The 2009 All-WPA Team

The day after the Pro Bowl rosters are announced there are the obligatory "snub" articles in local papers around the country. In the DC area, the annual London Fletcher snub article is simply reprinted from last year's Post. So what about those Vincent Jacksons and Cedric Bensons who were unfairly left off the roster in favor of big name stars who may not have had a particularly good year? Who really earned their ticket to Miami?

I'll compare players using two different stats. Win Probability Added (WPA) measures each play's increase or decrease in a team's chances of winning. For every play that a player is mentioned in the play-by-play description, including penalties, turnovers and everything else, the WPA is tallied in his name. WPA is a narrative stat. It tells the story of what happened and is very context-dependent. It measures performance when it matters most. It has limited applications in terms of predicting future player performance, but it my mind it's perfect for comparing Pro Bowl and MVP contenders--even Hall of Fame candidates once there's enough data.

Expected Points Added (EPA) measures the progress toward scoring for each play. Every down, distance, and distance to go combination has an expected net point value. Every play either increases or decreases a team's expected points by a particular amount. It's less context-dependent than WPA, so it provides a nice balanced look at a player's contribution.

Unfortunately, with rare exceptions, we can only measure WPA and EPA for the positions that are typically mentioned in the play-by-play description: QB, RB, WR, and TE. Kick returners, kickers, and punters can also be done, and I'll look at a couple selected KRs, but I'll save kickers and punters for another time.

The tables below list both Pro Bowl selectees plus some of the higher profile 'snubs,' many of whom were suggested by readers. This is also an update on the MVP race, but it really isn't much of a race as you'll see below. The actual Pro Bowl selectees are in bold. It's pretty clear from this list who the biggest snubs are and who doesn't belong, so I won't belabor the analysis. I'll leave everyone to their own conclusions.



Position
Player
WPA
EPA
QB
P.Manning
9.8
185.1
QB
P.Rivers
7.6
190.1
QB
D.Brees
7.2
135.0
QB
A.Rodgers
6.6
116.5
QB
B.Favre
5.7
131.2
QB
T.Brady
4.6
118.8
QB
M.Shaub
7.2
125.8
QB
B.Roethlisberger
5.9
68.7
QB
T.Romo
5.6
70.5
QB
D.McNabb
4.3
52.1
RB
A.Peterson
1.7
2.2
RB
D.Williams
1.7
0.9
RB
C.Johnson
1.7
29.4
RB
M.Jones-Drew
1.6
-0.6
RB
R.Rice
0.7
14.0
RB
S.Jackson
0.6
-34.3
RB
F.Gore
1.8
7.0
RB
R.Grant
1.1
-1.1
RB
C.Benson
1.1
-32.8
RB
R.Williams
0.9
-14.0
RB
T.Jones
0.7
-10.3
RB
S.Slaton
0.2
-14.6
TE
D.Clark
2.3
47.3
TE
A.Gates
2.3
78.0
TE
V.Davis
1.7
38.4
TE
J.Witten
1.2
29.0
TE
T.Gonzalez
3.1
48.4
TE
V.Shiancoe
1.2
43.8
TE
B.Celek
1.1
38.6
WR
R.Wayne
3.6
42.8
WR
A.Johnson
3.0
55.5
WR
M.Austin
2.9
64.3
WR
B.Marshall
2.7
59.1
WR
S.Rice
2.0
85.6
WR
L.Fitzgerald
1.6
25.6
WR
R.Moss
1.7
46.9
WR
W.Welker
1.6
59.0
WR
V.Jackson
3.5
89.8
WR
R.White
2.0
22.1
WR
C.Ochocinco
1.5
32.7
WR
S.Smith
1.4
62.3
WR/KR
D.Jackson
1.9
43.2
WR/KR
J.Cribbs
0.3
-8.8
WR/KR
P.Harvin
1.4
54.3



(Note: my previous post on the MVP race should be disregarded due to a bug in my code that caused some plays to be ignored.)

  • Spread The Love
  • Digg This Post
  • Tweet This Post
  • Stumble This Post
  • Submit This Post To Delicious
  • Submit This Post To Reddit
  • Submit This Post To Mixx

16 Responses to “The 2009 All-WPA Team”

  1. Paul says:

    Shouldn't Rivers be in bold? And I knew Matt Schaub should've gotten in. Better than Brady...

  2. Joe G says:

    Good stuff.

    My first guess is that EPA is a more "stable" stat than WPA since a couple game winning TD passes by a QB can really add up like 2 WPA. Agree?

    Is Stephen Jackson at -34 EPA correct?

