Kasmir

I thought it might be instructive to look a Patriot WR statistical productivity year to year for comparative purposes, given the interest in the debate. It will come as no surprise that I continue to use Football Outsider's statistics.

http://www.footballoutsiders.com/stats/wr.php

Specifically, in this case I'm using their DPAR measure, as measures total point productivity above replacement across the season, vs DVOA, which measures average performance per play. You can add total WR DPAR together to get a team WR DPAR, but you can't add WR DVOA without weighting it ny passes attempted. Also, DPAR is in points, which is a useful measure.

Here are the summary total WR DPAR results:

2006 32.6
2005 60.5
2004 61.6
2003 50.0
2002 22.4
2001 42.6

So perhaps to the surprise of some, the statistics support the eyeballs that see this year's WR productivity as being weak comparatively, the weakest of any year except 2002, the one non-playoff year the the bunch.

The individual data (see below) shows that the bulk of this fall off being caused by losing the contribution of our top WR (Branch) and not replacing it. Our #1 this year (Caldwell) is producing at a #2 level. What's worse, our #2 (Brown) is not producing at the #3 levels of the past 3 years. So it's as if we lost our #1 and #3 receivers.

The numbers suggest retaining Branch might have given us nearly the same productivity as 2005/4. Retaining Givens alone might have restored us to 2003 numbers. Retaining both and adding Caldwell might have given us the best WR we had in this period.

Notes:

2006 is projected, i.e. I used 16/14 of the present 28.6, i.e. I assume that performance to date will project for for the last two games

Playoff performance is not included. This is the only choice to get a baseline for comparison across the years involved.

WR's with less than 10 catches are not included.

WR rushing, passing or teams contribution not included, even though this can be measurable David Patten in 2001 contributed 8.6 pts of value receiving and 4.7 more rushing.

Football Outsider's caveat's apply particularly:
Also synergies with TE's, the running game, defense, and special teams are not included, although do note that the FO system is based on contribution per play, so many of these factors are indirectly included: WR whose NFL stats are padded by 7 yard catches on 3-10 will not be credited much in this system, for instance.

Individual Data: (name, DPAR)

2006 (14 games)
-----
Caldwell 13
Brown 5.3
Jackson 5.0
Gabriel 3.9
Gaffney 1.4

2005
-----
Branch 27.5
Givens 12.8
Brown 11.6
Dwight 8.0
Davis 0.6

2004
-----
Givens 21.4
Patten 18.6
Branch 16.7
Brown 3.6
Johnson 1.3

2003
-----
Branch 19.3
Givens 18.0
Brown 14.8
Patten 0.8
Ward 0.6
Johnson -3.5

2002
-----
Patten 9.0
Branch 8.2
Brown 5.2
Givens -0.6
Hayes -3.8

2001
-----
Brown 31.0
Patten 8.6
Glenn 5.7
Johnson -2.7

PATRIOTSFANINPA

Wow Branch Played For Us In 2006 5.3? . . Must Have Missed That!

Kasmir

Brown of course, tx for the catch.

PATRIOTS-80

Is there any way to compare what our TEs have done as opposed to later years? For a while there Watson was our leading receiver.

Thanks for all your stats. They really are enlightning.

Gumby

Nice research.

I guess the only stat that really surprised me in that is Brown's #. He has been such an consistent part of the offense and always seemed to be coming up with catches at critical points (not so much this year; but still every once in a while).

Do they provide those ratios broken out by game as well? It would be interesting to see if we are showing any tendancy to improve as the year progresses. If so then there may be some optimism for the playoffs.

But right now it seems we will live and die with TEs and RBs in playoffs as we have during the year.

Kasmir

Since FO compiles play-by-play values, they must have game breakdowns, but I don't see them on the site.

Their team DVOA roll-ups do show results both unweighted and with a greater weight for the most recent 8 games. By that measure, we're slightly better now as a team than we were earlier in the year, but a large part of that is that our opposition (AFC East, etc) is better than people thought.

As for Troy, the FO stats are not that kind to him even in his best years. Even though his 2001 was the most productive single year (in this period) for a Patriots receiver, his DVOA (value per play) wasn't that great that year (17.3%) -- Terry Glenn actually had the highest DVOA (20.4%) on the team that year. OF course, Glenn's 5.7 DPAR contribution didn't get him a ring; I sure hope Charles Johnson's -2.7 contribution also didn't get him one!

