Writers of Pro Football Prospectus 2008

Most Recent FO Features

FitzgeraldLar04.jpg

» Scramble for the Ball: The DVOA Schism

Mike and Tom try to figure out what kind of secret sauce Arizona is feeding the media to sit at the top of the power rankings and in the middle of our DVOA rankings.

23 Jan 2008

New York's DVOA Streak

by Aaron Schatz

Now that we're down to just two conference champions, there's no point in doing a whole table of "postseason weighted DVOA" with all 32 teams. The Patriots and Giants are the only teams worth talking about, and besides, the weighted DVOA for the Giants still looks ridiculous because they were so mediocre from Week 10 through Week 16.

So let's talk about the Giants and the Patriots -- well, mostly the Giants.

First, here are the ratings for the conference championships.


DVOA: 2007 Week 20
TEAM TOT OFF DEF ST
NE 41% 18% -14% 9%
SD 21% -6% -29% -2%
NYG 48% 10% -46% -9%
GB -40% -25% 12% -3%

Many readers have asked for VOA as well, without opponent adjustments. Here you go, although the adjustment for "fumble luck" remains.


VOA: 2007 Week 20
TEAM TOT OFF DEF ST
NE 29% 9% -11% 9%
SD -22% -11% 9% -2%
NYG 28% 8% -29% -9%
GB -40% -29% 8% -3%

We know the Giants are playing better now than they did before Week 17, but we're still left with the question of how many weeks are worth using to judge the Giants when we look at matchups for the Super Bowl. Does every single game before Week 17 no longer count? Is four games really enough to judge a team? If we want to also look at games before Week 17, how many should "count," and how important are they relative to the last four games? There's really no clear answer here.

As an experiment, however, let's look at just the last four weeks of the season. Now, we're left with one other issue here, which is that the opponent adjustments are based solely on the regular season. New England's performance in Week 17 is being judged as if they played a Giants team that was completely average -- but we think the Giants are much better than that now. So I've gone back and re-done opponent adjustments (for all 32 teams) so they now include every game from Weeks 1-20, rather than just Weeks 1-17.

So, with that in mind, here are the DVOA ratings for New England and New York only over the last four weeks: Four games for the Giants, and three games for the Patriots.


DVOA Weeks 17-20
TEAM TOT OFF DEF ST
NE 44% 45% 1% 0%
NYG 43% 28% -8% 6%

Yes, that's correct. The Patriots have a higher DVOA rating than the Giants even if we only measure the past four weeks. It's really not statistically significant, but the difference becomes more substantial if we take out the opponent adjustments, which are still heavily rewarding the Giants for the very act of staying in the game with the Patriots.


VOA Weeks 17-20
TEAM TOT OFF DEF ST
NE 30% 39% 9% 0%
NYG 17% 24% 13% 6%

None of this is meant to belittle the accomplishment of the Giants over the past four weeks. They've done something extremely rare: put up four straight games with a DVOA rating over 40%. Here is a table showing every team that has a streak of four or more games with a DVOA over 40%, along with the average rating over the streak, and the rating for each game.


Streaks of 4+ Games with DVOA Over 40%, 1996-2007
Team Year Weeks Avg.
DVOA
1 2 3 4 5 6 7 8 9 10 11
NE 2007 1-12* 71% 63% 98% 78% 66% 51% 41% 75% 114% 66% 78% 54%
STL 1999 11-15 70% 82% 106% 54% 63% 43%            
TB 2002 2-6 60% 86% 63% 42% 64% 46%            
IND 2007 3-8* 59% 43% 72% 64% 63% 51%            
KC 1997 12-15 78% 48% 56% 109% 99%              
PHI 2004 10-13 77% 52% 62% 95% 98%              
IND 2004 10-13 75% 73% 98% 64% 64%              
GB 2003 15-18 71% 41% 117% 78% 46%              
OAK 2002 16-20* 59% 42% 69% 72% 54%              
GB 1997 13-16 53% 43% 47% 65% 56%              
NYG 2006 5-8 53% 46% 59% 65% 40%              
NYG 2007 17-20 49% 47% 55% 46% 48%              
*includes bye week

One of these accomplishments sure stands out, doesn't it? Of course, since Week 13 the Patriots have fewer games over 40% (three) than the Giants do -- but two of those three games happen to be the Patriots' two postseason wins.

Notice who else shows up on our list of teams with four straight games over 40%? Hey, it's the 2006 Giants. These wins came primarily against poor teams (they beat Washington, Atlanta, Dallas, and Tampa Bay) but after opponent adjustments, the streak is awfully impressive. What happened to this "hottest team in the NFL" after Week 8? They followed their four-game win streak with a four-game losing streak and ended the year 8-8. Whoops.

