Writers of Pro Football Prospectus 2008

10 Jan 2008

Geek's Guide to the NFL Playoffs

Seriously, I'm everywhere this week. Maybe I'll put an addition on the house or something. Here's the first two parts of a playoff dialogue between myself and K.C. Joyner. Think of this as the intelligent antidote to all the phenomenally stupid playoff previews that say Team X will beat Team Y because they "have swagger" and "feel disrespected."

Posted by: Aaron Schatz on 10 Jan 2008

61 comments, Last at 03 Feb 2008, 12:06am by Andrew

Comments

1
by Wanker79 (not verified) :: Thu, 01/10/2008 - 1:51pm

Aaron, the link takes you to the second page (KC's response to you). It's only minor, but just thought I'd let you know.

2
by Craig (not verified) :: Thu, 01/10/2008 - 1:55pm

A director of pro scouting wrote that "phenomenally stupid playoff preview". I think I just lost some respect for Scouts, Inc.

3
by MDZ (not verified) :: Thu, 01/10/2008 - 2:19pm

#2
Scout's Inc. hires a bunch of failed front office and GM types. Randy Mueller was there main guy a few years ago, and his "insight" then makes it no surprise that he built the 2007 Miami Dolphins (or that he gave Joey Porter $20 MM guaranteed). Something tells me that Jeremy Green was only the director of pro scouting because his father was the head coach. Todd McShay is decent on the draft and college football. There's a reason they are Scouts Inc., not actual scouts anymore.

4
by Tim Gerheim :: Thu, 01/10/2008 - 2:42pm

The link is fixed, and now points to Aaron's part.

5
by Jon (not verified) :: Thu, 01/10/2008 - 3:10pm

Of all the people on the Giants to criticize...Steve Smith? You mean the rookie who missed most of the year with a hamstring injury after a promising start? He hasn't had trouble getting open from what I've seen at all. He's been a signficant upgrade over Sinorice Moss since returning, and looks to be a big part of the team's future.

The true whipping boys for the Giants should be Toomer and Moss. Toomer has been dreadful outside the past two weeks. His hands have turned to stone, and he can't get open at all. Moss suffers from Todd Pinkston disease, as evidenced in the SNF game vs. Washington.

I don't think Eli is the one who wants to go deep on every play. That's on the coaches. Eli has played well recently after Coughlin ordered Gilbride to call a simpler game. The other best moments of his career have come in the no-huddle, where he can improvise and isn't being goaded on to throw it deep on every play. When the team struggled, play action was completely forgotten. Shockey was turned into a glorified Mark Bavaro. The WRs, who couldn't get open, ran deep out routes on every play and couldn't catch the ball when it actually arrived.

That more than anything has why the Giants have underachieved this year. I know some people will compare the team's DVOA to its record and conclude that the coaching staff has done a wonderous job. Nothing could be further from the truth. Gilbride's playbook is absolutely maddening, while Coughlin shows blind loyalty to marginal players like R.W. McQuarters and Reuben Droughns.

6
by deshawn zombie (not verified) :: Thu, 01/10/2008 - 3:26pm

Jeremy Green was director of Pro-scouting for Cleveland during one of the worst drafting stretches that any franchise has ever had in any sport. I belive he selected a golden retriever to play tackle at one point because he thought it had good upside. He is not a smart man, and is and was very bad at his job.

7
by deshawn zombie (not verified) :: Thu, 01/10/2008 - 3:32pm

I do realize that he probably wasn't actually that involved in scouting and drafting players, but I believe in guilt by association. Any scouting department in which JG had a prominent position would have to be terrible by definition. The draft record just sort of proves it.

8
by The Original Sam (not verified) :: Thu, 01/10/2008 - 4:43pm

Does KC Joiner really compare Wes Welker and Donte Stallworth to Reggie Wayne? Is that really a good comparison?

9
by The Original Sam (not verified) :: Thu, 01/10/2008 - 4:46pm

Aaron: "Everyone is looking for reasons why the Jaguars or Colts will beat the Patriots, but the Patriots were the greatest regular-season team of all time."

--

Ahem. "greatest regular-season team of the DVOA era" ... I don't think your statistics and say otherwise, and a speculative debate one way or the other can't "prove" anything in that regard. Before we crown their asses, why don't we see if they can in fact win the Super Bowl?

