DMN Blog: Football Outsiders: Do Cowboys or Eagles have better cornerback crew?

[ABQcowboyJR]

i too think the dbs are even
i think of it like this

Newman > Samuel
Lito > Henry
Jones = Brown

this is JMO

To think that Jones and Brown are even is a grave mistake. Right now at this very instant in time you might be able to compare them and say they are close. Jones is one of the best corners in football when in playing form. I'm very interested to see how he will do this week.



I'm not such a big fan of Samuel. I'm looking forward to seeing the match up between him and TO when it comes.

 

[Deep_Freeze]

To think that Jones and Brown are even is a grave mistake. Right now at this very instant in time you might be able to compare them and say they are close. Jones is one of the best corners in football when in playing form. I'm very interested to see how he will do this week.



I'm not such a big fan of Samuel. I'm looking forward to seeing the match up between him and TO when it comes.

That matchup won't be interesting, TO will have 2 TDs and Samuel will have that stupid Roy Williams looking face like he's not responsible for the TDs.

As for the corners, they are even right now, but ours should be better by the end of the season with a healthy TNewoman and especially with Pacman back into his normal self.

 

[ABQcowboyJR]

That matchup won't be interesting, TO will have 2 TDs and Samuel will have that stupid Roy Williams looking face like he's not responsible for the TDs.

As for the corners, they are even right now, but ours should be better by the end of the season with a healthy TNewoman and especially with Pacman back into his normal self.

I know. Im counting on a few celebrations

 

[AmishCowboy]

I think you are underrating the Eagles LB's. They are young, big, and fast...this is not the LB's from 1 or 2 years ago.

Bradley could be Special, But I don't think Gocong and Gaither are any thing to write home about.

 

[AdamJT13]

K.C. Joyner, who is the most respected guy when it comes to computing a cornerback's YPA, has Terence Newman at 6.1 YPA for 2007. That was good for 14th in the NFL. The stat is in his new book, Scientific Football 2008.

Joyner has Pacman at 5.4 YPA in 2006, like the article states. That was good for eighth in the NFL according to Joyner.

Why does he have the Joyner number for Pacman in 2006, and Newman's number is not even close to Joyner's number for Newman in 2007?

Whether you agree with Joyner or not, his YPA analysis is the best in the business, and he has Newman at 6.1 YPA in 2007. The 6.1 YPA number for Newman (2007) is in his new book.

This article short changes Newman's YPA for 2007.

Joyner uses only plays that he calls "direct coverage." Football Outsiders counts plays when the defender is close to the receiver. So they're not counting the same types of plays.

Almost a full yard of Newman's YPA from Football Outsiders comes from one play, when Newman bumped Steve Smith and released him into the middle zone, then Smith caught a short pass and ran for a 57-yard gain. That play doesn't have anything to do with Newman's coverage ability, but they're counting it against him because he was closest.

 

[28 Joker]

Joyner uses only plays that he calls "direct coverage." Football Outsiders counts plays when the defender is close to the receiver. So they're not counting the same types of plays.

Almost a full yard of Newman's YPA from Football Outsiders comes from one play, when Newman bumped Steve Smith and released him into the middle zone, then Smith caught a short pass and ran for a 57-yard gain. That play doesn't have anything to do with Newman's coverage ability, but they're counting it against him because he was closest.

AdamJT13,


Thanks for sharing the information, and I agree with you. That explains things for sure. I thought Football Outsiders' YPA number for Newman was way too high if you just watched him play the games. Counting that Steve Smith play would hurt your YPA for sure. I'm thinking like you. I never put that play on Newman, so I didn't consider it a factor.

I wonder why Henry's YPA is lower than Newman's when Henry was the closest defender to Burress on that long TD on opening day. It seems like that play would have hurt Henry's YPA when Football Outsiders analyzed his play.

Joyner's method of "direct coverage" seems to be better parameter for measuring the stat.

 

[JonJon]

What I am more interested in knowing is how the coaching staff compares. I must admit, our talent has been better than that of the Eagles for a few years now, but we have been outcoached. With the new changes we have, who has the advantage now?

 

[DaBoys4Life]

this guy sounds like an eagles fan

 

[ThreeSportStar80]

I would say that by the end of the year, Dallas will have the better trio...

 

[Nexx]

QB Cowboys
OL Cowboys
DB Eagles
Running Back Eagles
WR Cowboys
Dline Cowboys
Linebackers Cowboys
Coaching Eagles

 

[msreddy999]

I would say QBs are pretty much even. When McNaab is healthy (and he is right now) he is as good as any not named Brady and Payton

 

[Doomsday]

The most surprising metric we can talk about with the Cowboys, though, is the "Stuffed" stat. That tracks the percentage of runs that gain either zero or negative yards on first down, or less than 25% of the yards needed for another first down on subsequent downs. The Cowboys were "stuffed" 26% of the time, making them 24th in the league.

