Back to Chat | Baseball Prospectus Home

Chat: Aaron Schatz (Football)

Welcome to Baseball Prospectus' Friday July 18, 2008 12:00 PM ET chat session with Aaron Schatz (Football).

Aaron Schatz is the lead author of Pro Football Prospectus 2008 and the senior writer at Football Outsiders.

Aaron Schatz (Football): Once again, I would like to welcome all the Baseball Prospectus readers to the wonderful world of that other sport, the one played with the brown, oblong ball. There are some people who have this strange belief that you are only allowed to like one sport. They don't want to read about other sports, and they take the chance to denigrate those other sports whenever they get a chance. If you aren't one of those people, Hello! I'm Aaron Schatz and once again I'm lead writer on the fourth annual Pro Football Prospectus, available in stores now. It's produced by yours truly and the great writers of FootballOutsiders.com. I'm here to take questions on the upcoming NFL season, your fantasy football drafts, the changes in our advanced football statistics, and anything else that strikes your fancy. Because in a one-sport world, what on earth would baseball fans do between November and March?

Brian (Brooklyn): I ordered last year's book and ordered this year's edition a few days back so how about a spoiler? The 2008 Giants are a Super Bowl contender or a team that will win between 7 and 9 games?

Aaron Schatz (Football): Hmmm, the BP linking software seems to believe the 2008 Giants play in San Francisco... The tagline on the Giants chapter is that the Giants will be better than the team from the regular season and not as good as the team from the postseason. The Giants played like an 8-8 team last year, not a 10-6 team (until January) but there are a number of positive indicators that balance out the fact that luck is likely to regress to the mean. Eli Manning was much better on first and second down than third down, which is a good indicator for growth in 2008. They had a lot of injuries last year, and should be healthier in 2008. There is strong continuity on the offensive line. The book gives them a mean projection of 9.6 wins, but that is before Michael Strahan's retirement. Take a Pro Bowl defensive end off the team and you get more of a 9-7 mean projection.

akachazz (DC): Hey Aaron, Love the work you guys have been doing. Do /will you have projected standings for the upcoming season either on your site or in your book? Those are really fun to look at.

Aaron Schatz (Football): Thanks, man. The book lists mean projection for all 32 teams. We'll be doing an updated version of those projections on FootballOutsiders.com sometime in early September before the season starts, taking into account things like Strahan's retirement, the Brett Favre nonsense, and preseason injuries.

By the way, this points out an important difference between Baseball Prospectus and Pro Football Prospectus. The baseball guys build their team projections off the individual projections made by PECOTA. While our team projection system does include some variables related to personnel, the team projection does come first in football, and then that helps create the KUBIAK individual fantasy football projections.

Chip (Chicago): Why is FO so down on the Bears this year? I would have assumed that the significant injuries and strong weighted DVOA performance (#14 Total, #2 DEF; #25 OFF - hey that's incredible for this Grossman/Orton/Griese team) would dominate any third down effect. What major variable effects am I missing with Kubiak?

Aaron Schatz (Football): Well, we can start with the fact that the offense is just a black hole of suck. Frankly, their offensive projection seems a bit high when you look closer and the personnel... We've got a rookie left tackle and very little offensive line continuity. We have two quarterbacks who have pretty much proven they are awful. The running game is going to use a rookie and a guy coming off a Lis Franc. The top wide receiver is 32 and averaged 11 yards per catch last year, and the number two receiver may be a return man whose hands weren't good enough to play wide receiver in college. Egads. As for defense, you have the third-down effect, as you note (Bears were second on third down, 14th and 16th on first and second). You also have the fact that not all injuries are created equal. From what we can tell from basic research so far, defensive line injuries tend to recur more than injuries to, say, offensive linemen or wide receivers or defensive backs. And defensive line injuries happen to be Chicago's major injury problem over the past couple years.

Doug (York, PA): Any chance we'll see you guys at a book signing in the D.C. area again? Maybe Harrisburg?

Aaron Schatz (Football): We were pretty shocked when Politics and Prose decided not to host us this year, but they were apparently all booked up in September and didn't want to do early August. Instead, I'll be at Barnes and Noble Johns Hopkins in Baltimore on Tuesday night, August 5.

Corkedbat (Dallas): Thinking of speed scores, do they change throughout a RBs career (before the age 29 decent or in case of injuries)? Is there a way that could measured for analysis in future PFP's?