If we loosen our restrictions to look at three-game streaks, we'll find other teams that went on surprising late-season runs but didn't finish up with a Super Bowl title. The 2002 Jets are a good example. The Jets snuck into the playoffs at 9-7 by winning their last two games 30-17 over New England (DVOA: 46%) and 42-17 over Green Bay (DVOA: 78%). In the first round of the playoffs, they annihilated Indianapolis by the ridiculous score of 41-0, which had a one-game DVOA of 130%. The streak isn't as long as the Giants' streak, but the wins are far more dominant. What happened to the Jets in the divisional round? Oakland 30, New York 10.

There's one other question about the Giants: Should we have seen this coming, either objectively or subjectively?

Objectively, I don't really see the evidence. There have been plenty of comments in our discussion threads, saying that the Giants' postseason run shows a major flaw in our DVOA formula. None of these comments, as far as I can tell, give any suggestions that would improve things. Any change that is intended to raise the Giants' rating for the 2007 regular season has to work for every other team since 1996. If we really want to use this to improve the ratings, we need to find other teams that had the same markers and were also underrated by the formula. I'm all about the suggestions, and I'm always looking to make the ratings more accurate, but comments that say "The Giants totally prove DVOA sucks" don't really get us anywhere.

(As an aside, people around the Interwebs do seem to have an awfully hard time telling the difference between hardcore statistical analysis of the regular season and subjective preseason predictions that are meant to be somewhat pithy.)

Subjectively, there seems to be a general meme in the press that the Giants really had a great passing game all along, they just needed to cut down on the mistakes. This is a fun bit of 20-20 hindsight, but to how many other mediocre offenses would this statement also apply? Isn't that exactly what happened to the Washington Redskins after Todd Collins came in to replace Jason Campbell at quarterback? How did that work out come playoff time? Don't the Arizona Cardinals have the talent to become a top offense "if they just cut down on the mistakes?" How about the Chicago Bears? The other popular idea is that a team's performance on the road means a lot more than their performance at home, and the Giants were 7-1 on the road during the regular season. As I explained a week ago, that didn't really do much for the 2001 New York Jets.

The Giants have been outstanding for the past four weeks. They are going to go into Arizona and make a game of this thing, no doubt about it. But nothing the Giants have done over the past four weeks invalidates the idea that they were a mediocre, overrated team from Week 1 through Week 16.

Posted by: Aaron Schatz on 23 Jan 2008

153 comments, Last at 07 Feb 2008, 1:41am by Rick

Comments

1
by Cyrus (not verified) :: Wed, 01/23/2008 - 4:52pm

I think DVOA is fine, as is. People just need to accept the fact that a somewhat "average" team through the regular season can catch on fire in the playoffs.

Colts defense last year, Steelers team the year before...

The only difference is the Giants won't win the Super Bowl, because I'm a Patriots fan and refuse to accept that possibility.

"I reject your reality, and substitute my own."

2
by John Morgan (not verified) :: Wed, 01/23/2008 - 4:54pm

I don't think we can even be sure the Giants will make a game out of it.

3
by Gil (not verified) :: Wed, 01/23/2008 - 4:55pm

"Any change that is intended to raise the Giants’ rating for the 2007 regular season has to work for every other team since 1996."

Isn't that a mis-statement? It really has to work for the majority of the teams to make the system more accurate as a whole, not all of the teams.

Aaron, don't be so defensive. If people still can't understand the concept that a model like DVOA can't be 100% predictive, they never will.

4
by Brandon (not verified) :: Wed, 01/23/2008 - 5:02pm

DVOA is fine - I've "monitored" it over the past two/three years and seriously believe in the numbers you're putting out. Anyone who tries to use unsubstantiated reasonings to dismiss DVOA either doesn't understand it or can't beleive a statistical model can be this accurate.

Haven't read much about it but the fact that Shockey has been out the past what, 5 games, which seems to have settled the offense and let Eli have the reigns if you will.

With the "right" person in the huddle calling the plays and not mucking around with the balance of the team seems to be a good thing with this QB.

Loss of Shockey = Giants winning #42?

If that's the case and if I were Shockey, I'd be worried...

5
by pat on the back (not verified) :: Wed, 01/23/2008 - 5:02pm

External validity is a bcomplex...

6
by Gerry (not verified) :: Wed, 01/23/2008 - 5:03pm

I think your chart exemplifies exactly what I have been saying about the Giants.

*Two* such streaks over the past two seasons? Perhaps one such streak would be fluky, but two?

