Aaron Schatz, the founder of FootballOutsiders.com, will be the lead author of Football Prospectus 2005.
Aaron Schatz: Hello there, everyone! I want to thank the BP guys for the opportunity to use their living room for our little football talk, we'll try to clean up all the tackles and helmets before we leave. To introduce myself briefly, my name is Aaron Schatz, I launched a website called FootballOutsiders.com roughly 18 months ago, and our goal is to do for NFL analysis what BP does for baseball analysis. If you are a BP reader just learning about Football Outsiders, I'm happy to answer questions about the basics of our methods, or about how baseball analysis compares to football analysis. If you are a longtime FO reader, I'm happy to answer questions about specific teams and players, how they are doing this year, and how our methods rate them different from conventional stats. This is not the correct BP chat for your baseball questions, although for the person who asked about Glendon Rusch, I will add that Glendon Rusch is the face that Voros McCracken sees in his nightmares.
Bill Johnson (New Mexico): Welcome to the world of BP chats! For starters, how about introducing yourself? Where'd you come from, how'd you get connected to the BP crowd, and so on? I know about FootballOutsiders, but what's your connection to the (infinitely preferable...) world of baseball? Whatever your answers, welcome to a really exciting operation!
Aaron Schatz: Where I came from is a long story, I've been a radio disc jockey, a market research analyst, and for three years I wrote a column about Internet search trends called the Lycos 50. I started FO as a side project, when I had a couple questions about the Patriots and decided to count box scores in old school Bill James style in order to find out the answers. That turned into a website, and we exploded on the scene thanks to our association with Gregg "TMQ" Easterbrook after his unfortunate dismissal from ESPN.com Page 2. (Ironically, I now contribute to Page 2.)
As far as connections to the BP crowd, they approached me a few months ago. They were interested in restarting the Pro Football Prospectus brand, and felt that the work we were doing over at FO was very much in the vein of the work that BP was doing. Since my goal was always to do football research as good as BP (and with writing as witty) I was quite flattered. Actually, the whole gang at FO is pretty psyched. For those wondering, our website will remain footballoutsiders.com and we do post new content pretty much daily during the NFL season.
YankeesSuck0213 (Hiding in the bushes at Billy Beane's House): DVOA, as good as VORP? Do they measure the same things?
Aaron Schatz: Ah, our friend who dislikes the Yankees (hey, who doesn't?) is asking about our main statistic, DVOA. It stands for Defense-Adjusted Value Over Average. What we do is take every single play during the NFL season, give it a number which represents success on two levels (total yardage, and yardage towards a first down), and then compare it to similar plays from the past two years based on down, distance, location on field, opponent, and so on. You can read more about it at http://www.footballoutsiders.com/methods.php.
The similar stat to VORP is actually not DVOA but rather DPAR, which stands for Defense-Adjusted Points Above Replacement. The replacement idea was, frankly, completely stolen from BP. DPAR is a total, not a percentage, and it takes into account the fact that an average player does have significant value in the NFL because he can take attention away from the other aspects of the offense, and if he is a running back he can eat clock time.
Our stats are not as good as stats like VORP, to be honest, for a number of reasons. First, we've only been playing with these things for a couple of years, whereas baseball research is decades old. So we have much to learn. Second, the baseball folks can play with numbers going back years and years. Since I need every single play to compute DPAR and DVOA, I can only do years where I have complete pbp. (So far, I have back through 1997 and I've computed back through 1999).
There is also the problem of differentiating between various parts of the same team. Baseball offers a very isolated matchup of hitter and pitcher. With football, everything is an interlocking part. Eventually I hope we can isolate things a bit more. For now, when I say "Curtis Martin has more DPAR than any other running back in the NFL this season" what I'm actually saying is "Curtis Martin, combined with the Jets offensive line and with Chad Pennington selling the play fake, has more DPAR than any other running back in the NFL this season."
