Writers of Pro Football Prospectus 2008

12 Jan 2009

Aaron Schatz on the B.S. Report: 1/12

Bill Simmons went 0-4 on his picks this weekend, and wants to know if anybody can figure out what's going on in this year's playoffs. Hey, we can! Well, as long as your question does not concern the Arizona Cardinals. Anyway, Bill and I look back at the Divisional round and preview the Conference Championships, plus his usual discussions with Cousin Sal and Michael Lombardi.

Posted by: Aaron Schatz on 12 Jan 2009

21 comments, Last at 14 Jan 2009, 11:30am by ernie cohen

Comments

1
by pawnking (not verified) :: Mon, 01/12/2009 - 5:59pm

Actually, when you consider DVOA, pretty much everything except the Cards beating the Panthers has been unsuprising this playoff year. Even Arizona's first win isn't that shocking when you consider they played it at home. Ditto the San Diego/Indy game.

Keep up the good work, guys. Your Baltimore/Eagles superbowl matchup is gaining you a lot of street cred this year.

2
by Staubach12 :: Mon, 01/12/2009 - 6:36pm

Aaron,

I have to say that you've been great when you appeared on the BS report this season. You can tell Simmons that I never listened to his show until you came on, but now I'm a regular listener. That's right, FO actually brought a new listener to ESPN.

3
by Fink (not verified) :: Mon, 01/12/2009 - 7:34pm

Agreed with Staubach. I'm a huge fan of Simmons and have been reading his columns regularly since highschool (7+ years), but the podcasts are just too long to listen to regularly. I do check out the NFL ones though, generally fast fowarding to the part with Aaron. Go Eagles this weekend! (Section 104 Row 14 Black #20!)

4
by ernie cohen (not verified) :: Mon, 01/12/2009 - 8:32pm

1) Aaron mistakenly identified too many men on the field as the only penalty subject to review. There are several others (I don't know if this is complete):

- illegal touching of a forward pass
- illegal forward pass (QB past scrimmage, or multiple forward passes on a single play)
- illegal touching of a kick
- pass interference (I'm not sure, but I think you can review that the ball was tipped)

2) There is evidence that the PIT-BAL game should be close, but the first two games are evidence of the opposite. If anything, you would expect the average margin to estimate the spread (albeit with a really big confidence interval). So you would expect a 3.5 point spread, plus 3 points for home-field advantage (arguably more). That's a 6.5 point spread, which is pretty big.

It really amazes me that people don't consider PIT winning both regular season games, even by small margins, as evidence that they will win the next one. More people seem to employ the bizarro-world logic that BAL is more likely to win because they lost both, and it's really hard to beat a team 3 times in one season. (The guest that followed Aaron on the show even gave this as evidence.)

3) There are quite a few reasons to like PIT in this game (BAL injuries, PIT health, PIT bye week, recent performance, homefield). I'm still waiting to hear a reasonable reason to like BAL in this game. (The funniest one I've heard was from the guest following Aaron: he said that the Ravens have more free agents, and that they have extra incentive to try hard. I would think the conference championship and a trip to the SB would suffice.)

6
by jonnyblazin :: Tue, 01/13/2009 - 1:26am

I don't get your logic for picking the spread. If two two teams are relatively equal, the home team should get 3 points. Since the Ravens are slightly better than the Steelers according to DVOA, 2.5 would be a good line. 6 is criminally high.

The Ravens are 14-4 against the spread this year with two of those "losses" coming by 1 and 2 points, so you can bet against them at your own risk. But its obvious that the betting public consistently undervalues them. The Steelers, on the other hand, with their massive throng of nationwide supports, tend to inspire homer bets that skew the line, especially at this time of year.

I don't think BAL is all that injured aside from Suggs, who should be playing this Sunday. Rolle might be a question mark as well. The Ravens have an extra days rest to heal, of course. The Steelers aren't injured, true, but one tap on Ben's noggin could change that quickly (I'm sure the Ravens aren't so underhanded as to attempt to take out the most important player of the opposing offense...). As for recent play, I think beating SD by 11 at home really isn't more impressive than beating MIA by 18 and TEN by 3 on the road.

"I'm still waiting to hear a reasonable reason to like BAL in this game."

Well, the Ravens outplayed the Steelers in Pittsburgh this year, and the score was tied at the end of regulation. McClain ran effectively in both games against the Steelers, while fast Willie has never been able to do anything against Ray Lewis and co. his entire career. Mark Clayton has gotten much better over the course of the season at corralling the deep ball, and PIT showed some vulnerability vs. SD defending the bomb, even though SD wasn't able to fully capitalize on some opportunities they had.

