In this week's Varsity Numbers, Bill Connelly takes a page out of baseball's playbook and attempts to isolate power from efficiency.
19 Aug 2005
by Aaron Schatz
Hello from Chicago, where it is hot and sunny and those f&%#ers from Northwest Airlines have lost my luggage. This is also the first stop on the tour where I am not even remotely optimistic about the local team, which is a bit of a bummer. On the other hand, I'm going to be on WGN 720 tonight, which has been a dream of mine since the age of, I dunno, eight or so. I'm a Cub Fan! Cub Fan! And a Bud Man! Bud Man! Holy Cow! Let's get some runs! Jerry Azumah spelled backwards is Hamuza Yrrej, and I'd like to thank Arnie back in the truck.
(Yes, it is odd that I have always dreamed of appearing on WGN even though I grew up in "The O.C." and then Boston. Blame cable.)
Here are some more questions and comments about articles on Football Outsiders as well as our book Pro Football Prospectus 2005. (On sale now!) We've put the best ones together in this mailbag -- which can also be used as an open thread to discuss the book.
Don't forget that we have a new contact form which you can use to e-mail any of the writers. We'll be doing a regular weekly mailbag during the season, and to be honest asking a question through the contact form is probably a better way to get your question answered than asking it in a discussion thread. (Especially this week, where I was lucky if I had any chance between hotel stops to look at the discussion threads.) I also encourage folks to pose their questions to other members of the FO staff based on their various specialties -- Mike and Mike are good for NFL history and all those players at positions that don't have DVOA and DPAR, Russell's the man for college football, Will of course for injuries, Al and Vivek for fantasy football questions, Ryan for economics, Benjy for technical questions about the website, and Jason if you want to suggest where he can hide Randy Moss's afro in next week's cartoon.
Be aware that I reference plenty of our innovative FO stats here, not to mention their unfamiliar terminology, so if you are a recent addition to the readership you might want to read this first.
Zac Hinz: Aaron, I know you mentioned that you guys would be making publicly available all the projections for players that had changed greatly since the book was published. When will you be doing that? Also, is the team projection system from PFP 2005 really saying that 8-8 might be good enough to win the NFC North? (According to the system, all four teams are more likely to be 8-8 or lower than they are to be 9-7 or higher.)
Aaron: Jack, I'd like to take the second part of that question first. Sure, the book predicts that 8-8 might be good enough to win the NFC North, but that's not likely to be the case. The chances are fairly strong that at least one of those four teams will have a winning record. (The chances are even stronger that this team will not be Chicago.)
The other issue is the Vikings defense, and let me explain. People ask me if there are variables in the DVOA projection system for players coming and going. There are a couple different things in there, but they are pretty hard to do and so they aren't as robust as I might like. But this generally is not a problem. Teams act predictably from year to year, making personnel moves to fill the holes on their team while losing players where they have depth or age. Going back over the past few years with the DVOA projection system, there are plenty of teams that the system gets wrong, but it is rare that the system gets that team wrong because of the impact of major personnel moves from the off-season. (Usually you either have a major injury issue or a sudden, unpredictable team turnaround.)
However, there is no doubt that every year there are one or two teams that have made such drastic personnel changes that our statistical projections aren't going to be nearly as accurate. Last year, for example, the San Francisco 49ers were that team. (Tennessee had lost some players to salary cap cuts, but their decline was caused for the most part by injuries to McNair, Calico, and two or three sets of defensive backs.) This year there is no question that team is Minnesota, with the experimental defensive transplant. Mike Tanier is working on a piece about similar defensive overhauls but we really have no idea if those players are going to fit together well, driving the Vikings to an 11-5 record, or not fit together at all, dropping the Vikings to 6-10. And the DVOA projection system doesn't really do a good job of guessing at this because it is so rare for a team to add so many new starters on one side of the ball.
(Speaking of the movie Quiz Show, by the way, does John Turturro hold the world record for Italian guys playing Jewish guys on film? Probably.)
And now, the first part of the question, and the one you are all waiting for. Yes, I promised I would make available projections that we had changed this pre-season. I must admit that I am very hesitant to change projections based on camp rumors of "this guy is going to be used more" or "this guy looks good this year" and so forth. But there are injuries, and there are holdouts, and there is the Curse of Mike Shanahan, well known to all fantasy football players who can't figure out who the hell Denver is playing at running back.