I used to say that football analysis was still in the "Bill James mimeographicing in his garage" phase. I think we've moved on to Bill's first book for Ballentine at this point but that still means we are two decades behind baseball research. There's lots more to learn!
Will AKA RCS (New Brunswick): The worst, least true yet popularly believed misconception about football is....
Aaron Schatz: ...that the players are idiots. The whole George Will "Men at Work" meme of baseball as graceful game also puts across the idea that football is just a bunch of fat sweaty guys running into each other. But football players have to make decisions that are as intricate as the ones made by baseball pitchers and hitters who are trying to outguess each other. Is this a blitz coming? Where is the quarterback planning to throw? Is the lineman next to me about to be double teamed? And so on. Sure, some of the players aren't so smart (Dexter Manley) but teams are now putting a premium on intelligence and the ability to understand a playbook that sometimes changes each week. The Patriots especially.
OrioleDog (Maryland): Football Prospectus sounds like a cool concept. Couple of Colts questions.
With James and Harrison both impending free agents, who would do you prioritize?
Do you foresee enough upside in the current defensive personnel to provide enough support for a historically great offense to win a championship in the next few years, or will a major overhaul be needed to make title hopes realistic?
Aaron Schatz: This is a pretty tough question. The problem is that both James and Harrison are reaching the age where players at their positions decline. Harrison's decline clearly started last year, and has continued this year. He's still an excellent receiver, of course, but he's not worth the money if he wants to be paid as one of the top five in football. Before the season started I absolutely thought that the Colts should let James walk after this season. As many people know, I am an advocate for the running back by committee both for performance reasons (you can get two running backs who are each good at a different aspect of the game) and economic reasons (a committee is often cheaper). James had clearly not been the same back since returning from the ACL... but this year he has taken a clear step back towards his pre-injury ability level. He's currently #2 in our RB DPAR numbers behind Martin and he's remarkably consistent from play to play -- he rarely breaks it long but he rarely gets stuffed.
Still, running backs break down FAST and then you end up running them out there for mediocre run after mediocre run like Eddie George. I think that if the Colts can sign Harrison for something reasonable, that takes his age into account, he's the one they should keep. Otherwise -- although this would be hugely unpopular in Indiana -- they should let them both go and spend the money on defensive free agents.
As for the Colts' Super Bowl chances, they moved ahead of Pittsburgh in our ratings this week, and their defense seems to be edging up towards league average. I write about this more in my Week 11 DVOA commentary today over at footballoutsiders.com.
Kordell (Chicago): Do you think stats can give insight into everything, or are there some things, like line play, that can't be analyzed with stats. And are there any analysts who are particularly good at breaking down line play without using stats?
Aaron Schatz: I hope that eventually we can get insight into everything with stats, but I don't know if we'll ever get to a point where a statistic can be the total measurement of a football player's ability -- just him, without the effects of the rest of his team. We do try to do a bit with line play, using a stat called adjusted line yards. Go to footballoutsiders.com, click on the Offensive Line menu item under JUST THE STATS.
As far as analysts who break things down without using stats, he doesn't do line play as much as he does secondary play, but Ron Jaworski is just amazing. I always learn something new when I catch him breaking down tape on some show.
geer08 (Birmingham): John Madden made what seemed like a great comparison last night: Tom Brady is Montana, and Peyton Manning is Marino...I thought that sounded about right (but I'm bummed because it likely means no rings for the Colts)...thoughts?
Aaron Schatz: I actually made virtually the same comparison last week in my QB ratings over on ESPN.com's Snap Judgment. You can find those here: http://sports.espn.go.com/espn/page2/story?page=snap/week10/rankings
I compared Manning to Elway, though, not Marino, because I do think that the odds are that at some point Indianapolis will put a team around him that can win it all, even if it doesn't happen until he's 37 years old.
