If you’re familiar with Aaron’s work, you’re probably frantically MapQuesting for directions to Third Place Books right now – allow us to save you some time. Just click on this link, and you’ll be good to go. If you haven’t familiarized yourself with Football Outsiders or Pro Football Prospectus, here’s some background from the site itself:
Football Outsiders brings you a series of brand new, in-depth statistics you can't find anywhere else. With these stats, we will attempt to bring objective analysis to football that matches the revolution in baseball writing and analysis over the past 20 years. We have new methods for analyzing skill players, offensive and defensive lines, special teams, and total team efficiency. Right now these statistics are complete for the years 1997-2005, and they are updated each week during the 2006 season.
We don't just have reams of stats, though; we'll also have in-depth articles explaining these statistics as well as articles to answer specific questions and challenge conventional wisdom about the game. We'll also have articles that aren't necessarily based on statistics, but still give a more intelligent viewpoint on professional (and college) football, combining fan obsession with a bit of acerbic wit. Our lineup of regular columns is split between those articles which appear first here on Football Outsiders, and those which appear as part of our partnership with FOXSports.com.
Aaron is a longtime friend of Seahawks.NET – his Q & A articles before the NFC Championship and Super Bowl XL were amazing, as you can review by clicking on the following links:
NFC Championship Q & A (Four Parts)
Yep – this is an event not to be missed! Again, come see Aaron Schatz for an exclusive book signing/Q & A at the Third Place Books in Lake Forest Park.
As a tune-up for his first visit to Seattle, Aaron answered ten new questions about the new and updated methods he and the FO staff are using, as well as some Seahawks talk:
Seahawks.NET: First of all, for those who aren’t in the know, can you give us a Brief History of Time – how did Football Outsiders, and the complementary book. Pro Football Prospectus, come about? How did you get from there to here?
Aaron Schatz: First, dinosaurs roamed the earth. When they died, mankind took over the earth and eventually discovered that while it was fun to play a game where you kick a ball, it's even more fun to play a game where you get to pass it and run with it and measure your success in 10-yard increments.
We fast-forward to December 2002, when my hometown team, the Patriots, missed the playoffs after winning the Super Bowl the year before. A local writer named Ron Borges said this was because the Pats couldn't establish the run. But he also thought Oakland would win the Super Bowl and of course Oakland ran less than anyone else. So I decided to test the concept by counting all the running plays in the play-by-play. (It turns out that first-quarter runs do not cause wins; actually, fourth-quarter running is the effect and winning is the cause.)
Eventually I had this database of every play from the 2002 season and I started to goof around with it and created these new stats. I sent the articles around to a few people I knew and when no site was interested in running them I decided to launch my own site with the help of some friends. It was quickly clear there was a mass untapped audience for intelligent football writing, and within a year Baseball Prospectus had approached me and asked if I was interested in doing an annual book for their publisher, based on their style. Since my original goal in starting Football Outsiders was to basically be the BP of football, I jumped at the chance.
I was laid off from my regular job in February 2004 -- I wrote a column called the Lycos 50, monitoring the most popular searches on the Internet -- and gradually I got so much work writing about football that I was able to go off unemployment, and this has officially been my career ever since.
.NET: Can you explain what you call your “Statistical Toolbox” in PFP 2006? What are the metrics you use to measure performance, and why are they either a strong addition to, or more effective than, traditional football statistics?
Schatz: The main stat is called DVOA, or Defense-adjusted Value Over Average. The basic goal here is to filter out all the issues of context that exist in measuring NFL yardage. Three yards is different on third-and-1 than on second-and-15, three yards are harder to get against Washington than against Atlanta, that sort of thing. I break down every single play during the season and compare to a league baseline for that down, distance, location, and score (fourth quarter only), and then adjust based on opponent strength. We also do a version of DVOA which compares players to replacement level, rather than average, and adds up all their plays to figure out who has the most total value; this is called DPAR (Defense-adjusted Points Above Replacement).