Aaron Schatz (Football): Well, they are based on the 40-time at the combine, so they are really meant to project rookies. We don't get 40 times for guys who are veterans.

bartleby (chicago): Aaron, love your work. What have you done to incorporate RB usage patterns into KUBIAK? or is that even possible? That seemed to be a major weakness in your fantasy projections last year.

Aaron Schatz (Football): Yes, it has been a major weakness in our fantasy projections every year. Of course, it is a major weakness in everybody's fantasy projections every year. It's a very difficult thing to project because it seems to have very little to do with how good players actually are. We're working on it every year, and we hope we do a better job this year than last year, and a better job next year than this year.

josher464 (NYC): Aaron, I have to question the blase way your site has treated a potential move of Buffalo over the border, as if fans will simply follow the team. I certainly haven't heard that sentiment expressed, and I think it's a shame that such a well written site would support the betrayal of such a loyal fan fase. Comments?

Aaron Schatz (Football): Well, we certainly don't mean to treat Buffalo fans badly. I think we've written about the potential move of the Bills in the same way we write about most things -- an intelligent regard for the facts mixed with more snark than Peter King and less snark than Deadspin. I hate to say this, but Buffalo simply is not a major league market. It is something like the 35th biggest market in the United States, even including Rochester and Syracuse, and it is not growing. The Bills do not have the nationwide fan base or public ownership that allow the Green Bay Packers to remain in a sub-optimal market. Nearly everyone in the NFL believes that if the Bills are still in Buffalo when Ralph Wilson passes away, you couldn't find an owner who would buy the team and leave it there. The team is pretty much going to move -- the question is where and when. For a lot of people, "when" is less important than making sure "where" is somewhere reasonably close to Buffalo and the team's current fans rather than, say, Los Angeles.

Skins Fan (Portsmouth, RI): Aaron give you us a clue of your top 5 QBs for the year?

Aaron Schatz (Football): I assume you mean in fantasy football, not DVOA or DYAR. Brady and Manning are the top two, with the order depending on your taste for risk. Combining numbers and stability, Drew Brees is probably number three. KUBIAK spits out the next two quarterbacks as Donovan McNabb and, believe it or not, Eli Manning -- as I said earlier, there are a lot of reasons to believe he'll improve that have nothing to do with the postseason. Ben Roethlisberger would be top five if not for injury risk -- the guy never seems to play all 16 games. Also, if it is becomes clear that Chad Johnson really is sticking around in Cincinnati, and we re-do things, it might move Carson Palmer up a bit. I don't think he would hit top five, but it would be close.

Steve (Georgia): Okay, I'll get it out of the way early: what's the deal with DYAR and True Yards?

Aaron Schatz (Football): This was the first question posted. Not very specific, though, is it? Let's see if we can do better...

shamah (NYC): In three sentences or less, what's the difference between DPAR and DYAR?

Aaron Schatz (Football): The translation between "success" and yards is more accurate than our previous translation between "success" and points. It does a better job of incorporating the differences in opportunity between players, and it is more tuned specifically to each position being measured.

DYAR also includes improvements that would have been made whether we changed the formatting from points to yards or left it the same: more accurate measurement of replacement level, better opponent adjustments, and the incorporation of defensive pass interference.

Oliver (Austin, TX): Does DYAR give "extra credit" for gaining more valuable yardage? In other words, is one actual yard sometimes "worth" 1.5 "DYAR yards"?

Aaron Schatz (Football): Absolutely. In fact, one yard is almost never worth "one yard over replacement." The basis of the individual stats has not changed at all. The formula still starts with taking every single play and figuring out its success value based on both raw yardage and yardage towards a first down, then taking that value and comparing it to a "replacement level baseline" based on situation and opponent. The main change is that we translate that into yardage, not points, because yardage is easier for most football fans to understand. The writers at Football Outsiders are more concerned with accessibility than the most hardcore readers of the site, and that honestly makes us no different from the writers at Baseball Prospectus or Basketball Prospectus.

db610wip (Philly): Hey Aaron, huge fan of the KUBIAK fantasy football projection system, the return on investment year after year is just oustanding. This year, however, I have noticed that in standard formats 1qb/2rb/3wr, WRs load the top 10, to the point where andre johnson and colston become viable picks 1st rd picks. Is this a glitch in the system, overestimation of value? Or do you think this says something general about draft strategy, that elite WRs are worth loading up on?