10
by B (not verified) :: Thu, 01/10/2008 - 4:49pm

6: Maybe he was trying to draft Ron Dayne and got confused.

11
by JasonK (not verified) :: Thu, 01/10/2008 - 5:15pm

#9:

How would results in the playoffs have any bearing on whether a team had "the greatest regular season of all time"? As for the pre-DVOA era, there's only one other undefeated regular season on the books, and given the differences in season length and margin of victory, I don't think it's particularly controversial to call the Pats' 2007 #1.

12
by Heef (not verified) :: Thu, 01/10/2008 - 5:37pm

#10 - good thing he didn't end up with Taylor Dayne instead.

13
by Jason (not verified) :: Thu, 01/10/2008 - 5:41pm

"I don’t think it’s particularly controversial to call the Pats’ 2007 #1."

Try checking out the 1962 Packers. They lost only 1 game and had a larger scoring differential than the Patriots.

The Packers outscored their opponents 415-148.

Average GB 62 score 32-11

Average NE 07 score 37-17

**We won't even mention the THIRD undefeated team of all time the 1929 Packers (12-0-1) who outscored their opposition 198-22 and posted 8 shutouts in 13 games

14
by Dean (not verified) :: Thu, 01/10/2008 - 6:03pm

I won't insult everybodys intelligence by linking to it, but along the lines of phenomenally stupid playoff previews, there's an article on Sportsline which actually discusses the fact that some of Jacksonville's players are squat and stocky, rather then GQ Models, as a valid analysis. I'm sure there was a point they were trying to make but it was lost in their stupidity. I'm reminded of the moneyball passage about trying to convince old-boy scouts that "we're not selling jeans."

15
by DGL (not verified) :: Thu, 01/10/2008 - 6:05pm

#13: If what matters in determining regular-season greatness is first and foremost winning games, then any team with a loss or a tie in the regular season is pretty much disqualified from being called the greatest regular season team of all time.

If what matters is something other than winning games then, well, any discussion is playing tennis without a net.

16
by Jason (not verified) :: Thu, 01/10/2008 - 6:12pm

#15

1st- 1929 GB was undefeated as well with a VASTLY better differential

2nd-It is comical to think that 1 play would automatically disqualify or qualify a team as greatest ever. Likewise, NE wouldn't be disqualified as greatest team ever if Balt hadn't taken that timeout and had won that game giving NE merely 1 loss.

17
by MJK (not verified) :: Thu, 01/10/2008 - 6:18pm

Jason,

Trying to point to teams from 1929, or even 1962, and compare them to the 2007 Patriots is like pointing to Manchester United and saying that they are a better football team than the 2007 Patriots.

The rules and nature of the game have changed so much since then that it is essentially a different sport. One doesn't have to say "the greatest regular season of the modern era" because "modern era" is implied.

18
by Jason (not verified) :: Thu, 01/10/2008 - 6:30pm

1962 was only 4 years before Super Bowl 1. I find it pretty arbitrary to say 66 is valid but not 62. Modern era is not implied unless it is indicated.

19
by zip (not verified) :: Thu, 01/10/2008 - 6:35pm

#17 said exactly what I was going to say. Comparing the 2007 Patriots to the 1929 Packers is phenomenally pointless, and no one except a semantics Nazi would think that Aaron is slighting the 1929 Packers with his comment.

20
by zip (not verified) :: Thu, 01/10/2008 - 6:37pm

#18

The 1962 team you're pushing lost a game. We're talking about the "greatest regular season" -- seasons are scored by W-L record.

21
by Jason (not verified) :: Thu, 01/10/2008 - 7:15pm

"The 1962 team you’re pushing lost a game. We’re talking about the “greatest regular season” — seasons are scored by W-L record"

I'm pretty positive Aaron, Bill James, almost any sabretician would clearly admit that points scored/points allowed are better indicators of a team's ability than is its record. Heck most experts for example rank the 15-1 85 Bears ahead of the undefeated Dolphins. Only judging by strictly wins is simplistic at best

22
by The McNabb Bowl Game Anomaly (not verified) :: Thu, 01/10/2008 - 8:14pm

@ Jason-

OK, you're right, it wasn't fair of Aaron to slight the 1962 Packers. They were arguably a better regular season team than the 2007 Pats.