We have a ton of new stuff after that, let's see if I can summarize. Pythagorean wins are a quick and dirty estimate of how much luck each team had based on projecting wins from points scored and allowed. Success rate measures the consistency of a running back. Adjusted line yards try to separate offensive lines from running backs by cutting off long runs past the first two levels of blocking. Adjusted sack rate measures sacks per pass play, rather than total sacks, and adjusts for opponent and for the fact that there are more sacks on third-and-long so quarterbacks with bad running games will end up taking more sacks. Stop rate measures defensive players by the percentage of tackles that actually stopped the offense short of success. And on and on and on.
.NET: In talking about the book, three of the primary things that would seem to make it valuable are: the ability to predict breakout or downfall seasons, the ability to narrow down what makes teams win and lose, and the ability to assist people in fantasy leagues. What else are you and the FO staff trying to accomplish with this book?
Schatz: I want people to come away from reading PFP thinking to themselves three things:
Wow, I know more about football now.
Wow, I enjoy football even more now.
Wow, there was some seriously funny s%$t in that book.
I consider myself a writer, not a statistician, and I happen to have built a staff of good writers to help me out. A bunch of numbers without words trying to explain them are boring. Words are the meat, stats are the spice. There's no way I'll ever put out a book that's just page after page of tables.
.NET: Please cite one example of a player you correctly predicted to rise, and one you correctly predicted to fall in 2005, and the logic behind those projections.
Schatz: Well, this interview is for Seattle fans, so we've got to talk about my boy Matt Hasselbeck, the pride of Westwood (two towns over from my hometown of Sharon). One of the strongest trends we've discovered is that teams that struggle on third downs, but are strong on first and second downs, will improve the following year. It's a combination of better luck, regression to the mean, and the fact that teams that are weak on third down recognize the problem and make roster moves to solve it. A few more converted third downs create a huge improvement in overall offense, because every new first down might be followed by one additional set of downs, or two, or three, or whatever.
Seattle's offense on first and second down has been the same quality for three years now, but third-down offense went from best in the league in 2003 to one of the worst in 2004. So we said, OK, this has to bounce back in 2005, and Hasselbeck will bounce back with it. And lo and behold, Seattle was once again near the top of the league in third-down offense in 2005, and just as we predicted, Hasselbeck was one of the top quarterbacks in fantasy football. Actually, due to the injuries to McNabb and Culpepper, he ranked even higher than our prediction (fourth). Oh, and yes, his name was misspelled on last year's book cover, and that was Workman's fault, our publisher, not mine.
As far as a correct prediction of a player falling in 2005, I'll skip the super easy one (Curtis Martin) and talk about the one where we went out on a limb, Tony Gonzalez. It's really simple: tight ends generally begin to decline at age 29. Gonzo was coming off a 1,200-yard season but only two tight ends in NFL history ever had over 1,000 yards at age 29 or older, Todd Christensen in 1986 and Shannon Sharpe in 1997. In fact, only four tight ends (in eight separate seasons) ever had over 800 yards at age 29 or older: Christensen, Sharpe, Mickey Shuler, and Wesley Walls. Most people thought he would still be the top tight end for fantasy football, but we had him third. This year, people still have him in the top three, and we have him tenth or something. Willie Roaf's retirement just makes things worse because, just like last year in the games Roaf missed with injury, he'll have to stay back to block more instead of going out on pass patterns.
.NET: What is the Game Charting Project, and how does it help you put things together?
Schatz: This is something we started in 2005. There are tons of important stats in an NFL game that teams track but the league does not track publicly. We wanted to track these things, so we got a group of volunteers to track games. We marked down things like Yards After Catch, grading defensive backs on coverage rather than tackles, what formations teams used and how many rushers and blockers matched up on each pass. From this data came all kinds of keen stats -- far more than we could actually fit into the book. For example, I can tell you that the Seahawks used a fullback on 87 percent of running plays, fifth in the league. They used two tight ends on only 22 percent of running plays, 29th in the league. They were near the bottom of the league in both screen passes and screen passes by opponents. And in his limited time on the field, Ken Hamlin had enough charted passes to finish second among 83 ranked safeties in Stop Rate, or preventing offensive success when he was the closest defender in coverage on a pass. There's a lot more in the book, although you want to make sure you haven't eaten recently before looking at Jordan Babineaux's charting data.