The Julius Jones factor! It will be interesting to see how much that number changes with MB3 and Felix Jones carrying the load this year.

Newman > Samuel
Lito > Henry
Jones = Brown

Even if we concede that the Eagles top 3 are slightly better as a group then our top 3, there is no argument that Scandrick and Jenkins are better then Hanson and Graham.

Having the ability to slide Henry over to cover TEs and slide Newman over to the slot without leaving yourself void of talent on the outside is huge and something no other team in the league has the depth to do.

 

[LittleBoyBlue]

QB Cowboys
OL Cowboys
DB Eagles
Running Back Eagles
WR Cowboys
Dline Cowboys
Linebackers Cowboys
Coaching Eagles

I think that the Eagles players execute better at times. I dont think we get outcoached. Reid gets too much credit. Johnson just send the blitz and he gets credit for outcoaching??

 

[Skinsmaniac]

Football Outsiders is a complete joke. For instance a first down play that gains 4.5 yards is awarded one "success point," while one that gains 4.49 yards is awarded zero success points. This makes absolutely zero sense. Success should be measured over a continuum, not arbitrary benchmarks, but of course the illiterates at FO have no idea how to do this. Gaining twenty yards on two plays is worth six points while doing it on one play is worth four. Now, perhaps gaining those yards in two plays is better because it chews up more clock, but is there any reason to make it 150% better? That's completely arbitrary. Please, please for the love of our education system, don't listen to these fake stats and the snake oil salesmen who peddle them.

 

[AdamJT13]

Football Outsiders is a complete joke. For instance a first down play that gains 4.5 yards is awarded one "success point," while one that gains 4.49 yards is awarded zero success points.

I disagree with some of Football Outsiders' stuff, too, but your criticism is misguided and/or inaccurate.

First of all, it's not their statistic. It's from a book that was published in 1988.

Secondly, plays don't gain "4.5" or "4.49" yards. They gain 4 or 5, and nothing in between.

And lastly, they award fractions of points for the yards gained, so your example wouldn't be true. Here's how they explain it: "We then expand upon that basic idea with a more complicated system of success points. A successful play is worth one point, an unsuccessful play zero points. Extra points are awarded for big plays, gradually increasing to three points for 10 yards, four points for 20 yards, and five points for 40 yards or more. There are fractional points in between. (For example, eight yards on third-and-10 is worth 0.63 success points.) Losing four yards is -1 point, losing 12 yards is -1.8 points, an interception is -6 points, and a fumble is worth anywhere from -1.70 to -3.98 points depending on how often a fumble in that situation is lost to the defense - no matter who actually recovers the fumble. Red zone plays are worth 20 percent more, and there is a bonus given for a touchdown."

This makes absolutely zero sense. Success should be measured over a continuum, not arbitrary benchmarks, but of course the illiterates at FO have no idea how to do this.

Are you sure about that? The paragraph I posted above says they do.

Gaining twenty yards on two plays is worth six points while doing it on one play is worth four. Now, perhaps gaining those yards in two plays is better because it chews up more clock, but is there any reason to make it 150% better? That's completely arbitrary.

You do realize that it's used for a per-play average, don't you?

Four points on one play = 4.0 per play.

Six points on two plays = 3.0 per play.

So, doing it on one play is 33.3 percent better.

 

[Temo]

AdamJT13,


Thanks for sharing the information, and I agree with you. That explains things for sure. I thought Football Outsiders' YPA number for Newman was way too high if you just watched him play the games. Counting that Steve Smith play would hurt your YPA for sure. I'm thinking like you. I never put that play on Newman, so I didn't consider it a factor.

I wonder why Henry's YPA is lower than Newman's when Henry was the closest defender to Burress on that long TD on opening day. It seems like that play would have hurt Henry's YPA when Football Outsiders analyzed his play.

Joyner's method of "direct coverage" seems to be better parameter for measuring the stat.

Henry had a lot more attempts at him last year, so one play does not hurt him. Newman had one of the least % targeted numbers in the league, plus he didn't play in all the games. So that one pass really hurts his numbers.

 

[Skinsmaniac]

I disagree with some of Football Outsiders' stuff, too, but your criticism is misguided and/or inaccurate.

First of all, it's not their statistic. It's from a book that was published in 1988.

Secondly, plays don't gain "4.5" or "4.49" yards. They gain 4 or 5, and nothing in between.

And lastly, they award fractions of points for the yards gained, so your example wouldn't be true. Here's how they explain it: "We then expand upon that basic idea with a more complicated system of success points. A successful play is worth one point, an unsuccessful play zero points. Extra points are awarded for big plays, gradually increasing to three points for 10 yards, four points for 20 yards, and five points for 40 yards or more. There are fractional points in between. (For example, eight yards on third-and-10 is worth 0.63 success points.) Losing four yards is -1 point, losing 12 yards is -1.8 points, an interception is -6 points, and a fumble is worth anywhere from -1.70 to -3.98 points depending on how often a fumble in that situation is lost to the defense - no matter who actually recovers the fumble. Red zone plays are worth 20 percent more, and there is a bonus given for a touchdown."