  3. Anonymous says:

    I wonder if it's possible to do a WPA/LI stat for football. That would seem to be a great stat to have.

  4. Joe G says:

    Also - should this be normalized on a per play basis?

  5. Brian Burke says:

    Jackson's numbers are correct, but I think there is a slight negative bias on the EPA numbers and a slight positive bias on the WPA numbers. It's really small, like .0002 per play or something. But it adds up over lots of plays.

    I can do average per play numbers, but for this application it seems like total accomplishment is most appropriate.

    WPA/LI can be done. I'm not happy with my current LI formula right now. To be honest, I don't understand what it tells us. This is my own failure, not a criticism. I'm just not that familiar.

    I'd agree EPA should be a more stable, repeatable stat.

  6. Brian Burke says:

    The thing with RB and negative EPAs is that RB are called upon to bludgeon themselves for 2 and 3 yard gains too often. Only a modest fraction of runs are actually positive. So paradoxically, the better the RB, the more often he carries the ball, and the more negative plays he accumulates.

  7. rk says:

    My only complaint with WPA is that, and correct me if I'm wrong, if your team immediately gains a large lead or is quickly put in a large hole, you don't have nearly the opportunity to accumulate WPA. I would think this would impact players on poor defensive teams (though I suppose if you are given the chance often to "come back" on offense you have more opportunity to accumulate WPA). Basically, I think WPA, because of its sensitivity to context far outside the players control, doesn't do as great a job of evaluating a player as EPA (of these two). That said, there is a lot to a player who makes a big play when it counts, so having a high WPA shouldn't count against a player, but I don't think a low WPA nails it either. Meh, just some quick thoughts.

  8. rk says:

    And, wow, for some reason my eyes skipped right over it, but Vincent Jackson was a monumental snub, way beyond what I thought was a big snub before looking into it. Jason Witten should never have been selected based on this.

  9. Brett says:

    Steven Jackson's EPA doesn't surprise me. He has only 4 touchdowns on 355 touches, by far the lowest rate of any major RB. He also has an average YPC and below average 1st down conversion rate. As Brian said, all the small plays add up negatively with no big plays to offset them. It probably says a lot more about the team than the individual in this case.

  10. DSMok1 says:

    These numbers look quite a bit like the numbers I got when I created an advanced stats-based fantasy football scoring system. The idea was to penalize players if they didn't get above about 2.5 yards per play (including targets for WRs) and to (HERE IS A BIG DIFFERENCE) count all fumbles the same whether they were recovered by their team or the other team. Check out the scoring in the league--the link is in my name.

  11. Martin says:

    What do you think of R.Wayne? Best non-QB i WPA, but 4. worst WR (3. if don't count J.Cribbs, or R.White) Is it a lot of tagets, or what?

  12. Brian Burke says:

    It's primarily context. Wayne's big TD catch late in the Jaguar game was worth 0.56 WPA alone. He caught the TD to beat the Patriots, which was worth a lot too. Manning has lots of other targets, which robs Wayne of some EPA. But he's made big catches in critical situations to boost his WPA.

  13. Joe G says:

    When a receiver catches a pass that takes the team's WP from .30 to .70 - do the QB and WR both get the .40 WP added or do they split it and get .20 each?

  14. Brian Burke says:

    The both get the full credit, +0.40.

  15. Michael Beuoy says:

    I wonder if this approach could tell us the converse about defensive backs?

    It would be hard to determine who the really good corners are using pure stats, since qb's simply don't throw in the direction of a good corner. However, I do believe that the play-by-play data indicates who was covering the receiver that made the catch.

    Could you sum up the WPA for each completed pass a cornerback was involved in? And sort in descending order? A sort of WPA Hall of Shame list.

  16. Unknown says:

    I read John Clayton's article about the best quarterbacks in the history of the NFL the other day and it has stuck with me since reading it.

    People will say that it's not about numbers, it's about wins. This ignores good quarterbacks on bad teams. Others will say it's only numbers. Others will say it's about championships.

    Thinking about that, I tried to get all smart today and come up with a better way to compare quarterbacks. I was thinking about using a formula with passer ratings, winning percentages, and championships. I came up with this:

    (RATING) X [1. + (WP)] X [1.0(#CH WINS)]
    For example, 88.9 x 1.641 x 1.02 = 148.81

    I was so proud of myself but I know nothing about passer rating so I started thinking about receivers and YAC. Then I asked about sacks. I couldn't find the data I wanted so I searched and found your site. I am amazed at the data you have collected and the scrutiny you give it! This is really an awesome job on your part.

    Have you ever statistically concluded who the best QB's of all time are?

Leave a Reply

Note: Only a member of this blog may post a comment.