ROTB Scores an Interview Football Outsiders' Aaron Schatz

Before we get into hardcore football season, I'll take one last opportunity to provide a final season preview for the Arizona Cardinals. Since I'll be using tons of stats from Football Outsiders throughout the season, I jumped at the opportunity to interview their President, Aaron Schatz. If you're unfamiliar with the fine work that Football Outsiders puts out there then I encourage you to check it out their website throughout the beginning of the season. Schatz is also the lead author of their annual Pro Football Prospectus, which is the source for any NFL stat including great situational stats. As you might expect, my questions and Aaron's answers were pretty lengthly so the entire Q&A session is after the jump, but here's how we got it started:

CG - First of all, I'm sure some of us here at Revenge of the Birds are unfamiliar with the work you guys do at Football Oustiders.com and the content of Pro Football Prospectus. I've been a fan of Football Outsiders since stumbling across the site last season and did my best to understand DVOA and DPAR, but I'm sure there's something that I missed. Can you give us a quick overview of the process that FO goes through to produce your various rankings and what kind of content can be found in PFP?

Aaron Schatz - First of all, I should point out that we've changed our individual
stats from DPAR to DYAR. They are now stated in terms of yards, which
we think will make them easier for fans to understand.

Football Outsiders was created to apply advanced statistical analysis
to pro football, although we've expanded since then, with columns that
are more scouting-based, and college football material. The stats we
do come in two varieties:

1) Stats based on the standard play-by-play, the Bill James-style
stuff. The basic idea is to break through the context of football
stats, and to look at each play in terms of success based on that
specific situation -- third-and-2 being very different from
second-and-15. DVOA is the main rating, which gives each play a
success rating and then adjusts it based on situation and opponent. We
also do things like analyze run tackles separately from pass tackles,
and filtering out the luck involved in fumble recovery.

2) Stats that try to fill the holes in the standard play-by-play,
collected by our armada of volunteer game charters. This is the
material like tracking formations, play-action, quarterback hurries,
defensive coverage, and so forth.

CG - Much has been made of Edgerrin James reaching the 30 year old plateau recently and questions have abounded about his ability to remain productive given his workload over the years. What if any information has FO compiled on backs after they reach that age? Is there any kind of correlation between the number of carries before a back starts to break down or is it more based on age?

AS - Al Bogdan addressed this in an essay on Warrick Dunn in the Atlanta
chapter of PFP 2006. Here are some of the findings:

a) The older the running back, the more likely he is to see a drop in
yards per carry from season to season.
b) Once a running back has around 2,000 career carries entering a
season, we see a negative correlation between the number of career
carries and the yards per carry the back will gain in the upcoming
season.
c) The relationship between career workload and YPC becomes stronger
than the relationship between age and YPC around 2,750 carries. (Edge
has 2,850 career carries.)

I think these paragraphs will be relevant to all Tim Hightower fans:

"There does not appear to be much of a relationship between the number
of carries an experienced running back has and how many yards he will
gain in a season. However, there is a negative relationship between
career carries and yards gained per carry. Given that the correlation
with yards is stronger than with yards per carry, that means the older
backs have fewer carries as well. By contrast, younger backs with more
career carries have fewer yards per carry but no fewer yards. They
must be carrying the ball at least as often as they did in their
previous seasons."

"One possible explanation for these results is that teams have a
quicker hook with older running backs who lose effectiveness than
younger ones with similar carries in their past. A younger back who
has lost effectiveness from the wear of carrying the ball 2,000 times
will be given a chance to work through his perceived slumps.  Older
backs who gain fewer yards per carry than they did earlier in their
careers may be replaced more quickly by a younger running back sitting
on the bench, or may be more likely to get injured and then replaced
by the young back on the bench. "

CG - Much has been made about Anquan Boldin's demand for a new contract and subsequent demand for a trade. Most seem to think that a contract somewhat similar to Larry Fitzgerald's would make Boldin a much happier Cardinal. There's been a debate for a season or two now as to who the better receiver is and who is more valuable to the team. What does FO think is the better receiver and how close are they?