Just to make sure, though, let's put them in a time machine and have the teams play each other. Hmm, the Pats all seem to be a hell of a lot bigger, stronger and faster than the Packers and play a far more sophisticated scheme. Gosh, maybe the Pats are better after all.

What's my point? Only that you can't compare across eras and it's probably not a good idea to try. When Aaron calls the Pats "the greatest regular season team of all time," he means they are the greatest of the teams that can be reasonably compared. Offhand, I'd say that only goes back to the liberalization of the passing rules in the early 80's.

23
by Fan in Exile (not verified) :: Thu, 01/10/2008 - 8:30pm

#22 I hate it when people say that. If that's what he meant he should have said it. He said all time so we have to assume he meant all time. He's the one who compared them to the 1929 and 1962 Packers when he said that.

It's always a mistake to say anything about all time when there are packer fans around because they're so freakin caught up in past glory.

24
by Jason (not verified) :: Thu, 01/10/2008 - 8:43pm

"you can’t compare across eras and it’s probably not a good idea to try"

That's silly. A large reason for advanced stats should be to help compare players and teams across eras.

Saying "Bronco Nagurski, Jim Brown, Walter Peyton, and LT are all good running backs let's leave it at that" adds practically no useful insight.

It's also debatable even if you compare the Pats to merely "modern teams". A handful of teams lead the league in both points scored AND points allowed, a dominant trait NE failed to accomplish for example

25
by brian c (not verified) :: Thu, 01/10/2008 - 8:49pm

Poindexter, does this mean we r(are) the geeks or you 2?

26
by Jake (not verified) :: Thu, 01/10/2008 - 9:41pm

#13"“I don’t think it’s particularly controversial to call the Pats’ 2007 #1.”

Try checking out the 1962 Packers. They lost only 1 game and had a larger scoring differential than the Patriots.

The Packers outscored their opponents 415-148.

Average GB 62 score 32-11

Average NE 07 score 37-17

**We won’t even mention the THIRD undefeated team of all time the 1929 Packers (12-0-1) who outscored their opposition 198-22 and posted 8 shutouts in 13 games"
Math lesson time?
415 - 148 = 267. Divide by 14 and you get 19.07 pt/game difference.
07 Pats? 589 - 274 = 315. Divide by 16 and you get... 19.68 pt/game.
Don't base your argument on a stat unless you've checked it. The Patriots are a better regular season team by your own metric. When you also consider only three games against teams with winning records (versus six playoff teams plus the 10 win Browns this year) and that pesky LOSS, it becomes obvious grasping at straws.
And the 1929 Packers are no more relevant to American Football than rugby is. The forward pass wasn't really even legal in the league and teams played anywhere from 6 to 18 games. There were three expansion teams. Its was proto-football at best.
The team that has an argument is the 1985 Bears and even they just fall short.

27
by Tom (not verified) :: Thu, 01/10/2008 - 10:20pm

Why did the Giants draft Moss and Smith? Eli's main problem is accuracy, and seems to do a lot better with big targets like Burress and Shockey. So they draft two midgets. I just don't understand.

28
by The McNabb Bowl Game Anomaly (not verified) :: Thu, 01/10/2008 - 10:24pm

@ Jason-

There is an article in one of the Pro Football Prospectus books, I think 2006 and I think by Michael David Smith, explaining in great detail all the various eras of football and why it is a fool's errand to try and compare across them, even with the benefit of advanced statistical analysis. It simply can't be done in a valid way.

29
by daltadave (not verified) :: Thu, 01/10/2008 - 10:49pm

I agree that it is futile to compare teams from different eras. The game has changed too much. 1962 teams were allowed 33 man rosters. There was no room on the Packers for punting, kicking or return specialists, they were all regular position players. No team could run from multiple sets or bring in defensive players to cope with specific situations. No roster space for kick coverage specialists. The rules favored the running game in a big way. Pack ran 63% of the time. It was a time when establishing the run really meant something.

30
by Zac (not verified) :: Thu, 01/10/2008 - 11:26pm

I'm a Packer fan, but including the '29 Packers is ridiculous. If you're going to include them, you might as well include the '22 and '23 Canton Bulldogs, who went 10-0-2 and 11-0-1 respectively. In fact the '23 Bulldogs outscored their opponents 246-19, which is a better ratio than the '29 Packers.