Are you sure about that? The paragraph I posted above says they do.



You do realize that it's used for a per-play average, don't you?

Four points on one play = 4.0 per play.

Six points on two plays = 3.0 per play.

So, doing it on one play is 33.3 percent better.

Hi Adam, as I see it, any misunderstandings I might have had about FO do not change the fact that their methods are bogus. First, whether or not it is originally their statistical method, it is the basis of their system, and thus they should be criticized for any methodological flaws. Secondly, football teams don't gain precisely 4 or 5 yards, so any statistical system that confines itself to those numbers is inherently flawed. My point about arbitrary benchmarks is still valid. Here is what they say on their website:
"On first down, a play is considered a success if it gains 45 percent of needed yards; on second down, a play needs to gain 60 percent of needed yards; on third or fourth down, only gaining a new first down is considered success."
Therefore on the usual first and ten, 4.5 yards is the benchmark. Their "expanding" upon the bogus statistics of the book only makes their system more arbitrary. Can anyone give a valid defense of why gaining eight yards on 3rd and 10 is worth .63 points? Or why losing 4 yards is worth -1 point and losing 12 is -1.8? These figures are pulled from you know where. Why is a redzone play worth 20% more than a regular play? Why not 21% or 30%? What is statistically significant about the 20 yard line? Why is a play that begins on the 20 yard line worth 20% more than a play that begins on the 20.5 yard line? It's complete nonsense, and an embarrasment to anyone with a rudimentary knowledge of statistics. I myself am no statistician, but I know enough to call B.S.

Additionally, I would like to point out that the founder of FO has admitted to never taking a statistics class. Not even in high school. This is not some Princeton mathematician we are talking about here.

 

[Temo]

Secondly, football teams don't gain precisely 4 or 5 yards, so any statistical system that confines itself to those numbers is inherently flawed. My point about arbitrary benchmarks is still valid.

Really? So I'll turn on ESPN on Tuesday and it'll say that Romo had 341.28 yards for the day? If you're going to put down FO's stats for not having incremental yardage, then I suppose you don't ever use official NFL numbers either?

And the part about how success points are determined... read "The hidden game of football".

I'm not saying FO is perfect though... the best football statistical research site on the web (though not as data-heavy as FO or some others) is http://www.advancednflstats.com/

And this is coming from a guy who majored in statistics.

 

[Apefist]

I think the Eagles have the edge in the secondary until the Cowboys prove otherwise. Dallas drafted a CB and signed another big free agent, but as a unit, they are unproven, a pedigree backfield until they shut down McNabb and his wideouts and Westbrook out of the backfield tonight, and then do as good if not better job shutting down the rest of the NFL this season. I believe they can do this. The Eagles secondary last season did well in the second game, but got eaten alive in the first meeting in '07. Newman has to improve his health, Jones needs to hit his stride and get some confidence, and Jenkins needs to get some pro experience. Hamlin is solid, and the experiment to replace Roy Williams on obvious passing situations seemed to have worked against Cleveland. Scandrick and Davis are solid nickel players. I anticipate the Eagles will show some no-huddle to trap Williams on the field and exploit his pass defense with a mismatch with Westbrook or a wideout.

The Eagles have regularly been the statistical elite secondary in the NFC for the past several years, even when other squads on their team have broken down. This could be the year the torch is passed to Dallas. The pieces are in place for it.

 

[AdamJT13]

Secondly, football teams don't gain precisely 4 or 5 yards, so any statistical system that confines itself to those numbers is inherently flawed.

All yardage in the NFL is measured in whole yards, so I guess any statistical system using NFL stats is inherently flawed. That's good to know.

My point about arbitrary benchmarks is still valid. Here is what they say on their website:
"On first down, a play is considered a success if it gains 45 percent of needed yards; on second down, a play needs to gain 60 percent of needed yards; on third or fourth down, only gaining a new first down is considered success."
Therefore on the usual first and ten, 4.5 yards is the benchmark. Their "expanding" upon the bogus statistics of the book only makes their system more arbitrary. Can anyone give a valid defense of why gaining eight yards on 3rd and 10 is worth .63 points? Or why losing 4 yards is worth -1 point and losing 12 is -1.8? These figures are pulled from you know where. Why is a redzone play worth 20% more than a regular play? Why not 21% or 30%? What is statistically significant about the 20 yard line? Why is a play that begins on the 20 yard line worth 20% more than a play that begins on the 20.5 yard line?

Some of your points are valid. (I don't think any of their stuff is "arbitrary," by definition, though.) Like I said, I don't agree with a lot of their stuff, either. And I have other criticisms of their work, too.