They have it in them to be a good, good team. Why they aren't consistently is very frustrating.

7
by nat (not verified) :: Wed, 01/23/2008 - 5:20pm

To believe in a close DVOA between the Giants and Patriots since week 17, you need to believe that the Giants have changed their nature since week 16, that the Patriots have changed since week 16, and that the Cowboys, Packers, Buccaneers, Jaguars, and Chargers have not changed. Otherwise using season-long opponent adjustments combined with 3-4 game DVOAs doesn't make a ton of sense.

Teams do change. Players get injured or heal. QBs learn new skills and gain confidence. Coaches abandon weak schemes. It's plausible enough to make the game interesting.

Here's wishing the Giants a good game, as good as week 17.

And may the best team (ever) win.

8
by Tom D (not verified) :: Wed, 01/23/2008 - 5:26pm

re 6:

Interestingly, at the end of regualar season, the Giants had really low variance. So they didn't really show the ability to play great this year. Personally, I think the secondary has been playing way over their heads the past few weeks, and I just hope they continue to do so for one more game.

Plus, Eli has actually played how Giants must have expected when they traded for him.

9
by jerry (not verified) :: Wed, 01/23/2008 - 5:39pm

While I understand you are somewhat defensive about the Giants disproving all your mathematical calculations, what you simply cannot measure is, for lack of a better word, heart. The human spirit defies any statistical analysis.

As for the Giants' 2006 collapse that you refer to, how exactly do the massive injuries that team suffered reflect in the DVOA? I don't know if they do or if they don't. If you remove the Giants' starting defensive line from play, then the players remaining obviviously still have to complete the schedule, but their lesser talent and quality are obviously reflected in the statistics. So, until you can come up with some individual metric that accounts for injury and inexperience you will always have some difficulty reconciling reality with your statistics.

10
by Cosmos (not verified) :: Wed, 01/23/2008 - 5:42pm

Who is this "Patriots" you speak of? Bah! Never heard of 'em.

11
by karl, miami (not verified) :: Wed, 01/23/2008 - 5:47pm

i want you guys to make money, but if this is what comes of it - i don't know.

just let the idiot chorus sing and keep doing what you're doing. eli manning has become a new person and the giants got a lot better. they were OK before the new eli, they are good afterwards. exactly how does that reflect poorly on dvoa, etc.?

just let it go, a-man.

12
by karl, miami (not verified) :: Wed, 01/23/2008 - 5:48pm

i'm saying ignore the idiots

13
by GlennW (not verified) :: Wed, 01/23/2008 - 6:01pm

Can't we just live with the fact that NFL football with all of its injuries, constantly shifting player roles and other largely undetectable human variables is patently unpredictable, period? Throw in the fact that you have a limited 16-game schedule to judge a team by, and you have a mess on your hands for the purposes of hard and fast analysis.

I'm absolutely fine with a system like DVOA which measures in realistic terms the quality of a team's individual departments in the recent past, but quite frankly, I wouldn't take the numbers to Vegas. Not that I know of any such system, but when we're left to contemplate whether the "real" Giants are the 4-game, 8-game, 16-game or 18-game variant, let's face it, we just don't have much of a clue as to when/where/how a team eventually comes to gel. There's just not enough constancy in any area, as you might have in a sport such as baseball. That should be the standard disclaimer-- "past performance is no indicator of future success".

14
by admin :: Wed, 01/23/2008 - 6:02pm

Comment number nine is a very good example of the problem.

The issue here is not being defensive about the Giants defying the math. Yawn, who cares. Upsets make sports fun. The issue here is "should we have seen this was coming, and if so why, and what can we learn for the future?"

Now, if someone had told you on January 1 that one of the four wild card teams would make a run all the way to the Super Bowl thanks to heart and the power of emotion, which team would you have picked?

Anybody who does not answer "Washington" is lying. The Giants' improvement in "heart" was just as unpredictable as their improvement in turnover margin. So that doesn't help us figure out how we can better predict which teams will make the Super Bowl in future seasons.

15
by Dan (not verified) :: Wed, 01/23/2008 - 6:06pm

I think that the giants have significantly improved. What I think most people are complaining about is not DVOA's output of the regular season giants being an average team over the whole season, but rather that you implicitly assume that team's "quality" stays the same and therefore you can use the entire season DVOA to declare a huge mismatch in a playoff game.