31
by Jason (not verified) :: Fri, 01/11/2008 - 12:41am

-Zac you are correct, when stating a team is #1 All Time in regular season then those 23 Bulldogs indeed are included

-#23 Jake there is tremendous irony in that you speak of Math lessons yet it is who who needs the help. GB 62 played 13 games yet you divided by 14!!!! So for some reason you are dividing by 14 when they only played 13 games, I would love to hear your logic for this....

32
by Jake (not verified) :: Fri, 01/11/2008 - 1:08am

Jason they played 14.
Game by game schedule and stats are available at pro-football-reference.
If you don't believe that site, you can find it many other places.
I think we've established your lack of credibility in this argument.

33
by Non Math Major (not verified) :: Fri, 01/11/2008 - 1:23am

re: 31

The Packers were 13-1 in the regular season. Point differential of 415-148 divided by 14 games. (link in name) Jake's math is correct. Why do you insist they played 13 games?

34
by Carlos (not verified) :: Fri, 01/11/2008 - 1:51am

Great article, but I too really have to question:

the Patriots were the greatest regular-season team of all time

Better than the 1991 Skins?

In Favor of Pats:
1. Greatest point differential in history (as far as I know), w/ #1 offense in league
2. Undefeated
3. Historically great offense
4. Very strong defense (4th by points) & ST

In Favor of Redskins:
1. 2d best point differential in history (I believe), w/ #1 scoring offense in league
2. #2 scoring D in the league
3. Great ST (probowl kicker and KR, niether were "reputation" picks as they were both very young)
4. 14-2, with losses coming by 2 and 3 points
5. Most important point: they did this in a 28 team league, w/o the talent dilution that comes from growing the league by 15%!!
6. 2nd most important comparitive poitn: On top of the higher concentration of talent, they also faced a much, much tougher schedule: 11 games against teams finishing at or above .500. The Pats played 8 such games. Aggregate opponent record: Skins 135-121. Pats 120-136. The Pats played a very tough out of division schedule, but their division didn't have another winning team. The '91 Skins had only one losing team in their conference, the 4 win Cards.

Interesting: profootballreference puts both teams at expected W-L of 13.8-2.2.

Finally, Belichick's pouring on the offense regardless of the score will also skew a sabermetricians fave measure of point differential. You can be certain that Gibbs didn't do the same (that's not a beef w/ the Pats! I'm a fan, and I enjoy watching Brady throw to Moss a lot more than I'd enjoy watching Cassel hand off to whoever the 3rd RB is!).

Anyway, I'm not necessarily saying the Skins were better than the Pats, just that I'm not prepared to crown the Pats @sses, primarily b/c of relative strength of schedule and the dilution of talent today. Be a heck of game between those two teams!

35
by Bobman (not verified) :: Fri, 01/11/2008 - 2:03am

Stop. Stop! STOP!

Please take a few deep breaths.

Resume.

Thanks.

36
by Bobman (not verified) :: Fri, 01/11/2008 - 2:21am

If I see the phrase (subscriber only) one more time I am going to crawl through my co-ax connection to the Internet and throttle Mr. Joyner.

Also, based on this discussion, should the Chargers even show up? I think the Colts are looking at a 14-point win (I have a "Denver Playoff Feeling"), but KC and Aaron haven't really said anything positive about them (aside from Rivers' 90+ ratings the last 4 games). The lack of respect makes this Colt fan nervous.....

Then again, the Colts don't often blitz and can't pressure the pocket like they used to, so maybe Rivers is in for a good game.

VERY interesting take Joyner has on Garrard. That would mesh with the "where the hell did he come from all of a sudden?" feeling a lot of folks have, but what a weird way to get "good"--make many boneheaded decisions but get lucky on them ALL. Frankly, it seems a bit of a stretch to me. But if it is true, he will likely crash and burn next year doing all the same stuff "because it worked before" but without the rabbit's foot, or whatever has been helping him.

37
by Jason (not verified) :: Fri, 01/11/2008 - 2:39am

Going by pure point differential is also not completely informative.