Aaron Schatz (Football): Yeah, isn't it wild? I was pretty blown away by this. I think it is connected to a couple things. First, the rise of the running back by committee over the past couple years. Second, in improving the WR projection system this year, we ended up with projections that did a lot less regression to the mean than in previous years. This helps to answer a constant complaint from readers in the past. People used to wonder why would always end up projecting the top wide receiver with, say, only eight touchdowns. To give a sample -- KUBIAK 2007 projected only one wide receiver with 10 touchdowns, and it ended up being wrong about that one (Marvin Harrison). KUBIAK 2008 projects 12 wide receivers with at least 10 touchdowns. KUBIAK 2007 projected only four wide receivers over 1,100 yards. KUBIAK 2008 projects 11 wide receivers over 1,100 yards. The actual number from last year was something around 15 guys with 1,100 yards and 12 guys with 10 touchdowns, so these new projections are a lot closer.

I think that, in general, fantasy writers have always assumed that wide receivers would be harder to predict than running backs, and that's another reason they tell you to rarely take a wide receiver in the first two rounds. I don't know if that really seems like it is still the case, especially after last year, when the top projected running backs were all over the map.

GTripp0012 (Chicago): Aaron, can you explain the Raiders projection (3.9 wins) being lower than last years Raiders' Pythagorean record (4.9 wins)?

Aaron Schatz (Football): Yes, last year's Raiders weren't using a first-time starter at quarterback and, as impossible as this seems, actually had better offensive line continuity.

Roger (Madison, WI): Aaron, I was shocked to see Marion Barber so low in your projections. After being top 5 in DVOA, 3 in DYAR, having a 50% success rate and a high APC, I was almost certain the system would shoot out Gore-like projections circa 2007 moving into the starting role? Is your damped Dallas expectations really the only reason for this or is there something else?

Aaron Schatz (Football): We're back to the problem of figuring out running back usage patterns. Remember, last year we had a major problem where we took young running backs with high DVOA and gave them too many carries. We tried to temper that this year, but that leads to assuming that Barber and Felix Jones will split carries much like Barber and Julius Jones did.

Corkedbat (Dallas): Honestly, you know the age of the Ravens Def and I am a Ravens fan having grown up in B-more, but do you really think they got a chance to be a top Def again this year? The projection, while I hope is true, just doesn't match up with anyone else's analysis.

Aaron Schatz (Football): Oh, absolutely, the Ravens have the best defensive projection of any team in the league. Age isn't really an issue, the Ravens have a good mix of veterans and youngsters like Landry and Ngata. They had a ton of injuries last year in the secondary, which is a good sign of bounce back. Their Adjusted Sack Rate dropped significantly (from second to 14th) and that's likely to bounce back.

Here's another reason for the Ravens projection: You might remember an essay in the 2007 book regarding "Bend But Don't Break Defenses" which introduced a stat called "points prevented per drive." A high number meant a team that gave up lots of yards but rarely gave up points, and a low number meant a team that tended to give up more points than you would expect from their yardage allowed. In general, teams with a particularly low PP/Dr will improve the following year -- some teams seem to be good in this area nearly every year (Patriots, Eagles) but there aren't any teams that are consistently poor. Here are the five worst teams in PP/Dr in 2007:

BAL -0.381
MIA -0.325
STL -0.296
ARI -0.194
NYG -0.182

Just for those curious, the top five:

BUF 0.298
IND 0.287
SD 0.279
GB 0.245
MIN 0.168

... with Philadelphia sixth and New England seventh.

Chad (Atlanta): Aaron, love the new book and all the new ideas you guys bring to the table as well as your great fantasy results from last year's Kubiak. I won 4 of the 5 leagues I was in using primarily your list. I was wondering do you guys plan to expand your fantasy coverage any time soon?

Aaron Schatz (Football): Thanks for the kind words and for not saying the name "Ladell Betts." I can't really give details yet because we're not really sure on the details yet, but there is a plan to do more fantasy football material. I'm not sure how much of that will go into action in 2008 compared to 2009. I do know that we'll be doing a weekly "start/sit" column as part of our relationship with ESPN.com.