16
by erik fast (not verified) :: Wed, 01/23/2008 - 6:09pm

RE: GIANTS and DVOA: The DVOA seems to work fine. So it didn't predict the Giant's run, who cares? They are who we thought they are! A mediocre team with a mediocre QB. It just so happens that they are playing way "above the rim" right now (sorry for the mixed metaphor, but I have wanted to use that phrase for years).
RE: #9: Heart and human spirit? That is a bit simplistic. Heart and human spirit will help a team be better, but not demonstratively better. As many have pointed out, the reason for their recent improvement is probably a combination of other teams falling off at the end of the season/playoffs and the lack of Shockey (big distraction to Eli even if I do like the guy alot).

17
by GlennW (not verified) :: Wed, 01/23/2008 - 6:13pm

> To believe in a close DVOA between the Giants and Patriots since week 17, you need to believe that the Giants have changed their nature since week 16, that the Patriots have changed since week 16

Good point. Never mind the Giants; the Patriots are a substantially different team than they were at midseason, especially on defense (the dropoff on defense being more significant considering the weather changes and the associated but expected decrease in offense). I know this sounds kind of strange when you're talking about a dominant 18-0 team, but I sense that the Patriots are almost staggering to the finish line (again, relatively speaking) and would be more than thrilled to come away with a nail-biting win in the Super Bowl. The Pats have still been great towards the end of the season but they've been dodging bullets for the past two months.

18
by mrh (not verified) :: Wed, 01/23/2008 - 6:18pm

One of my pet peeves about this talk about the Giants is the notion that they were 7-1 "on the road" in the regular season. They were 6-1 "on the road" as the term is commonly used and 1-0 on a neutral field.

Although it's unlikely that they would have lost to the Dolphins in Miami, they didn't play there. Playing at another team's home is harder than playing on a neutral field.

Twice in this three-season run of making the playoffs the schedule makers have helped out the Giants, once by giving them an extra home game (and one less road game) and once by giving them one less road game. Would they have made the playoffs anyhow both seasons? Maybe, perhaps probably, but that doesn't change the fact that they have had an easier schedule over the last three years than any team.

19
by jd (not verified) :: Wed, 01/23/2008 - 6:19pm

Dude did you read the second half of comment nine? Because it would answer your question about what happened to the hottest team in the NFL in 2006 after week 8.

20
by ammek (not verified) :: Wed, 01/23/2008 - 6:19pm

9 - Yes, and it's precisely because "heart" and "desire" and "moxie" et al are immeasurable that there isn't any point bringing them up. The discussion can only descend into breast-beating and nauseating exchanges of unfounded insult.

That's not to say the Giants haven't shown "heart"; they are clearly on the same page. The surprise is really that this is happening to a team run by Coughlin and Gilbride, who have a reputation for being, let's say, stubborn, and were reported to be losing the players as early as week two (or, if you prefer, towards the end of last season).

21
by GlennW (not verified) :: Wed, 01/23/2008 - 6:21pm

> The issue here is “should we have seen this was coming, and if so why, and what can we learn for the future?”

I would humbly submit that the answer is no, forget it, and don't expend any energy in trying to figure it out. No system is ever going to predict the postseasons of the 2003 Panthers, 2005 Steelers, 2007 Giants et al. As I understand it you're very much in the "tweak" stage with DVOA, and teams like the Giants represent sea changes which simply will not compute, at least not without a hammer and tongs...

22
by footballprofessor (not verified) :: Wed, 01/23/2008 - 6:26pm

I guess Dow Theory applies to football:

Trends exist until definitive signals prove that they have ended.

The Giants' trend was still in effect, and there were no definitive signals to prove that it ended, so they kept winning. Easy enough. The hard part can be identifying the signals.

23
by jd (not verified) :: Wed, 01/23/2008 - 6:39pm

The reason everyone is jumping all over this is that first your writers made absolutely idiotic predictions about how the giants would only win 2 games this year based on nothing other than a heartfelt dislike of Coughlin and Eli Manning. Honestly reading the off-season and pre-season Giants coverage on this website made me wonder if Coughlin stole your lunch money or something. Second you guys described the Giants - Bucs playoff game as a huge mismatch favoring the Bucs, which based on your regular season stats may have been true but based on talent level was a ridiculous statement. The Giants have drafted well the past few years and have also done well with free agents and other pickups. They are a talented team. The fact is the Giants play a bunch of talented first and second year guys who got better as the season went along and they were getting used to new offensive and defensive coordinators. Then the emergence of Bradshaw, the loss of Shockey and the return from injury of Steve Smith (who you guys described as merely a system guy in your book) gave Eli the opportunity to start reaching his potential. Basically a young talented team is peaking at the right time. And you guys completely missed it because all you care about is how they played two months ago.