For example:

Team A- Outscores opponents 500-300 = +200

Team B- Outscores opponents
300-150

Team B is the more dominant team (doubled the points of their opponents) even though Team A had the higher differential

NE 07- In their games they scored 68.3% of all points and their opponents scored 31.7% (589/863)

GB 62- In their games they scored 73.7% of all points vs 26.3% for their opponents
415/563)

**415/563 is a more "dominant" percentage than is 589/863

38
by Penrose 10,000 (not verified) :: Fri, 01/11/2008 - 5:39am

That Jeremy Green article is fantastic. And look at the comments, as opposed the the FO post comments we all know and love.

JG, you get R-E-S-P-E-C-T in my book... You know what you're talking about, you're staying consistent, and I can't argue any of your points.

Probably not a law student.

39
by Non Math Major (not verified) :: Fri, 01/11/2008 - 10:52am

Carlos, I enjoy your commentary here and understand you're just making the case for the Skins but don't forget this game.....

November 10, 1991

The 9-0 Washington Redskins are ahead of the 5-4 Atlanta Falcons 35-17 after 3 quarters. Redskins QB Mark Rypien throws an 82 Yard TD pass to Gary Clark with 12:12 remaining, 'Skins up 42-17. Shortly after that Rypien hits Art Monk with a 64 Yard TD pass with 10:00 remaining in the game, 'Skins up 49-17. Falcons QB Brett Favre Throws a Pick Six to Redskins LB Andre Collins with 0:52 remaing, 'Skins up 56-17 and game over.

Mark Rypien was 16-31-442 and 6 TD's.

40
by Disco Stu (not verified) :: Fri, 01/11/2008 - 11:17am

I don't get why Aaron keeps bringing up Witten in this article- his crazy day in the first game was 100% about the Mathias Kiwanuka experiment. Kiwi couldn't cover anyone. It was embarassing to watch.

41
by Disco Stu (not verified) :: Fri, 01/11/2008 - 11:19am

That 91 Redskins team was the greatest I've ever seen (I started watching football in 88 or so). They always get forgotten about- prob cuz of the lack of superstars- but for that 1 year they were absolutely amazing. Good job bringing them up for this discussion.

42
by Jake (not verified) :: Fri, 01/11/2008 - 12:00pm

#37 Well its convenient that your preconceived position is the correct one even when you clearly don't know much about the season and change the standards.
Re: 1991 Redskins. The '91 Redskins were a great team but they lost twice. They're not the 2nd best point differential team of all time. The 1999 Rams have a higher one for instance. One loss I think keeps you in the conversation because a fluke night can happen (although the 85 Bears loss I don't think was a fluke and is a sign that if you're doing a time machine comparison they couldn't dominate the game at all today) and the 85 Bears are often seen as the GOAT. A second loss though... you can still be an all time great team but I don't see how you can argue a two loss team was better than the undefeated and one loss teams.
Along with WL and point differential, another stat used by Eddie Epstein in "Dominance" to measure the strength of a team is yard differential (gained vs allowed). The Patriots would be about 1 yard/game behind the all time leaders (79 Steelers), and the 2nd place team (72 Dolphins) and 11/game yards ahead of the new 4th place team (71 Cowboys).

43
by Wanker79 (not verified) :: Fri, 01/11/2008 - 12:18pm

Re: 37

Oh, ok. So now that people have poked holes in your original statement all of a sudden point differential is "not completely informative" and a completely subjective "dominant" point differential is the way to go. So if someone discredits this point of view, are you just going to change tactics again and again until everyone gets bored with your inconsequential argument?

44
by Disco Stu (not verified) :: Fri, 01/11/2008 - 12:38pm

I love how we took a throaway line from a four page article and turned it into the basis for this thread. Gotta love the Internets.

Seriously though- there's a lot of good stuff in these articles. I like Joyner's last post- Phil Rivers IS overrated. SD is prob the cushiest QB spot in the league, and he's only slightly above average. I get the sense that on most other teams he'd be talked about as a near-bust.

45
by B (not verified) :: Fri, 01/11/2008 - 2:01pm

The '85 Bears is an interesting case in comparing them to the '07 Patriots. In the one loss the Bears had, the Dolphins used 3 WR sets and Marino's quick release to overcome the 46 defense, which was a great blueprint for beating the Bears. However, no other team in that year could duplicate the results, but the '07 Patriots are perfectly designed for that kind of strategy.

46
by MJK (not verified) :: Fri, 01/11/2008 - 2:22pm

Jason, I disagree.