The real comparison, for those who have not heard me make it, is that Tom Brady is Derek Jeter. My fellow Patriots fans HATE that, but it is true. Both are seen as "just winners." Both are seen as better than their stats would indicate. Both get far too much attention compared to other members of the team that contribute to victory. Both are overrated by people who love them and underrated by people who hate the people who love them. Both are darn good looking. And on and on. But Robert Kraft is about as far from Steinbrenner as is humanly possible.
Stan (Tennessee): Aaron,
You might want to add that the most complicated decision making takes place once the play starts. Receivers might have 3 or 4 different routes to run based on what the defense does once the ball is snapped. And the QB has to see the same defensive change and throw based on the same expected route adjustment. Doing this accurately in a split second with a 300 pound man about to smash into your body at full speed requires good decision making skills.
Aaron Schatz: Yep. Good point. Nothing to add there.
ckahrl (Burn Richmond, Burn!): What do you think about the difficulties of doing historical comparisons of players across time, given how many major rules changes have radically altered the game in the last three decades?
Aaron Schatz: Hey, look, it's our favorite Transaction Analyzer! Yeah, this is a really tough one. I've concentrated almost entirely on doing analysis of the current game and current players. Ryan Wilson and Michael David Smith from my website are more interested in historical comparisons, and they'll probably be writing more articles along those lines in the offseason. Mike wrote a whole piece in the preseason about comparing historical running backs based on how their yards per carry compared to their backups.
Football has changed far, far more than baseball, and far more often. They seem to fiddle with the rules every couple of years. When I am looking at historical data (for example, doing similarity scores) I usually begin things in 1978 with the start of the 16 game season and the liberalization of passing rules.
The good thing about this is that the "book" of "the way you are supposed to play" is much, much smaller in football than in baseball, because the rules have changed so often. That means coaches and front offices are far more open to new styles of play or roster construction than their equivalents in baseball.
Dave Retter (Chicago): Have you been contacted at all by any NFL front offices regarding your work? In general, have you find the football establishment to be more or less accomodating of statistical analysis than their baseball counterparts?
Aaron Schatz: Which leads me into this question... I think there is no doubt that the football establishment is more accomodating of statistical analysis than their baseball counterparts. Bud Goode has been consulting for some teams for decades. There is much more of a sense in the NFL of "we will try anything to win more games" instead of "well, this is the way they've been doing it since the 50's so it works."
In baseball, the objective analysis thing started with fans, was gradually picked up by major writers like Stark and Gammons, and then leaked into front offices. In football, there are some teams that have been doing video analysis and statistical research for years, but the major writers like Pasquarelli and King and Dr. Z are mostly still reciting the same stuff about how you need to establish the run, etc. My goal with FO is much more to revolutionize the coverage of football, the way you read about the NFL, rather than the way front offices manage things.
Not to say that I haven't been contacted by some front offices since I started this. Jim Schwartz, defensive coordinator of Tennessee, is very cutting edge when it comes to analysis, and I worked some projects for him last offseason. I've talked to Mike Eayrs of the Packers, though that was mostly me asking him questions and not the other way around. Two other teams have contacted me but I believe they wish to stay anonymous.
BTW, I'm not going to stop at 1pm... I realize I take a long time to answer questions so I'll stick around another half hour...
Justin (Los Angeles): Who should I start in my fantasy league next week? Brooks or Brady @ QB?
Are you a big baseball fan? Is there a Barry Bonds of Football? Also, which football stat is as overrated to the media as batting average for batters and wins for pitchers?
Aaron Schatz: I'll leave the fantasy start questions to my boys over at Scramble for the Ball, who answer such things in their mailbag... email them at Scramble@footballoutsiders.com and read them on Fridays (this week Thursday because of the early game).
I'm a huge baseball fan. I grew up with the Cubs, living in California and going after school to a sitter who was from Chicago and had WGN on every day. I carry a Sandberg baseball card around in my wallet. I moved to Boston at age 13 so I'm a Red Sox fan as well. The World Series win was a great, but I know other people felt a sense of vindication that I didn't quite feel because the Cubs are still my number one love. Even if I dislike the current manager.
