Writers of Pro Football Prospectus 2008

11 Sep 2009

Smarter Stats: The Bart Scott All-Stars

Jets linebacker Bart Scott recently contended that high tackle totals mean very little, especially among linebackers. He's preaching to the choir at FO, and this week's Washington Post stat feature reveals 2008's most efficient defenders, as measured in percentage of Defeats.

Posted by: Doug Farrar on 11 Sep 2009

19 comments, Last at 20 Oct 2009, 8:44am by yu hao


by pachoo5 :: Fri, 09/11/2009 - 10:43pm

Just as I thought, Babineaux is so underrated. He is the defensive stud few people talk about.

by TruFalcons (not verified) :: Fri, 09/11/2009 - 11:35pm

Excellent article, thanks for posting.

by Lou :: Sat, 09/12/2009 - 12:56am

cool analysis. i enjoy seeing so many Bears unexpectedly rank highly here. i wonder how much higher briggs is then the next highest 4-3 linebacker. i also wonder the value of having a lineman rank highly here, but have poor sack totals and if these numbers are consistent from year to year.

It would be even better if we could see defeats divided by total plays on the field. i know it would be absurdly difficult to track that with the charting project, but i know one of the chicago sports writers does it (biggs i think) and didn't some bills fan do it a couple of years ago. but maybe you guys could find a way to collect their data along with any other current sportswriter or fansite that tracks it and at least create a partial list.

one quibble, its been a while since i learned about significant figures, but are two decimal places really necessary here?

by SBL (not verified) :: Sat, 09/12/2009 - 2:19am

And unsurprisingly Lamarr Woodley, James Harrison, and Troy Polamalu made the list and hey even Lawrence Timmons gets some love. Go Steelers!

by bubqr :: Sat, 09/12/2009 - 4:22am

Corey Ivy ????


by KJT :: Sat, 09/12/2009 - 4:44am

Jason David was a bit of a surprise.

Hole in Zone could not be reached for comments.

by Gus A. (not verified) :: Sat, 09/12/2009 - 11:11am

Corey Ivy makes sense to those who watched a lot of the 2008 Ravens. He couldn't cover anyone so Rex Ryan used him frequently as a blitzer form the nickel or dime slot cover position.

by Jimmy :: Sat, 09/12/2009 - 1:20pm

It is a good article and a refreshing was to look at defensive production. It is always nice as a fan to see players on your team getting credit and the list only goes to confirm how much better Briggs is than any other 4-3 OLB in football. Nice to see Urlacher getting some props too despite his play not being all that great last year (maybe rumours of his demise are premature, I guess we will see one way or the other soon enough).

However Daniel Manning shouldn't be counted amongst the safeties as he played nickel back for the majority of last year. If he put up that kind of production in the couple of games where he played safety then he is going to win the NFL defensive MVP this year.

by Dice :: Sat, 09/12/2009 - 2:43pm

Surprised to see Finnegan there on the corners list; he's good but not real big, and I noticed the Steelers running his way a lot(and away from mid-30s Harper), so I guess he gets a fair number of chances to get stops/defeats.

by tuluse :: Sat, 09/12/2009 - 3:03pm

This seems like a useless metric for rating corners.

by zlionsfan :: Mon, 09/14/2009 - 1:12pm

Sure, it's a small part of what corners are expected to do, so looking at this list and ranking CBs solely by it isn't helpful ... but then there aren't many (any?) positions where one or two metrics can accurately encompass everything they do.

Even for DL and LB, this doesn't take into account how well they do on plays where they do not make a tackle, and that would be the vast majority of their plays.

This is just a look at the value of the plays that they do make, and one step toward better understanding their contributions as a whole.

by phillyangst :: Sat, 09/12/2009 - 6:40pm

So what is the correlation between Total Defense Tackles and Wins?

"DVOA loves Philadelphia!"

by greybeard :: Mon, 09/14/2009 - 12:40am

Bart Scott is not a Bart Scott All-Star.

by matthewglidden :: Mon, 09/14/2009 - 12:49pm

Removed and reposted below.

by matthewglidden :: Mon, 09/14/2009 - 12:48pm

FO included DVOA in the tackles table. Does anyone have a version of the defeats tables with team DVOA ranks?

by TGT2 (not verified) :: Tue, 09/15/2009 - 12:41pm

Those are all percentages. I think we need a totals list (of defeats) before we can say anything much. If a guy is cleaning up tackles downfield, does that mean his defeats should be devalued? If a middle linebacker drops back in coverage and makes alot of downfield passing tackles, does that mean his line of scrimmage run stops are any less worthy?

Take Freeney. His job was to run up field. Either he gets a defeat or he whiffs completely. Have you ever seen Freeney 2 yards or more downfield after the ball was snapped? I'm surprised his defeat percentage was as low as it was. Are his 14 defeats, but only 13 other plays, really better than another end who gets 20 defeats, but has 30 tackles chasing guys downfield?

Ivy was mentioned, and he's a good example of the other side of the coin. As Gus A. said, Ivy blitzing was a staple of the defense, so he has relatively high defeat numbers, but that's not why he has a large defeat percentage. His percentage is so high because he was so horrible in coverage and so damn slow that receivers just ran away from him. He didn't clean up anyone else's receivers down field, and someone else tackled all the guys that caught balls on him. If he had more speed or could cover decently, he'd have a much worse defeat percentage, even if he blitzed just as much and had just as many defeats.

The best measure would be defeats as a percentage of plays (pipe dream of knowing who's on the field every play). Without that, a list of total defeats is necessary with the percentages. Total defeats is still going to be position biased, but i think it would be a better measure of great plays. Percentages shows great:okay ratio. If you have 10 great plays, 10 okay plays, and 2 million horrible plays, you're still going to look great on the provided ratio of defeats, but you're going to be outstripped by other players on total defeats.

A side note on Bart Scott not being a Bart Scott All-Star. First, in total defeats, he might be, but I doubt it. Second, Scott's job last year was to be the Defensive Tackle of Linebackers. His job was to take on blockers and let Lewis go crazy. Scott essentially had support responsibilities. If you run the defeats number from when Lewis was hurt in previous years, and the Ravens were scheming to free up Scott, and you'll see very different numbers.

by TGT2 (not verified) :: Tue, 09/15/2009 - 12:46pm

Addendum to 16:

That was one of the reasons Scott gave for going to the Jets. Ryan promised to make him the feature linebacker instead of a support linebacker. While Lewis is still in Baltimore, the Ravens couldn't promise that.

Also, thank for the info. I didn't mean to be completely negative. Defeats is a great stat, I just don't see the use in defeats (as a percentage of involved plays) without total defeats. I'm used to FO giving out both totals and percentages for most stats (like the DB stats). I might be spoiled.

by EaglesFan999 (not verified) :: Fri, 09/18/2009 - 4:08pm

You don't even necessarily have to do it on a per play basis. How about just raw Stops & Defeats. I don't see Trent Cole listed in the charts but he's mentioned as having the most. Can you print the total list?

by yu hao (not verified) :: Tue, 10/20/2009 - 8:44am

NFL Jerseys Store Online provides nfl pro bowl,nfl replica, also you can find 2009 pro bowl,pro bowl voting,nfl pro bowl 2008 .UGG boots .Get the discont and free shipping