For example:

Team A- Outscores opponents 500-300 = +200

Team B- Outscores opponents
300-150

Team B is the more dominant team (doubled the points of their opponents) even though Team A had the higher differential

I don't believe Team B is more dominant at all. To me, dominance is about winning. A team that has a 95% chance of winning every game is more dominant than a team that has a 60% chance of winning every game.

The higher differential put up by Team A implies that they won by more every game. Which in turn means that they are less likely to lose a game on a fluke play or a bad call by a ref, even though Team B has a better point ratio.

Put another way, a team that tends to score about 34 points per game, and give up 21 (and hence tends to win by scores around 34-21) is more likely to win any given game, and hence is more dominant, than a team that tends to score just 14 points per game and give up just 7, even though a season of 14-7 wins will result in a better ratio (twice as many points) than the first team (1.6 times as many points). The second team will lose if they fumble a kickoff, or a batted ball gets returned for a TD, or a ref blows one critical call. The first team will not--they will just win by less.

There's also the issue of reasoning on the average--in generaly, you can't do it according to the laws of probability, unless the distributions involved are certain distributions (which points scored and points allowed are almost certainly not). For example of when reasoning on the average gets you into trouble in this context, consider a team that consistently scores 24 points every game, and gives up exactly 3 points in half its games, and exactly 27 points in the other half of its games. Hence, on average it would score 24 ppg and give up 15 ppg (which would give a 16-game season total of 384-240, which looks pretty dominant), but it would only have a 50% chance of winning any given game and would go 8-8. Hardly a dominant team.

What you really need to do to gauge dominance is to look at the probability a team has of winning any given game, which requires lookingat the distributions of points scored and points allowed per game by a team. You then calculate the probability that points scored will be greater than points allowed (i.e. that the team will win). A higher probability means a more dominant team.

The Patriots this year, I bet, have had a pretty high probability. I haven't done the analyses (it would be quite a bit of work, and probably suffer from small sample sizes--16 events really isn't enough to get a clear picture of a distribution). But I bet the intersection of the points allowed by them and the points scored by them would be pretty small...

47
by Kurt (not verified) :: Fri, 01/11/2008 - 2:55pm

46,

It doesn't quite work that way. Raw point differential doesn't tell the whole story either; a team which scores 100 points and gives up 0 has a smaller point differential than your 500-300 and 300-150 examples, but obviously has a higher expected win%.

I'm not sure how well Bill James' pythagorean theorem applies to football, but under that teorem a baseball team which scores 500 runs and allows 300 has an expected win percentage of .735, and a team which scores 300 and allows 150 has an expected win% of .800. Roughly the same is true for the 34-21 and 14-7 teams - fluke plays and bad calls are already built into those averages, and each of those plays the 14-7 team gives up just means more shutouts in other games.

I don't know how valid the point on distribution is, because I don't know that that's really a good test of quality. No team in real life would give up exactly 3 points in 8 games and exactly 27 points in the others, unless it was a function of the opposition. That's just playing games with numbers.

That being said, I think scoring differential counts for determining dominance. A team which wins 3-0 every week is not as dominant as the 2007 Pats, even though the expected win% based on points scored and allowed is better.

48
by Carlos (not verified) :: Fri, 01/11/2008 - 3:13pm

39: you brought a tear to my eye reminding me of the days when Gibbs had a pair of balls

41: is it a lack of superstars, or just some kind of crazy lack of HOF support after the fact? Monk, Clark and Sanders is about as good as it gets for 3 WRs -- and not to restart a tired debate, but Monk and Clark aren't obviously worse than Stallworth and Swann. Charles Mann, D Green and Wilber Marshall -- 14 probowls and 11 all-pro teams. Jim Lachey and Joe Jacoby were each 5 time probowl tackles. Brian Mitchell is kinda the Raffy Palmeiro of KR/PRs, hopefully w/o the juice. '-)

42: eh, wins. You completely ignore the SOS and dilution of talent points. Even by FO's own measure, the Pats faced only the 19th toughest schedule in NFL this year and ended w/ 13.8 pythagorean wins. Facing tough competition not only has a "primary" effect of increasing the likelihood of loss in that game, it also has secondary effects of putting more strain on your roster. Half the Pats games were against teams that finished with losing records. Their divisional opponents finished 19, 25 and 29 in DVOA. Actually, they're a lot like the '85 Bears when in comes to SOS, as the bears division didn't have another winning team in it.