I am not going to be sending out the spreadsheet again to people who have already donated, but I will list new projections here. These won't really make sense if you don't have the book because I am only going to give projections for a handful of players, and this way those who bought the book but not the spreadsheet will not be left out. The Chicago projections are based on the idea that Hutchinson will play 10 games, Orton 8 (Some overlap there, obviously). Grossman may not be out for the year after all, so feel free to adjust accordingly. We also include here the effect of the Todd Pinkston injury -- though we're still going with the idea that the whole T.O. nonsense will have blown over by Week 1 -- the Cedric Benson holdout, Mike Anderson being listed as the starter (for now) in Denver, and Jerry Rice passing Darius Watts on the Denver depth chart.
As for players not listed here who are rumored to be moving around the depth chart, you are always welcome to try to separate the wheat from the chaff when it comes to pre-season news, and adjust our ratings according to your own best judgment.
And yes, since a lot of people are asking, we are still sending out the projections spreadsheet to anyone who makes a $10 donation to the website. It proved so popular that we've kept it available even though the book is in stores. Obviously, we hope that you'll buy the book as well.
(Late note added Sunday: Please see comment 30 below for a revised Tiki Barber projection for those who believe the Brandon Jacobs hype.)
Charles Jake: Is there any way to statistically project whether a No. 2 WR can be a successful No. 1? Reggie Wayne was the best receiver in 2004 according to FO metrics, but I think everybody sees him as Robin to Harrison's Batman. So can we tell if he can make it as a No. 1 or if he'd do a Peerless? I bet Vikings fans would love to know this about Nate Burleson.
Aaron: Damn, interesting question from the winner of this year's Offseason Free Agent Contest. The answer is that we've never done research of this nature, but it sounds like an excellent idea to consider for next year's book. As far as Burleson goes, remember the following:
|Weeks 7-11 (Moss injured)||10.2||5.8||59||0.80||9.1||15.4%||67%|
So Burleson had more opportunities without Moss, but he did less with each one.
Peter Quinn: Do you have an RSS feed for your Ramblings?
Benjy: We do have one: http://www.footballoutsiders.com/wordpress/wp-rss2.php is the feel URL. However, it's a bit screwy, as all the links come up as index.php?whatever, instead of ramblings or articles. We're working on fixing that.
Benjamin: Andy sense I have never met you I quess it's fair to say you are a hater of sports. Most of your articles are negative, does that mean you have a low self esteem or you sucked at playing sports? I seems it gives you great players to put players down, and no, stats doesn't always tell the story. I quess if you knew the game (any game) you would know that. Keep your opinions to youself, especially about people you don't know, or suit up yourself and carry the ball just once.
Aaron: That was the Football Outsiders Somewhat Incomprehensible Angry E-Mail of the Week.
Tim Towns: Aaron, I bought your PFP and am enjoying it. Very fresh information and analysis. One question though. I could not ascertain if your 2005 projections included the schedule DVOA as a factor. I got the impression from reading your stat explanations that you guys did the projections and then sorted out the schedule DVOA. I'm wondering if one should start with the projection and then consider the schedule DVOA as another factor on top of the projection when trying to sort out the prospects for a team this year?
Aaron: Thanks for picking up the book, Tim. The reason why schedule is not considered when we project DVOA is that DVOA is, by its very definition, supposed to be schedule-neutral, adjusted for opponent strength. However, schedule is part of the formula that was used for the win projection system, the table of each team's chances to be a Super Bowl Contender, Playoff Contender, and so on. And the projections for team offensive DVOA, defensive DVOA, and schedule strength are also used in the KUBIAK projection system for fantasy football. For example, this is part of why Jake Delhomme projects to lose fantasy value this year -- Carolina projects to have a significantly improved defense which in turn drops the number of passes Delhomme will throw.
As I've noted before, we will be running the mean projected DVOA for each team as well as mean projected wins in the week before the season starts, just like we did last year.
Ryan Restivo: I read the link of the professors' report on the draft and I want to ask you guys: why do teams pick quarterbacks in the first round if they know it's such a gamble? Also, what other football books would you recommend other than The Hidden Game of Football?