7
by Mystyc :: Tue, 01/13/2009 - 4:14am

PIT's margin of victory over San Diego was far smaller than it could have been. The Chargers had (essentially) two garbage-time touchdowns while down by three scores. That said, I can't think of a good reason to seriously favor team - they are that close.

11
by ernie cohen (not verified) :: Tue, 01/13/2009 - 7:19am

1) I agree that if two teams are equal, the spread should be the homefield advantage. And I'm not saying that the spread shouldn't be 2.5 or 3 points, much less that it should be 6. I wasn't trying to pick a spread, period.

What I am saying is that people are twisting the evidence of the first two games in a way that is simply mathematically incorrect. The scores of the two games that they played is *not* evidence that they are equal; if anything, it is evidence that PIT is 3.5 points better. I agree that it is very *weak* evidence - far weaker than, say weighted DVOA. (It is evidence that they are not grossly unequal, but we didn't really need that.)

2) To whatever extent that injuries are relevant, it is clear that the change from the first two games and this one is in PIT's favor. In the first BAL-PIT game, PIT was not only down to their 4th string RB (who admittedly played unexpectedly well), but they also lost Simmons from the OL. On the other side, while Suggs and Rolle might be the only real question marks, it is clear that BAL took many more nicks than PIT did, played in a much more physically draining game, and is without the benefit of the bye week.

3) When I said I didn't see reasons to like BAL in this game, I meant that there is little evidence to believe that BAL will play better, or PIT worse, than we would expect from their previous games. So e.g. FWP's inability to run on BAL doesn't qualify (it's already factored in), but BAL's improved ability to throw deep is arguably one.

12
by DGL :: Tue, 01/13/2009 - 9:56am

If the scoring differential is weak evidence, and stronger evidence (e.g., WDVOA) is available, why cite the weak evidence?

Furthermore, because of the discrete scoring possibilities in football, a score differential is misleading unless it's an aggregate result for a very large number of games (like an entire season). Pittsburgh won the games by 3 and 4 points, but it's more reasonable to say they won by "one score". So they were one score better than the Ravens - which is equivalent to saying "they won".

So while I agree that to say things like, "It's hard to beat a team three times in one season" isn't exactly the pinnacle of reasoned analysis (I saw one article that led with the fact that the Steelers have never beaten the Ravens three times in one season+postseason. Then later it acknowledged that the Steelers and Ravens have never played three times in one season+postseason...), neither is saying the average margin of victory is three and a half points so the Steelers are three and a half points better than the Ravens.

DVOA and WDVOA say they're very close. The Steelers dominated 56 minutes of the game against the Chargers, while the Ravens dominated against the Dolphins; the Ravens and Titans were very close, but turnovers swung it to the Titans. The first two games were very close. The Steelers are playing at home and are the healthiest they've been all year, while the Ravens haven't had a week off since the middle of September. I'd give a slight edge to Pittsburgh - I see it as a 3-4 point game again - but if I were a betting person (and not a Steelers fan) I'd be awfully tempted by the six-point line that's up this morning.

18
by ernie cohen (not verified) :: Tue, 01/13/2009 - 8:24pm

I wasn't the one who cited the first two games as evidence. Aaron did, on the radio chat. That was why I responded to it here, instead of on the open game thread.

While not relevant to the discussion, I would say that based on their recent performance, PIT looks at least 2 points better than BAL. My divergence is mainly because I think the DVOA appraisal of the BAL (+25%) and PIT (-25%) performances against TEN are just out of whack with reality (TEN was about even with PIT, but was mostly destroying BAL), and I put substantial stock in the injuries and the bye week. Add the homefield advantage (+2 or +3), and I'd say a 4-5 point spread.

14
by Chief (not verified) :: Tue, 01/13/2009 - 3:27pm

The "it's tough to beat a team three times in a season" argument was made before the 1999 AFC Championship, too. IIRC, the Washington Post ran an article with a graphic showing how often teams had faced each other a third time, and how rarely one team won all three times.

Overlooked in the article and the graphic: Every time a team had won the first two games (rather than splitting them), the team won third. The Titans continued that trend, beating the Jaguars for third time.

A quick glance at the record books says that since then, the trend has been broken a couple times -- the Vikings over the Packers in the 2004 Wild Card round, and the Giants over the Cowboys in the 2007 Divisional round.

I didn't think to note how many times the 2-0 team won the third meeting during that time, but it was at least 4 or 5 times.

