mean that Crayton was the best receiver in the league, but it was a sign that the Cowboys could feel comfortable expanding his role in the offense, which they did when Terry Glenn went down. If Austin can keep up this level of performance for the entire season, I'd say that we could say the same about him.
2. Is there any statistical evidence at your disposal linking number of rushing attempts to wins?
Actually, it's ironic that you ask that -- the first article ever written on Football Outsiders, the website's raison d'etre, was analyzing a Ron Borges article which claimed that the '02 Patriots failed to win because they didn't establish the run.
There's a very strong correlation between rushing attempts and wins, but it's because teams are running the ball in the fourth quarter to run the clock out, not because of their performance running the ball in the first half. There's no relationship between number of carries or even success running the ball in the first half and winning.
Last week, after the loss to the Commanders, Bradie James said "They were committed to the run. They wanted to establish the run game. That's what they did. By allowing them to run the ball, we allowed them to make plays and stick around." This led to at least one article about how the Commanders' ability -- and dedication -- to running the ball on the Cowboys led to their success on Sunday.
It's not really the case. In the first half of each of the first three Cowboys' games, Dallas' opponents averaged just over 11 rushing attempts, gaining 3.7 yards per carry. On Sunday, the Commanders ran the ball 13 times, averaging 3.8 yards per carry.
In the second half of those games, Dallas' opponents (each of whom trailed for a good portion of the second half, unlike the Commanders) averaged just under nine rushing attempts, gaining 4.8 yards per carry. The Commanders ran the ball 21 times, averaging 5.4 yards per carry.
I don't think those two extra attempts in the first half had a lot to do with the Commanders winning that game.
3. How many receivers have been targeted 15 or more times in a game this season? How does T.O.'s production against the Commanders compare to the other receivers who were thrown that many passes?
Owens is one of eight guys to get 15 or more targets this year; his performance would be seventh, regardless of what metric you choose to use.
Using DYAR, our counting metric which takes into account the same context that DVOA does, T.O had -20.1 yards above a replacement player; in other words, if you'd taken a player off the street and stuck him in the lineup on Sunday, you'd have expected him to pick up 20 more yards than Owens. That's seventh-worst out of the eight receivers; only Muhsin Muhammad's Week 1 game as worse, at -41.9. The best was T.J. Houshmandzadeh's performance against the Giants in Week 3, which rated out as 58.3 yards above replacement.
If we want to use traditional metrics, Owens' performance was again far below average. He had only seven catches; only Muhammad was worse. He had 71 yards, again, only Muhammad was worse.
By the way, T.O.'s wrong; the teams relying so heavily on one targets are 3-5 this year. Last year, they were 9-28. If you're going to one guy fifteen times, it's not because he's a great player; it's because you're losing and you need to catch up.
4. How bad from a statistical standpoint is the Bengals' offensive line?
Let's go with the good news first. Their pass blocking has only been below-average; when you account for the quality of the defensive lines and the number of dropbacks they've faced, their Adjusted Sack Rate is only 23rd in the league.
The run-blocking...it's not pretty. They're last in the league in Adjusted Line Yards, which we explained earlier in the year. And it's not by a small margin. 31st place is Green Bay, and they're closer to 22nd place than they are to the Bengals.
Across the board, they've been dire. They're 31st in power situations, 29th in gaining 10+ yards, and 31st in frequency of getting stuffed. It's sad to think that this was an above-average line two years ago.
5. The Cowboys are 29th in FO's defensive line rankings. What goes into these rankings, and why are the Cowboys so low?
The defensive line rankings are exactly the inverse of the offensive line rankings we discussed in a previous edition. We're essentially attempting to isolate how a defensive line does versus the run, not how the whole team performs, by removing all the stuff that happens away from the defensive line and adjusting accordingly.
Last year, Dallas was eighth in the league, allowing 3.93 Adjusted Line Yards per carry. They were particularly effective at stuffing opponent, with 28% of opposition carries on first down gaining zero or no gain, and carries on second through fourth down gaining less than one-fourth of the yards needed for another first down. That was seventh in football.
This year, they're allowing 4.79 Adjusted Line Yards per carry. Teams have been successful 88% of the time in power situations (third and fourth down with less than two yards to go, with runs from inside the Cowboys 2 also considered) against them; that's 31st in the league. They're also only stuffing opponents 20% of the time, which is 26th in the league.
As to why that is? It's hard to say. Although we're beginning to adjust for opponent now, they've played some pretty impressive offensive lines in Cleveland, Philadelphia, Green Bay, and Washington. I think that as the season goes along, that'll creep up to around the league average.
8:00 AM Fri, Oct 03, 2008 | Permalink | Yahoo! Buzz
Tim MacMahon http://www.***BANNED-URL***/blogs/images/email-icon.jpg E-mail http://www.***BANNED-URL***/blogs/images/email-icon.jpg News tips
http://cowboysblog.***BANNED-URL***/NS_21CowsI-thumb-250x169.jpg
Bill Barnwell, a contributing editor at footballoutsiders.com and one of the minds behind the must-read Pro Football Prospectus, will answer five Cowboys questions each week of the season using the FO data and methodologies. Here's the Week 5 edition:
1. Miles Austin ranks No. 1 among receivers right now in your DVOA formula. Can you explain what that means and what, if any, conclusions can be drawn from Austin's ranking?