Rusty (Maryland): Question about prospectus '08: If the redskins' defense had a very high dvoa on 1st and 2nd down, but terrible dvoa on 3rd AND the PFP '08 text suggests they have good personal for a west coast offense, why is their projection so low? (I don't even disagree with the projection, just seems text and stats don't match the projection). are the depth and coaching issues so bad that they outweight the above positive signs?

Aaron Schatz (Football): Actually, Washington's defense is projected to be slightly above average. The problem is the offense. The system doesn't like offenses that are introducing a new head coach and a new offensive coordinator simultaneously. Obviously there are exceptions -- the 2007 Cowboys, for example -- but in general these teams tend to do a little worse than the projection would suggest using other variables. There are a lot of other little projection bits that also end up negative.

Mike (Washington, DC): I'm a new reader of PFP and just received my copy a few days ago. I love the book so far and am just getting familiar with DVOA, DYAR, etc and am still reading the Statistical Toolbox section. Who would be your Top 3 breakout stars this year in Fantasy for this upcoming season? (maybe this is covered later in the book, I just havn't gotten there yet)

Aaron Schatz (Football): It is covered, in fact, in an article called "Top Fantasy Risers and Fallers." If I were to name just three, I would say:

1) Marshawn Lynch. One of the rare running backs these days who is not in a committee, and the Bills should play a fairly easy schedule of run defenses (Dolphins, Jets, AFC and NFC West).

2) Santonio Holmes. He's good. Very good. His quarterback is good. The cornerbacks of the AFC North are not. The other starting wide receiver was, but he's older and declining.

3) Anthony Gonzalez -- although if Harrison is clearly healthy by the time we reach the end of the preseason, Gonzalez may not get the passes we're projecting for him.

Mike (DC): The Lewin Career Forecast: Is there a link to where these numbers are posted? I've read that Jason Campbell had one of the higher scores. How do you think he will do this year and in years to come with, hopefully, the Redskins?

Aaron Schatz (Football): Just to get to the Redskins again... I think we all still like Campbell for the long term, but once again he's stuck learning yet another system, and that's probably going to set him back a bit. The guy has had something like 11 offensive coordinators in 12 years if you go back to high school. That team also is depending on a couple of rookie wide receivers, and as they grow and develop in future seasons, it will help Campbell grow and develop. The Lewin Career Forecast numbers aren't posted on the site, but you can find them in articles in both Pro Football Prospectus 2006 and 2007. They need to be slightly re-worked so they can translate into DYAR instead of DPAR.

Let's hit a few more short answers...

BL (Bozeman, MT): How do evaluate KC's draft, and where does that organization go post-Jared Allen?

Aaron Schatz (Football): Everyone likes that draft. Build from the lines out. They know they won't be any good this year.

Josh (New York, NY): Why doesn't KUBIAK consider postseason performance? I would think performance in a postseason game would be at least as predictive as performance in a regular season game.

Aaron Schatz (Football): I've tried it and it hasn't come out predictive. Like most things that don't work in the system, I'll probably end up trying it again in the future and that might change.

Keith (Fairview Park): Is there any way to get yardage projections for players if you just purchase the online kubiak projections.

Aaron Schatz (Football): Yardage projections for every player are in the online KUBIAK projections, sure. Even the DVOA projections -- which we did this year for the first time -- are in the downloadable spreadsheet as of Tuesday.

Arthur (Philadelphia): What should a Dolphins fan look for this year? I'm not just talking about the McCown/Beck/Henne QB mashup. Sparano is the 5th head coach in 5 years. Anything exciting in the new staff? Do I cross my fingers and hope Parcells still has some magic?

Aaron Schatz (Football): You should look for a calendar that says "September 2009" on the top of it.

Doug (York, PA): Does New England break the Super Bowl Loser's Curse?

Aaron Schatz (Football): Didn't Seattle already break the Super Bowl Loser's Curse? The Patriots once again look like the best team in the league, and once again that's not a guarantee of a championship, but they are the most likely winner of Super Bowl XLIII, absolutely.

Eric J (Fredericksburg VA): So, umm... Favre. How many teams have you projected him onto, just out of curiosity, and what sort of effect did he have on them? In particular, which team does he upgrade the most?