The shocker for me was actually Troy's 2002 numbers. Even though he had 891 yards that year, he had only a 5.2 DPAR. In other words, FO claims Chad Jackson has contributed more to the success of the 2006 team than Troy's receiving did to the 2002 team ! Brown rated a miserable -9.3% DVOA despite being targeted for 141 passes. I don't know how this compares to the "eyeball" memory of that season, as the 2002 season is now a painful blur for me.

There are a lot of interesting nuggets buried in this data. I didn't post the DVOA's because it would have cluttered things up too much, but beware that there is sometimes a big discrepancy between DPAR and DVOA; for example Branch was 3rd in DPAR in 2004 because he was injured, but he was by far our best receiver in DVOA (36.2%).

There is also a tendency for regression to the mean, particularly with receivers who have high DVOA one year on not many attempts. Givens was fantastic in 2003 on a DVOA basis with 55.9%, which was 3rd in the league (and would be first in many years), but he regressed when targeted more (105 times vs 55) in the next year (and perhaps because Branch was hurt so much), but he ended up leading the team in contribution (DPAR) for the only time, even though his performance per play (DVOA) fell to 16.6%.

Of course all statistical approaches have anomalies, but the FO system seems pretty good and has helped my understanding of the game. Again, you can't eyeball everything, you can't remember everything you see, and sometimes you see and remember only the things you want to. Stats don't have those subjectivity problems.

Kasmir

I'll take a look. Again, they're not my stats, they're Football Outsiders, who have really marvelously innovated in football statistics. All I'm doing is collating them for our interest.

Kasmir

Tight End Statistics.

http://www.footballoutsiders.com/stats/te.php

Summary:
2006 8.9 DPAR, 144 att
2005 12.8 DPAR, 96 att
2004 24.1 DPAR, 80 att
2003 15.7 DPAR, 107 att
2002 13.1 DPAR, 84 att
2001 3.2 DPAR, 29 att

Perhaps surprisingly, TE production is also significantly down, despite being featured a lot more. Both Watson and Graham have negative DVOA, i.e. they're averaging below replacement level per play (recall that DVOA/DPAR penalizes fumbles). Perhaps this is a negative synergy with the WR production problem, or perhaps there's another systemic problem (Brady having an off year?). But TE performance is not a strong point according to FO stats.

Individual breakdowns:

2006
-----
Watson 5.5 DPAR, -2.8% DVOA, 91 att
Thomas 1.3 DPAR, 7.5% DVOA, 11 att
Graham 1.0 DPAR, -7.9% DVOA, 24 att
-------
Total 7.8 DPAR (8.9 annualized), 126 att (144 annualized)


2005
-----
Watson 8.5 DPAR, 12.2% DVOA, 54 att
Graham 4.4 DPAR, 17.4% DVOA, 25 att
Fauria -0.1 DPAR, -12.8% DVOA, 17 att
-------
Total 12.8 DPAR, 96 att


2004
-----
Graham 13.6 DPAR, 32.5% DVOA, 48 att
Fauria 8.0 DPAR, 42.3% DVOA, 20 att
Weaver 2.5 DPAR, 16.8% DVOA, 12 att
------
Total 24.1 DPAR, 80 att


2003
-----
Fauria 8.0 DPAR, 22.1% DVOA, 45 att
Graham 6.7 DPAR, 8.0% DPAR, 62 att
------
Total 15.7 DPAR, 107 att


2002
-----
Fauria 11.1 DPAR, 30% DVOA, 40 att
Cleeland 1.4 DPAR, -1.7% DVOA, 21 att
Graham 0.6 DPAR, -8.9% DVOA, 23 att
-----
Total 13.1 DPAR, 84 att


2001
-----
Wiggins 3.7 DPAR, 13.5% DVOA, 21 att
Rutledge -0.5 DPAR, -20.9% DVOA, 8 att
------
Total 3.2 DPAR

edzo44

So just out of curiousity, for 2006, what were Branch and Givens numbers with their new teams? Better or worse?

__________________
We've had a bunch of practices. We're going out there everyday. They're tired. They're sore. There's a little bit of information overload, but it is pro football. Nobody ever said it was going to be easy."

- - - Bill Belichick

Kasmir

I covered that in another thread, but here it is again:

I should also add that Branch has a DPAR of 6.4, Givens -2.2.


http://208.109.107.176/new-england-p...t=46535&page=6