17
by Lord K :: Tue, 01/13/2009 - 4:56pm

Seems like an easy thing to research. What is the combined record of teams that swept the regular season matchup meeting the third time in the playoffs?

It might be useful to divide these figures into eras/decades, to see if there is any trend in any particular direction.

21
by ernie cohen (not verified) :: Wed, 01/14/2009 - 11:30am

By my count (which wasn't too careful), in the SB era, teams have met thrice in a season, after one team swept the first two, 18 times. 12-6 in favor of the sweepers. The lucky losers were

83: OAK over SEA
89: PIT over HOU
94: CHI over MIN
98: ARI over DAL
04: MIN over GNB
07: NYG over DAL

5
by Carlos (not verified) (not verified) :: Tue, 01/13/2009 - 12:01am

Bill Simmons sounds like an 11 year old. With a voice like that, he should be in print journalism. Or at least get a voice coach and learn to speak from his chest. Yikes.

8
by hector :: Tue, 01/13/2009 - 5:24am

Nice use of the strawman on that one, Aaron. I know the Jaworskis and Cosells of the world never called Delhomme a game manager - he's been compared to Favre's style so many times, it's almost boiler plate at this point. I'd like to see where you pulled that "conventional wisdom says Delhomme is a game manager" from; seems conveniently manufactured to me.

Do a web search on "Jake Delhomme" and "gunslinger" and examine the wealth of results. Player and term have been linked for years. I know the FO mindset is "I never heard this before, I must be inventing it!" but it doesn't apply here.

Here's one of the "gunslinger" ties to Delhomme, from ESPN.com.

- http://sports.espn.go.com/espn/page2/story?page=gallosnibbe/081016&sport...

9
by Lord K :: Tue, 01/13/2009 - 6:33am

Given that Howard Cosell has been dead for 13 years, I'd say you are right about that.

10
by hector :: Tue, 01/13/2009 - 7:09am

If you're not familiar with Greg Cosell, you're missing one of the very best football analysts around the game today.

16
by Lord K :: Tue, 01/13/2009 - 3:59pm

Nope never heard of him, but I guess that's what living in the South Pacific does.

13
by Temo :: Tue, 01/13/2009 - 10:53am

I have to agree. I've actually never heard Delhomme described as a game manager.

15
by Bill Barnwell :: Tue, 01/13/2009 - 3:29pm

I don't know about the exact term game manager, but a cursory search of post-SB stories reveals:

During the best quarterback shootout in Super Bowl history, Delhomme morphed from a guy who simply hands off to a master of the passing game. He went from caretaker of one of the league's best running games to a well-rounded quarterback with an unlimited future.

http://sportsillustrated.cnn.com/2004/writers/john_donovan/02/02/delhomm...

In reality, "a guy who simply hands off" == game manager.

19
by hector :: Wed, 01/14/2009 - 5:07am

Well, I'll give you credit, you did find an example, even if it took going back five years to land it. I'll maintain until the cows come home that the overwhelming consensus on Delhomme for years has been he's a wild-eyed gunslinger.

Looking around the internet for a generic "Delhomme scouting report" this is the first thing I found (tsn.ca, which, don't laugh, has good scouting reports on pro athletes):

Delhomme Assets: Tough, fiery, and competitive. Great arm and a quick release. Moves around the pocket well and makes plays on the run.

Delhomme Flaws: Makes some terrible decisions. Misreads defenses and will throw into coverage. Not a great touch passer over the middle. Coming off an injury that could sap some arm strength.

Now, to me, that sounds a lot more like Brett Favre, gunslinger, than Brad Johnson, game manager.

Here's what Greg Cosell wrote about Delhomme earlier this year (and Cosell has forgotten more about football than John Donovan, that SI source, has ever known) . . . tell me if this sounds like a game-manager description:

"Delhomme's overall record as the Panthers' starting quarterback is 46-29. It's tough to argue with a winning percentage of 61 percent. But Delhomme has always been a tough watch. His decision making at times is arbitrary, his accuracy variable. He is not a precision passer, but more of an intuitive playmaker with toughness and moxie. That's fine up to a point, but his performance fluctuates too much week to week."

Playmaker = gunslinger. Just about everyone is seeing the same thing.

20
by hector :: Wed, 01/14/2009 - 5:49am

One final Delhomme gunslinger reference, just for fun (from a year and a half ago):

"Delhomme is 32 and has started to acquire a Drew Bledsoe-like reputation as a gunslinger who racks up a lot of yards, but makes too many game-ruining mistakes."

Sounds familiar? Pro Football Prospectus, 2007. Lead writer, Aaron Schatz.