Aaron Schatz (Football): I don't project guys onto teams until they go to those teams, really. I wrote about this a week ago, but I think the only team that makes sense is Baltimore. The Packers will never allow him to go to a team in their division, and I can't see them trading him within the conference either. Baltimore is the one team in the AFC with a clear hole at quarterback, a championship-caliber defense, and the ability to take on a veteran with one or two more years left because they have a young franchise quarterback waiting in the wings.

T-Wrap (Knoxville, TN): Haven't picked the book up yet--just can't bring myself to do it until August, baseball season and all, but it's taking all of my restraint to hold out! Can you give me a quick rundown of the best OL's? There's still a perception that Denver makes RB's of average skill Pro Bowlers, but that ship seems to me to have sailed a few years ago. So, relatedly, what less publicized RB's do you see as putting up decent fantsy numbers?

Aaron Schatz (Football): Pro Football Prospectus comes written in short bits that are easily digestible between innings! As a bonus, much of the KUBIAK projection system was perfected while watching Red Sox games. Yes, Denver's offensive line looks like a mess this year. The best OLs... New England, Cleveland, Dallas, Green Bay, Indianapolis, the Giants, and the Vikings left side. Not in that order. Less publicized running backs... Sammy Morris, Steve Slaton, Pierre Thomas, Mewelde Moore in REALLY deep leagues... and in shallow leagues, Earnest Graham. Hello, kids, starter on a winning team? Good numbers last year? Offensive line with both youth and continuity? Other top running back has major injury and may never play again? I have no idea why this guy is so low on everyone else's list.

wammer (austin): Is there any reliable way to predict kickoff and punt return yardage on a team-wide basis, or does it fluctuate too much from year to year?

Aaron Schatz (Football): I've honestly never tried. We've had a few requests to include it in KUBIAK. I do want to work on return touchdowns because I know in some leagues, if you have a WR and he scores a touchdown on special teams, it counts -- obviously, this boosts Devin Hester tremendously.

Jack (SD): I always hear people asking for their top 5 picks in 2008, but I am going to have pick #6, care to extend that list for me?

Aaron Schatz (Football): Well, it depends on your league, right? Let's say the top five are Tomlinson, Westbrook, Peterson, Moss and Stephen Jackson... the sixth guy would probably be either Joseph Addai or your choice of Manning or Brady, depending on the rules of your league.

db610wip (philly): I hear you about the WRs in KUBIAK this year- but- would you then suggest taking a WR in the first? 2 of your first 3 rd? 3 of your first 5? if the value over baseline is there of course..

Aaron Schatz (Football): One more question and this one is really important.

We say this in the KUBIAK download, on the instructions page. Readers have written this on the site numerous times. The KUBIAK projections, just like the PECOTA projections, work best if you use your noggin. Remember that a fantasy draft -- and ESPECIALLY a fantasy AUCTION -- is not just about picking the best players. It is about getting the best value based on what the rest of the owners are thinking. As we say in the download, "Make sure you don't draft players too early simply because our projection for that player is much higher than conventional wisdom." That's why we include Average Draft Position information.

Obviously, it depends on the structure and rules of your league, but I would not take a wide receiver in the first round unless it is Randy Moss. The trick is to figure out when other players are likely to take wide receivers, and take them one round earlier. If most of your league starts with running backs in the first two rounds, make sure you take your top receiver in round two. If most people start RB-RB-QB-WR, make sure you have two wide receivers by the end of round four. But that doesn't mean take them in the first two rounds.

One thing that you can easily take advantage of -- people have a hard time understanding how much small scoring changes or changes in roster size or league size can change the values of players. You can't use a list from a random magazine if you play in a PPR league that has only eight teams and starts four wide receivers. The KUBIAK download makes your life a lot easier by letting you come up with a value list tailored to your league's size and rules.

Aaron Schatz (Football): OK, that's it for me. Thank you all for coming, drive safely, tip your waitresses. Please purchase Pro Football Prospectus 2008 if you have not done so yet, and come to FootballOutsiders.com for the best NFL coverage anywhere (and some pretty good college football coverage too). Also click the picture of the book on our site for more details on the PFP 2008 tour: Baltimore August 5, New York August 7, Indianapolis September 9, Chicago September 10, Philadelphia September 11, Boston September 17, and anywhere else if someone wants to pay for me to fly there.

Baseball Prospectus Home  |  Terms of Service  |  Privacy Policy  |  Customer Service  |  Newsletter  |  Masthead  |  Contact Us