Obviously the Pats are a phenomenal team, and perhaps they are the greatest team of the post-merger era. I just don't see it as a slam dunk.

49
by Orion (not verified) :: Fri, 01/11/2008 - 4:54pm

Seems to me that point differential is a mostly meaningless statistic in football. There aren't enough games in a football season that you can really say anything without looking at the context.

Example: Team A goes 16-0, easily winning every one of its games by 20 points. Team B goes 16-0, but squeaks by in eight of its games, winning by one point, and blows out the other eight, winning by 40.

Now, Team B has a greater point differential, but I don't think very many people would bet on them beating Team A in the Super Bowl.

Example 2: Team C gets a three touchdown lead, then runs out the clock to win by 21. Team D gets a three touchdown lead, then continues to run up the score to win by 42. Was Team D any more dominant?

Look at this year's Patriots: ran up the score a couple of times early in the season (or so the accusations went), then won several close ones toward the end. How much sense does it make to just add up the totals? This isn't baseball, where such things end up evening out due to the staggering number of games they play -- and where it is never a correct strategy to choose not to score.

(Disclaimer: I'm not expressing an opinion about how this year's Pats compare to other great teams, and I have no problem with anyone running up the score in the NFL -- some tiebreakers depend on points scored, even if it's unlikely that it would every come up.)

50
by xbox (not verified) :: Fri, 01/11/2008 - 5:19pm

Is it really worth debating which team was the "greatest regular season team of all time" of all time?

Prior to Aaron using that expression, I had not thought it was something worth measuring. Probably because the contenders for greatest regular season team of all time went on to win the Super Bowl and so their regular season achievements became secondary.

If the Pats do not win the Super Bowl, I doubt they will get much credit or notice in the years to come for their great regular season.

51
by zip (not verified) :: Fri, 01/11/2008 - 5:33pm

If the Pats do not win the Super Bowl, I doubt they will get much credit or notice in the years to come for their great regular season.

To the contrary, I think they will be as legendary as the 15-1 Vikings from 98 that lost to the Falcons in the NFCCG, both for a ridiculous offense and a postseason failure.

52
by MJK (not verified) :: Fri, 01/11/2008 - 5:35pm

Orion makes a good point when he/she reminds us that "it isn't baseball".

In baseball, there is no clock--just a set number of scoring opportunities, so you try to score as many runs as possible each time your team is at bat. Yes, when you have a huge lead, maybe you don't worry about hitting as well, and yes late in tight games you might vary your strategy somewhat (i.e. bunting, sacrificing, etc) especially in the NL, but for the most part, a team is trying to get as many runs out of each three outs as possible whether the score is 4-2 or 12-5. And the team's likliness to score does not change all that much as the score changes.

However, in football, you don't get a set number of opportunities, but a set amount of time. Hence your strategy and your likliehood of scoring changes dramatically based on the score.

If one must look at point differentials for measuring dominance, maybe a better measure of dominance would be to just look at points scored versus points allowed in the first half, or in the first three quarters...

53
by Travis (not verified) :: Fri, 01/11/2008 - 5:40pm

From Joyner's latest post:

New York was already thin at this position after placing Corey Webster on injured reserve last week, so they could potentially be playing fourth- or fifth-string cornerbacks in this game.

Webster went on IR before the start of last year's playoffs, but is healthy this year.

54
by MJK (not verified) :: Fri, 01/11/2008 - 5:41pm

Kurt,

Obviously, the case I mentioned of allowing 27 points in half your games, and 3 in the other half is a bit extreme. You're right--I was playing with numbers a bit; I picked it as an overly simplistic example to demonstrate why just reasoning on "points per game" could lead to a wrong inference.

However, I wouldn't be at all surprised to see that the distribution of points allowed for some teams was bimodal, meaning that they play lights out defense some weeks, and get torced the others, with little in the middle. There's a lot of things that could cause this. Maybe they are very well geared to stop a particular kind of team (e.g. Minnesota, with a fantastic run defense and a fairly porous pass defense). Or maybe they are a bad defense that plays in an open stadium with lousy weather (i.e. the Jets). Or maybe they are a savy but aging defense where, as long as their offense controls TOP, they are lights out, but once they are on the field too long and get tired they get lit up relatively easily (I think this is the case of this year's Patriots).