The Barry Bonds of football played in the sixties and his name was Jim Brown.
Which football stat is overrated that much, geez, I could do a whole list. I would say touchdowns by a single player. That final yard between the one and the goal is harder to get than the yards that came before it, but not that much harder. Because of fantasy football we all have this idea in our heads that a touchdown is worth sixty times as much as a yard. So you will have guys like Edge run for 200 yards against a very good Chicago run defense, but he only gets one touchdown because Manning passes for the other four. Don't even get me started on those games Bettis was having at the beginning of the season. When you give an offense first and goal at the one-yard line, they will score 88% of the time, so it really isn't that much of an accomplishment.
Rich (Boston): Do you think FO might eventually branch out into analyzing situational stuff (like when to go on 4th-and-short, when to go for two, etc), or will you be sticking only to player/team overall evaluations?
Aaron Schatz: Actually, we've done some work on situational questions. Look in our archives for the following articles:
Better to Have Rushed and Lost
Guest Column: Going for Two
How Many Points is a Turnover Worth?
Setting the Pace, Parts I and II
We'll do more in the future, probably for Pro Football Prospectus 2005. I also recommend our friend William Krasker and his website footballcommentary.com. He may be contributing to the book as well.
e (Chicago): Week 10's WR stats have Hines Ward on top. Ward's catch % is at 88% while most of the other WRs are between 50%-70%. Would his high DPAR (points above replacement) be similar to a baseball player's inflated OPS due to a high singles rate (good luck)?
Aaron Schatz: No, I don't think so at all. Eventually I hope to figure out how much of the completion/incompletion is generally the quarterback's fault, and how much is generally the receiver's fault. For now, however, the bit of study that I have done shows that a receiver DOES have a significant impact on the percentage of balls thrown his way that are complete passes. Just ask Eli Manning about this. Ward is in this strange position of being underrated even though he's a Pro Bowler -- the guy is just an amazing receiver. Roethlisberger wouldn't be having the year he is having without Ward. Our numbers say he is neck and neck with Terrell Owens and Reggie Wayne as this year's most valuable WR and I think that's accurate.
As an aside, after Week 11's games, a TIGHT END actually passed all those wide receivers. Antonio Gates now registers as the most valuable receiver in football for 2004. Amazing for a guy who didn't play college football.
bronnerea (Smithfield, RI): Comparing Brady to Jeter is silly. A QB has much more impact on winning than a shortstop does. The QB makes the critical decisions on the field that have the biggest impact on the outcome. What decision making does a shortstop do? Jeter is nothing more than a media-created cheerleader. And yes, I am a Pats fan....
Aaron Schatz: Oh, no, I think you are correct that the QB is the most important player on the football field. But no matter how important he is -- even if he is the only player on the offense that matters in the slightest -- he can never be responsible for more than half the team's performance. Actually, about 43%, the split in a team's performance is about 43% offense, 43% defense, and 14% special teams. I feel that the New England need to insist on Brady as the best quarterback in football, instead of just one of the five best quarterbacks in football, is a huge sign of disrespect to Richard Seymour, Rodney Harrison, Mike Vrabel, and all the other players who make the Patriots defense so consistent, not to mention the other offensive players and the special teams guys like Vinatieri. The best team doesn't have to have the best quarterback, it is ok to say they have a very very good quarterback. Which leads me to...
Pete (Parts Unknown): Where do you rank the current Patriots among the all-time great teams?
Aaron Schatz: People have a hard time dealing with the Patriots because of the fact that they don't stand out in one particular area. BP fans are used to discussion of players who are good at everything but great at nothing, I dunno, call it "Bobby Grich Syndrome." The Patriots have this on a team level.
They were a good offense last year, they are a great offense this year. But they aren't the best offense -- that's Indianapolis. They were a great defense last year, they are a great defense again this year. But they aren't the best defense -- that's Baltimore. They have good special teams, but not great special teams. They have a good kicker, a clutch kicker, but not the best kicker (that would be David Akers in Philly).