Aaron: I recently put together a piece for the Wall Street Journal listing a recommended reading list for the intelligent NFL fan so it turns out I've got an answer to your second question quite handy:
Hmmm, maybe I should put together an Amazon page with all these good football books linked so our readers can buy them and kick some of the money back to us. Of course, Joyner's book is self-published and both Hidden Game and Thinking Man's Guide are out of print. I left Epstein's book off the WSJ list by accident, so I apologize to Eddie and encourage folks to pick up a copy.
Your first question refers to this article by Michael David Smith but the right person to answer it is Ned Macey, who has done some preliminary research regarding the drafting of quarterbacks. Take it away, Ned.
Ned: Speaking only in terms of the draft, I found that teams should only draft quarterbacks in the first or second round. I looked at every QB drafted between 1990 and 2002. I then termed a QB a success if he he had a career passer rating over 70 and 1000 attempts, not exactly high benchmarks. At that standard, 52% of first rounders and 46% of second rounders were "successes." Only 6 out of 51 players drafted in rounds 3-5 were successes, and only 7 out of 44 players drafted in round 6-7 were successes. If we raise the threshold to either 80 passer rating or at least one selection to the Pro Bowl, we are at 36% of first rounders, 23% of second rounders, and 10% of rounds 3-7. (As an aside, Brad Johnson and Trent Green were eighth round picks before the draft shrunk, but not much else came from those later rounds when they existed.)
To me the clincher was that of players drafted after the second round in this period, only two went on to win playoff games with the team that drafted them: Neil O'Donnell and some guy in New England who isn't too bad. That's two quarterbacks out of 122 draft picks. Among first-round quarterbacks, 8 out of 25 have won playoff games for the team that drafted them.
Their are hundreds of people with the skill set to be almost-NFL quarterbacks, your Jim Sorgis or Sage Rosenfelses or all the guys in NFL Europe. Out of those with questionable physical attributes, some have that special something to be great NFL QBs, like Brady, Bulger, etc. A first round quarterback is not a guarantee, but he is much more likely to become a good or great player than a late-round pick.
Of course, a handful of undrafted quarterbacks have become stars, like Kurt Warner, Jake Delhomme, and Jeff Garcia. Why do teams draft first round quarterbacks instead of just looking for the next Delhomme? That is a question that I mean to investigate. With the amount of money invested in high first round picks, you would want a slightly safer investment. Someday I intend to look at how successful teams are based on how they acquired their quarterback (free agency, draft, trade, etc.). I also intend to read War and Peace, remodel one of my bathrooms, and end world hunger, so I'm not sure what my time table on that one is.
To sum up this slightly disjointed response, teams continue to select quarterbacks in the first round because they are most likely to be the best quarterbacks. It is hard to think that the Colts, Eagles, Vikings, Jets, Falcons, or Steelers are unhappy with that strategy. Whether or not this is the most efficient use of resources is yet to be seen, but it is not like the Cardinals, 49ers, or Dolphins have necessarily found a better way, and holding out hope that you'll find the next Tom Brady is like spending your rent money on scratch tickets because you just know that one of them will win you a million bucks.
Baxter: Any advice on getting PFP 2005 in the UK other than Amazon?
Aaron: Hmmm, not a clue. I doubt you'll find it in stores over there so online is probably the way to go. If any of our other UK readers have ideas for Baxter, stick those in the discussion thread.
Richie Wohlers: Does there exist a listing of cumulative opponents DVOA for 2005 broken down into offensive and defensive? For example, what is the Colts' cumulative opponents' defensive DVOA for 2005?
Aaron: In the book we list strength of schedule based on cumulative projected total DVOA of opponents, but that's not separated into offensive and defensive. That's easy enough to do, however, so let's do it here. Some notes for your edification:
This Team's Offense
This Team's Defense
This Team's Offense
This Team's Defense
This is part of why we think Michael Vick and Curtis Martin are a little screwed, and Donovan McNabb and Matt Hasselbeck are not. I'm a little surprised that Cincinnati faces the third-hardest schedule of defenses, but I think that's still the effect of playing Pittsburgh and Baltimore twice each despite exchanging the AFC East for the AFC South on the schedule.
More mailbag next week.
53 comments, Last at 24 Aug 2005, 11:29pm by Sid