55
by Kurt (not verified) :: Fri, 01/11/2008 - 7:06pm

In baseball, there is no clock–just a set number of scoring opportunities, so you try to score as many runs as possible each time your team is at bat. Yes, when you have a huge lead, maybe you don’t worry about hitting as well, and yes late in tight games you might vary your strategy somewhat (i.e. bunting, sacrificing, etc) especially in the NL, but for the most part, a team is trying to get as many runs out of each three outs as possible whether the score is 4-2 or 12-5. And the team’s likliness to score does not change all that much as the score changes.

That's not quite true - the Yankees outperformed their expected record based on run differential for years, in part beacuse the back end of their bullpen was so weak,and those are the guys pitching the last three innings of a 10-1 game (true of all teams, obviously). Certainly it's not as much of an effect as in football, where you can sit on the ball, but it's still there.

No arguments with 54 - honestly, there's only so much you can extract from 16 games.

56
by Jsaon (not verified) :: Sat, 01/12/2008 - 5:05am

At it's most fundamental level, dominance relates to one's ability to score points at a significantly higher pace than one's opponents.

Take 2 Teams

Team A Outscors it's opponents 400-100 an incredible mark of +300

Team B Outscores it's opponents 500-190 for a +310

Jake and MJK contend Team B is the more dominant Team despite the fact that they allowed almost twice as many points yet scored barely 1/4 more points. I can;t imagine most would agree with their assessment

57
by ammek (not verified) :: Sat, 01/12/2008 - 5:37am

This is a depressing thread. Aaron and KCJ had a heap of interesting things to say in their chat, and the folks out there have picked up on one throwaway comment and launched into yet another aimless, irrelevant all-time-greatest-of-all-time dispute.

Can't we leave this to the offseason?

58
by Matt Saracen - QB1 - Dillon Panthers (not verified) :: Sat, 01/12/2008 - 6:58am

Jsaon/Jason: You don't give up easy. I respect your argument about percentages, but I would rather a team that won by a higher amount. But I agree with you that just cause the Pats went 16-0 doesn't make them a lock for #1. They could easily have lost against the Ravens for example, and probably deserved too.

'91 Redskins was a team I'd never considered. My rule of thumb in this kind of argument has always been that teams deserve 1 mulligan since you can always have a bad day, but 2 bad days? Now you are stretching the allowances. But that's just me...

Re #52 MJK: Comparing the scoring per quarter would be interesting. Maybe you could weight the first 2 quarters a little heavier than the 3rd and a lot heavier than the 4th (but only in games where there is Team A has a decent lead at the end of the 3rd). Otherwise the 4th is just as important. Then you could come up with a composite figure and recalculate Pythagorean wins etc.

Does anyone know if there is a site that tracks the quarter by quarter scoring? I had a 5 minute browse but couldn't locate one (unsurprisingly I guess haha).

59
by Matt Saracen - QB1 - Dillon Panthers (not verified) :: Sat, 01/12/2008 - 7:04am

#57 Ammek: Yes I agree. I was quite interested by KC's David Garrard comments. I'm been pimping him and voted him to the Pro Bowl. Maybe I need to watch him more instead of judging just on stats. That's why I like reading KC when I can to get the opinion of a tape junkie, but of course I can't read a lot of his articles (subscriber only, Bobman)

60
by Non Math Major (not verified) :: Sat, 01/12/2008 - 11:22am

re: 58 - link in name for a site that has pro football boxscores going back to the 1920s.

61
by Andrew (not verified) :: Sun, 02/03/2008 - 12:06am

How in the world one thinks a differential of +310 is inferior to a differential of +300 is beyond me. Let alone the fact that a 10 point difference isn't that significant to begin with so your examples show that each team is fairly equal.

To put this into some context here, I used a league average metric similar to ERA+ to measure how great the New England offense was compared to the rest of the league and here's what I got.

(Below 100 is below league average, 100 is league average and anything above 100 is above league average)

2007 NE: 170

Compare that to the 1998 Minnesota Vikings and how their offense compared to the 1998 league and you get.

1998 MIN: 163

I don't have an all-time PS+ list, but I hope I have added a little context to just how dominate New England's offense was.