What it means is that Austin, relative to every other receiver in the league, has done the most with the chances thrown in his direction.That performance is adjusted relative to opponent, down, distance, and situation; it doesn't take into account differences in coverage or the quarterback or offensive line he plays with.
Does that make it a flawed stat? Yes and no. Obviously, no one's saying that Miles Austin's the best wide receiver in the league, and they shouldn't. It's a much more useful stat, though, than saying "Miles Austin has 171 yards". Antonio Bryant has 225 yards, but he's been thrown 37 passes; Austin's only been thrown 12. Austin's been far more effective, and DVOA bears that out.
A good comparison to Austin might be Patrick Crayton, even though they're very different stylistically. In 2006, Crayton had the highest DVOA in the league as the Cowboys' third wide receiver. It didn't mean that Crayton was the best receiver in the league, but it was a sign that the Cowboys could feel comfortable expanding his role in the offense, which they did when Terry Glenn went down. If Austin can keep up this level of performance for the entire season, I'd say that we could say the same about him.
2. Is there any statistical evidence at your disposal linking number of rushing attempts to wins?
Actually, it's ironic that you ask that -- the first article ever written on Football Outsiders, the website's raison d'etre, was analyzing a Ron Borges article which claimed that the '02 Patriots failed to win because they didn't establish the run.
There's a very strong correlation between rushing attempts and wins, but it's because teams are running the ball in the fourth quarter to run the clock out, not because of their performance running the ball in the first half. There's no relationship between number of carries or even success running the ball in the first half and winning.
Last week, after the loss to the Commanders, Bradie James said "They were committed to the run. They wanted to establish the run game. That's what they did. By allowing them to run the ball, we allowed them to make plays and stick around." This led to at least one article about how the Commanders' ability -- and dedication -- to running the ball on the Cowboys led to their success on Sunday.
It's not really the case. In the first half of each of the first three Cowboys' games, Dallas' opponents averaged just over 11 rushing attempts, gaining 3.7 yards per carry. On Sunday, the Commanders ran the ball 13 times, averaging 3.8 yards per carry.
In the second half of those games, Dallas' opponents (each of whom trailed for a good portion of the second half, unlike the Commanders) averaged just under nine rushing attempts, gaining 4.8 yards per carry. The Commanders ran the ball 21 times, averaging 5.4 yards per carry.
I don't think those two extra attempts in the first half had a lot to do with the Commanders winning that game.
3. How many receivers have been targeted 15 or more times in a game this season? How does T.O.'s production against the Commanders compare to the other receivers who were thrown that many passes?
Owens is one of eight guys to get 15 or more targets this year; his performance would be seventh, regardless of what metric you choose to use.
Using DYAR, our counting metric which takes into account the same context that DVOA does, T.O had -20.1 yards above a replacement player; in other words, if you'd taken a player off the street and stuck him in the lineup on Sunday, you'd have expected him to pick up 20 more yards than Owens. That's seventh-worst out of the eight receivers; only Muhsin Muhammad's Week 1 game as worse, at -41.9. The best was T.J. Houshmandzadeh's performance against the Giants in Week 3, which rated out as 58.3 yards above replacement.
If we want to use traditional metrics, Owens' performance was again far below average. He had only seven catches; only Muhammad was worse. He had 71 yards, again, only Muhammad was worse.
By the way, T.O.'s wrong; the teams relying so heavily on one targets are 3-5 this year. Last year, they were 9-28. If you're going to one guy fifteen times, it's not because he's a great player; it's because you're losing and you need to catch up.
4. How bad from a statistical standpoint is the Bengals' offensive line?
Let's go with the good news first. Their pass blocking has only been below-average; when you account for the quality of the defensive lines and the number of dropbacks they've faced, their Adjusted Sack Rate is only 23rd in the league.
The run-blocking...it's not pretty. They're last in the league in Adjusted Line Yards, which we explained earlier in the year. And it's not by a small margin. 31st place is Green Bay, and they're closer to 22nd place than they are to the Bengals.
Across the board, they've been dire. They're 31st in power situations, 29th in gaining 10+ yards, and 31st in frequency of getting stuffed. It's sad to think that this was an above-average line two years ago.
5. The Cowboys are 29th in FO's defensive line rankings. What goes into these rankings, and why are the Cowboys so low?
The defensive line rankings are exactly the inverse of the offensive line rankings we discussed in a previous edition. We're essentially attempting to isolate how a defensive line does versus the run, not how the whole team performs, by removing all the stuff that happens away from the defensive line and adjusting accordingly.
Last year, Dallas was eighth in the league, allowing 3.93 Adjusted Line Yards per carry. They were particularly effective at stuffing opponent, with 28% of opposition carries on first down gaining zero or no gain, and carries on second through fourth down gaining less than one-fourth of the yards needed for another first down. That was seventh in football.
This year, they're allowing 4.79 Adjusted Line Yards per carry. Teams have been successful 88% of the time in power situations (third and fourth down with less than two yards to go, with runs from inside the Cowboys 2 also considered) against them; that's 31st in the league. They're also only stuffing opponents 20% of the time, which is 26th in the league.
As to why that is? It's hard to say. Although we're beginning to adjust for opponent now, they've played some pretty impressive offensive lines in Cleveland, Philadelphia, Green Bay, and Washington. I think that as the season goes along, that'll creep up to around the league average.