.NET: What is the KUBIAK Projection System?
Schatz: That's the projection system for individual fantasy football stats, based on a ton of variables including DVOA splits, height, weight, age, college stats for quarterbacks, team projections, and all kinds of other things. It's jokingly named after a backup quarterback from the eighties because Baseball Prospectus named their projection system PECOTA after a backup infielder from the eighties. (Note: You can order the KUBIAK System, and all subsequent included updates, here.)
.NET: In the 2005 Pro Football Prospectus, you said that Matt Hasselbeck was on the verge, and he was a primary reason that the Seahawks should be taken seriously as a Super Bowl contender. Good call! Going into the 2006 season, how do the Football Outsiders projections view our little team?
Schatz: Loves them. Loves them so much it is scary. 12.5 mean projected wins, and it would have been higher except the projection system for teams has a little thing in it that reduces the possibility that a team will project to have 15 or 16 wins or losses. I see no reason to believe that San Francisco will turn it around this year, although they have more talent for the future than most people think. We're much more negative on St. Louis than conventional wisdom. I suppose Arizona could finally put it together and finish 9-7, but even still that's not going to compete with Seattle, and my god, have you seen their offensive line this preseason? Poor, poor Edgerrin James. As for Seattle, the main worries are replacing Hutchinson, improving the secondary, and hoping that 430 carries (including postseason) doesn't lead to a big Alexander decline or injury. Nonetheless, the passing game is so good and the front seven so good that even that won't keep them from the division title.
The only problem, of course, is that this is basically what our numbers said last year about Philadelphia, and while we were right in almost all our playoff picks, Philadelphia was not one of them. I can't imagine that all the horrible things that happened to the Eagles will happen to the Seahawks -- locker room turmoil, injuries at every position, all three division rivals improving simultaneously -- but then again, I couldn't have imagined them happening to the Eagles either.
.NET: Based on your analysis, what are some things that even the most diehard Seahawks fan might not know about their team and its players?
Schatz: Other than the data in question five above, I can mention that last year D.J. Hackett had more DPAR than any other receiver with fewer than 50 passes thrown to him.
.NET: Once again, you’ll be doing your (sometimes controversial) weekly Power Rankings for FoxSports.com, based on your own numbers. Have you tweaked those numbers at all in the off-season?
Schatz: Yep, DVOA got its first overhaul in two years this off-season which improved things, making it both more predictive for the future and better correlated with wins in the past. Right now, I'm in the middle of finishing a new formula for early in the season which will do a better job of predicting the remaining games by combining performance in early games with preseason projection.
Of course, even with the older formula, our power ratings turned out to be a lot more prescient than people may realize. All that angry e-mail we got when we had Atlanta ranked 17th at 6-2, or Washington ranked 10th at 5-6 -- well, those ratings didn't seem so silly by the end of the season.
.NET: Anything else you’d like to tell us about what you're doing?
Schatz: The quick summary on the 2006 season: projects to be very much like the 2005 season, except Philadelphia much better and Minnesota much worse. The NFC East and AFC West are killer divisions, but Chicago and Seattle seem like they have smooth sailing to the playoffs.
I'm lucky to do this for a living. It's a pretty sweet way to make a living, so I thank everyone who reads the site and buys the book. I much prefer to do Q+A at the book events, rather than just give a spiel and sign books, so people should come to the one in Seattle and feel free to ask me anything.
If you have any questions about Aaron's book signing, call Third Place Books at 206-366-3333, or click here.
Aaron Schatz is the Editor-in-Chief of FootballOutsiders.com, Lead Writer and Statistician for the “Football Prospectus” annual volume, and an NFL analyst for FoxSports.com. He has also written for the New York Times, the New York Sun, the Boston Globe, The New Republic Online and Slate, and has done custom research for NFL.com and a number of NFL teams. Before creating Football Outsiders, he spent three years tracking search trends online for the internet column, “The Lycos 50”. He has a B.A. in economics from Brown University and lives in Framingham, Massachusetts, with his wife, Kathryn, and daughter, Mirinae.