AS - First of all, we need to remind people that we don't yet fully
understand the relationship between one receiver's stats and another
receiver's stats. There's no doubt that there are a few slot receivers
who look really good in the FO stats because they are getting
single-covered while the stars are being paid attention by safeties.
Each player's place in the team's system will affect his stats.

However... there's no contest here. Based on FO stats, Fitzgerald has
been the more valuable receiver, and it isn't really close. Three of
the four highest DYAR totals ever put up by Arizona wide receivers
belong to Fitzgerald in the past three seasons. (The other is David
Boston in 2001.)  Even as a rookie, when the Arizona offense was so
bad that both players came out as "below replacement level,"
Fitzgerald had more DYAR (-31) than Boldin (-74).

Last year was the first time Boldin had a higher DVOA than Fitzgerald
-- i.e., more value per play -- but Fitzgerald still ended up with
more DYAR because he had 167 targets compared to just 100 for Boldin.

CG - Cornerback is a position that most feel is one of the weakest positions on the team. How to Roderick Hood and Eric Green stack up against other corners around the league?

AS - Hood ranked 17th in our game charting stats for Success Rate, 22nd for
Adjusted Yards Allowed per Pass. Green was 71st and 59th respectively.
Rolle was 73rd and 79th (as a cornerback). I might as well just quote
from the unit comment for the Arizona secondary in PFP 2008:

"Arizona cornerback Roderick Hood was among the most targeted
cornerbacks in the league, and the Cardinals struggled against both
number-one and number-two receivers. Scouts feel Hood is clearly more
suited to be a nickel back, which is what he was in Philadelphia
before Arizona signed him in free agency. Yet Hood was one of the top
starting cornerbacks in the NFL according to our game charting
project. Where is the disconnect?"

Apparently, it is between the scouts and our numbers, not between our
play-by-play numbers (DVOA) and our game charting numbers (Success
Rate). Yes, Hood occasionally looks like a fool when he tries to jump
a route and guesses wrong. (The worst example of the year was probably
Vinny Testaverde's 65-yard touchdown pass to Steve Smith in Week 6,
when Hood wasn't even facing the right way.) But the Cardinals team
numbers against wide receivers were bad because of Eric Green and
Antrel Rolle, not Hood. Hood had better numbers than the other two
corners against number one wide receivers, number two wide receivers,
AND 'other' wide receivers."

CG - Are there any situations/formations that the Cardinals excelled in last season that a casual fan might not realize (ie. shotgun, play action, four wide etc.)?

AS - They were third behind New England and Jacksonville in passing DVOA in the third quarter. Of course, they usually weren't throwing with a
lead like those other two teams.

While Warner had a higher passing DVOA than Leinart overall, Leinart
had a much higher passing DVOA under pressure (five or more pass
rushers).

For two straight seasons, the Cardinals defense has sucked extra hard
against the shotgun.

CG - Finally, how does FO project the Cardinals this season, both as a team and a couple players individually?

AS - I think the general mean projection is for the Cardinals to go about
7-9 again. It's another building season, really. I think if the
Cardinals really do take a step forward and challenge for the
playoffs, the reason will be defense, not offense. I still think they
are another year away, though. I also think that playing Warner over
Leinart is a mistake. They need to have patience and trust that
Leinart will improve with experience. The only evidence that Leinart
is a bust is the fact that Whisenhunt won't just play the guy and let
him take his knocks. Warner gives them a better chance to go 9-7, but
is that the point, or is the point to try to make it to Super Bowl
XLIV or XLV?

As far as specific players, you know, the usual. Fitzgerald will be
great, Boldin good, James mediocre.

Finally, let me just say we need more Arizona fans reading FO. You
guys are seriously outnumbered by the Seattle people, not to mention
all the East Coast fans. Come over and defend your team in our message
boards!

-----------------------------------------------------------------------------------------------

First, many thanks to Aaron Schatz. So what do you guys think? Is anyone already a Football Outsiders or Pro Football Prospectus reader? Agree or disagree with what he had to say?