The Patriots also don't blow people out very often. Last year, they never did until the final week. But they were by far the most consistent team in the NFL. Every week they had an above average performance, but they never had a spectacular performance (except Week 17). This year has been similar, except that they were horrid against Pittsburgh and amazing against Buffalo.
They also have to deal with the leftover feelings about 2001. The 2001 Patriots were a fluke. That doesn't make their championship any less deserved -- they won it fair and square by beating the other teams -- but they ranked 16th in DVOA and had a ridiculous amount of luck in a number of different games. In 2002 they were a better team even though they missed the playoffs. Last year, they were definitely the best team in football by January. This year they are #1 in our ratings. The Patriots have improved every single year since Belichick took over, but it doesn't look like it on the surface because that 2001 team won the improbable championship.
Anyway, as far as all-time great teams, I'd love to know what Eddie Epstein (who wrote DOMINANCE as well as BASEBALL DYNASTIES with Rob Neyer) would say. As a one season team, I don't know if last year's Pats would be in the top ten. But as a two season team, going 9-1 as the best team in the league for a second straight year, that's mighty impressive.
Stephen (Kingston, RI): Having read Easterbrook regularly, I've seen a few of your columns cited. However, do some of his mantras (kick early, go for it late, and "stop me before I blitz again!") have statistical backup?
Aaron Schatz: We first got Gregg to link to us when we analyzed his mantra: "Clang on First Bars Run on Second."
We discovered he was wrong, an incomplete pass on first down didn't necessarily mean a run on second down would not succeed.
The whole article we did about "Better to Have Rushed and Lost" was also for Gregg, an analysis of his idea that teams pass too much and should run more often in short yardage situations. This time our research completely agreed with him.
I'm hoping to revisit his Maroon Zone idea for PFP 2005, and also research the "Kick Early Go For It Late" thing. I'm guessing we will find that one is wrong. We're big believers in going for it on fourth down when you are deep in the red zone because if you miss, you have the other team pinned back into their own end and more often than not you will just get the ball back and be able to score on your next possession anyway. Often when Gregg criticizes a team for eschewing a field goal try, he doesn't notice that they just get the ball back and kick the exact same field goal on the next offensive drive, so it isn't like they gave up three points.
The blitz stuff is harder to do because the play by play logs do not tell you when a play is a blitz.
KennyGee (O Town): who is the best defensive player in football? my money's on ed reed of the ravens
Aaron Schatz: We've really become big Ed Reed fans this year. We picked Ray Lewis as the top defensive player last year, but I think he has finally been knocked off his perch. The one issue I have with Reed is that people don't normally think of a safety as the most important defensive player on the field.
People who have been reading FO know that we are big believers that the most VALUABLE defensive player is Green Bay nose tackle Grady Jackson. The Packers defense immediately turned around when he arrived last year at midseason. This year, when he got injured, the defense fell apart, and the Packers are undefeated since Jackson returned from the injury. He's good, but more importantly he allows all the players around him to return to roles with which they are more comfortable because he takes care of clogging the middle.
Creating some sort of individual defensive player statistics is in big bold 30 point highlighted letters on my list of things to do. Right now we're stuck with subjective conjecture.
Sean (Houston): AAron, I've been a FO reader for a while now and enjoy the articles but especially the statistical analysis. I have gathered that NFL play-by-play from NFL.com and ESPN.com are your main data sources. What are the ways that you ensure data correctness from these volatile web sources?
Aaron Schatz: On Saturday mornings, I add an extra paragraph to the Shmoneh Esrei asking that the NFL count its plays correctly.
No, seriously, I'm totally at their whim and mercy. I do try to seek out play corrections that the NFL makes days later and doesn't really tell anyone about. The non-standardization of NFL play by play recording and statistical records has to be my biggest pet peeve about my new career.
Vince (Stanford, CT): Who was the better Rashaan? Sheehee or Salaam?
