New Feature: Advanced Stat Team Pages

I’ve added a new page to complement the individual player advanced stats. Until now, there were stats for all defenders, offensive skill players, and player stats by team on both sides of the ball. I’ve recently added a team-level page, which is a summary of each offense and defense in terms of advanced stats. You can always get to them via the 'Stats' menu item in the site header. There’s Win Probability Added (WPA), Expected Points Added (EPA), and Success Rate (SR) for every offense and defense since 2000. Stats for the run and the pass are broken out.

For example, so far in 2010 New England’s offense is back firing on all cylinders, leading the league with 43 EPA. This basically means that the NE offense has generated about a net point advantage of 43 points over three games. That’s about 14 net points per game. Everyone knows about their passing success, but they’re running the ball very well too. They’re 3rd in the league with about a 50% SR in the running game.

The much-maligned Jets offense has actually outplayed its heralded defense so far this year. The offense ranks in the top 10 in terms of EPA, and the defense is rather mediocre at +4 EPA. (Negative numbers represent better defenses.)

Despite their 1-2 start, the San Diego defense ranks 1st in EPA and 2nd in SR. San Diego’s woes stem from offensive turnovers and poor special teams play.

The Kansas City defense is for real, at least so far. They 4th in the league in terms of EPA, right there sandwiched between Baltimore and Pittsburgh. The Chiefs success isn’t just from random big-play turnovers. They’re winning battles consistently, topping the league with a 60% defensive SR.

I could go on and on. There’s a ton of analysis to be done with these stats. We can also go back a decade and look at some fun things. Was the 2000 Titans defense really better than the legendary Ravens defense that year? How do they compare to the 2002 Buccaneers D? Which squad was “better,” the 2000 Ravens defense or the 2001 St. Louis offense? What about the 2007 Patriots offense?

One thing to note is that these stats are not adjusted for opponent strength, something that’s on my list of projects for the year. Opponent adjustments are particularly important early in a season. It’s a fairly simple and obvious thing to do, but it’s a complex task in terms of programming and database management.

  • 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

12 Responses to “New Feature: Advanced Stat Team Pages”

  1. Dave says:

    Great stuff. Thanks for the work.

  2. xtimmygx says:

    Hey Brian,

    First off, love the site. Didn't really know where to put this so I'll put it here. A couple of questions about the Jets-Dolphins game.

    1) I know it didn't matter in the end but as the Jets were driving and getting ready to score I was thinking to myself that they would definitely go for 2 if they did score. However, the announcers barely even mentioned it and it didn't seem to be something that Rex Ryan really considered. Did they make the right choice on this?

    2) I was looking at the WP graph for the game it seems that after Brandon Marshall got pushed out at the 11 yd line the Jets winning pct was 99%. That seems way to high to me. I don't know the odds of scoring a td there, but we can say its 50% (seems reasonable off the top of my head since teams convert 1st downs 67% of the time), 2 pt conversion is another 50% or so. That means the odds of converting both of these is 25%. Then the odds are basically even because most likely the game goes to overtime. This means to me that the WP should have been much closer to ~87%.

    Sorry for the long post.

  3. Martin says:

    How about the Broncos going for it on a 4th down in the 4th q, down by about 12? I said go for it, by the announcers in Denmark said kick it. What does WP say?

  4. coldbikemessenger says:

    Is it better to rank teams by EPA or WPA?

  5. Brian Burke says:

    EPA is going to be more predictive. EPA per play might be best. WPA is going to be a measure of exactly how past games unfolded.

    My feeling is to not have "one stat to rule them all" but to provide tools that look at performance from different perspectives. Once people have an intuitive grasp of what each stat is, they can use whichever one makes the most sense for the question at hand.

  6. Adam Davis says:

    Something appears to be broken on your team pages. On both offense and defense, the Jaguars appear to suck.

    Oh, wait...

  7. Brian Burke says:

    Very funny, Adam!

  8. coldbikemessenger says:

    "The proportion of plays in which a player was directly involved that would typically be considered successful. Specifically, SR is the percentage of plays resulting in positive Expected Points Added (EPA)."

    "Everyone knows about their passing success, but they’re running the ball very well too. They’re 3rd in the league with about a 50% SR in the running game."

    Dallas is over 50% success rate as well running the ball, yet the Cowboys run EPA is much lower.
    I am not sure what what the run and pass success rates are and how they relate to EPA?

  9. coldbikemessenger says:

    Also wanted to say I am really enjoying these charts.

  10. Brian Burke says:

    CBM-Thanks. Success Rate is a derivative of EPA. If a play yields a positive EPA it's considered a success. So in DAL's case, with a high SR but low EPA, it probably means they have been grinding the ball successfully but without breakaway gains. It's also likely they've had some bad luck with fumbles.

  11. Sampo says:

    Very nice, Brian!

    Is there another way to get to the page other than this post? I can't find it from the dropdown-menus.

  12. Brian Burke says:

    Thanks. I haven't had time to update the menus, but I'll get to it.

Leave a Reply

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