Aaron Schatz: Vince had to know this question would get answered because we have a soft spot in our hearts for the XFL. I've always been a believer that a spring football minor league could work if done correctly. But the XFL, well, for you baseball fans, imagine how excited you would me if ESPN started to show games from the Winter Leagues down in the Dominican to brighten up the baseball offseason. Now imagine if they broadcast those games, but only if all the players would wear clown makeup, and second base would occasionally explode on stolen base attempts.
Anyway, Rashaan Salaam actually had a 1000-yard rushing season in the NFL, and he was second in the XFL on yards per carry behind current CFL star John Avery. He was the better Rashaan.
DrManhattan (NYC): Hey, Aaron - long time, no chat.
Do you think the Red Sox will be able to stick to their analytical guns and let Pedro & Varitek walk if they get pre-2001 offers, or will Lucchino & friends overrule Theo in the name of PR?
Aaron Schatz: This has nothing to do with football but is a chance for me to point out an article that I wrote for The New Republic Online. Yes, in addition to being Mr. Football Stats, I'm also the unofficial sports dude for TNR.
Listening to sports radio around here, it is clear that I was 100% correct that many fans would throw a fit if the Red Sox kept to the idea of economic efficiency -- even if they are used to it from the Patriots.
Craig (Chicago): Who would win a game played by CFL rules, the Toronto Argonauts or the Miami Dolphins?
Aaron Schatz: Giggle.
OK, one more...
Joseph T. (USA): I noticed Sunday that guys like Jerome Bettis and Curtis Martin are ahead or close to guys like Eric Dickerson and Tony Dorsett on the all-time rushing yards list. Have rushing numbers become easier to come by in recent times, sort of like home runs in baseball?
Aaron Schatz: Absolutely, for a number of reasons. First, going to the 16-game schedule in 1978 meant guys have more games to get those yards in (obviously that's not an issue with Dickerson, who played in the 80's). Second, teams twenty years ago ran two back sets much more often, giving runs to the fullback or splitting carries among two running backs. Third, running backs, like players in pretty much all sports, have longer careers these days. Most of them stick around much longer in their decline phases, and a few of them have managed to put off those decline phases. For Curtis Martin to have more value than any other RB in football at his age is pretty amazing.
Oh, whoops, forgot this question that got asked early and I actually planned an answer for...
Dayn (Parts Unknown): Hey, Aaron-
Where would you rank Walter Payton all-time in terms of most productive players? It seems to me people forget what a great receiver he was coming out of the backfield. And he had a cool mouthpiece. Full disclosure: he's my favorite player ever.
Aaron Schatz: Dayn submitted this early so I threw it to the two guys who do more historical stuff, Ryan Wilson and Michael David Smith.
Mike says: "In my view, the best thing about Payton is what a workhorse he was. He led the league in carries four times, he was probably the best of his era at picking up the blitz (Jim McMahon often mentioned that), and, as the questioner says, he was good at catching passes out of the backfield. However, the downside of Payton is that there wasn't really an era when he was the hands-down best back. When he first entered the league O.J. Simpson still set the standard, then in the early part of his career most people would say Earl Campbell was better than Payton; late in his career Eric Dickerson was superior. So longevity and durability were his strengths.
Ryan answered with some numbers he's done up called zscores but I'll save those so we have something fun to play with in the offseason.
Aaron Schatz: Wow. Thanks for all the questions. I recognized some names but I hope many of you are BP regulars who are just seeing Football Outsiders for the first time and will come check out our work. We've got regular content, discussion threads, a weekly cartoon, and a blog of the best football articles from around the Web.
I'll probably be doing a few more of these chats before the NFL season ends, and maybe we'll even send over Mike or our college football expert Russell Levine. And I look forward to doing more stuff with the BP guys, and of course writing Pro Football Prospectus 2005. In the words of my onetime WBRU Program Director Alexa Tobin: "Rock on."