24
by wbenetti (not verified) :: Wed, 01/23/2008 - 6:43pm

I think that the only way to determine exactly how much the Giants have improved is to do a play by play look at each of their drives (both on offense and defense).

It seems to me like there were several key plays that kept drives alive or gave them large chunks of yardage. It feels less like "they were teh awesum!" and more like they were very fortunate at given points during the past 2 games. Penalties moving them into scoring range, etc.

The team has clearly improved, but aside from the game vs NE at the end of the season there were no statistical indicators that I could see that would point to this sort of turnaround. I say this as a Giants fan who watched almost every game of the year live, and re-watched several more.

They looked bad versus WAS and while they won versus an inferior BUF team, they didn't look great while doing so. The weather was a factor in both games, I admit, but regardless; the team should adapt to the conditions.

25
by JQM (not verified) :: Wed, 01/23/2008 - 6:50pm

As a Giants fan, as well as a devoted reader and believer in DVOA, I'll vote with those that say no adjustments should be made for this season's Giants team. DVOA's assessment of the team's performance those first 16 weeks was fair. I watched every minute of every game and never could've seen these last 4 weeks coming. How is one supposed to predict Eli will suddenly start playing better? If his improvement is a result of better playcalling, how is one supposed to predict Gilbride will suddenly stop being a dunce? What are the odds that at the same time things abruptly click into place for Eli, it happens to Corey Webster as well (If indeed things have clicked into place for both of them and this isn't a fluke)? All I can say is they are playing like a different team right now, and I'm extremely thankful. I can only hope the sudden improvement carries over to next season the way Indianapolis' defensive transformation in last year's playoffs carried over to this season.

26
by Bele (not verified) :: Wed, 01/23/2008 - 6:54pm

You're almost always going to have outliers in a statistical analysis. While you can usually refine models, it's rare to discover "hidden" variables that allow it to account for a few extreme outliers while still accurately accounting for the bulk of the variance. Anyway, if your model was perfect (which it never can be), you'd always know who is "best" and who would win and that would be pretty boring. The fact is that the Giants have been a different team lately, but there is very likely no objective quantifiable variable(s) that you could add to the model that would have predicted this dramatic change.

27
by MC2 (not verified) :: Wed, 01/23/2008 - 6:55pm

I know it's usually a stupid idea to compare teams from different sports, but with all this talk of how the Giants are "on fire" and how they're going to play a great game in the Super Bowl ("no doubt about it"), I can't help thinking back to last fall, when a couple of my friends said that I was "CRAZY" for predicting a Red Sox sweep vs. the "red hot" Rockies.

28
by Otis Taylor 89 (not verified) :: Wed, 01/23/2008 - 7:00pm

I think what explains the NYGs this year is this: They are a cold weather team, but not a bad weather team. They played their worst when there was wind, rain or both. When it's just cold, they are usually the bigger, tougher team. I think you can throw out two games, MIN and TB. MIN was just a bad game that happens to teams, TB was against the most overrated team of the '07 season.

29
by Pat (not verified) :: Wed, 01/23/2008 - 7:02pm

Wait, wait. So, I don't really care about "could we have seen this coming?" etc., blah, blah, blah.

The interesting thing from this article to me isn't "could we have seen this coming." It's "The Giants have played at the same level as the Patriots since Week 17."

The "without opponent adjustments.." part I don't really agree with - the best way to remove the effect from being boosted by playing New England close is to remove the New England game.

NE, Divisional: 73%
NE, Championship: 41%
NYG, Wild Card: 55%
NYG, Divisional: 46%
NYG, Championship: 48%

Obviously, New England's going to be better, due to the Jacksonville game, but it's not a lot, and wow the Giants have become suddenly consistent at performing at a 40-50% DVOA level.

I actually think the top end of DVOA is probably a little artificially spread out - in all ranking systems, it usually is due to the fact that you can only test teams against each other - so the Giants performing anywhere near the vicinity of what the Patriots have been playing at for most of the year really makes me think they might have a shot at it.

Then again, I don't know why I would think they wouldn't - New England's never won a Super Bowl by more than 3 points.

30
by JQM (not verified) :: Wed, 01/23/2008 - 7:07pm

#18 Are you arguing that the Giants winning 10 games (9 by your standards) on the road isn't significant, but that playing 2 less road games in 3 years gives them an enormous advantage over the rest of the league? I agree that the Giants road record is close to meaningless, and has been way overhyped. I also believe that if you use opponents' DVOA or winning percentage, you'll see their schedule wasn't even close to the easiest in the league over the past 3 years.

31
by GlennW (not verified) :: Wed, 01/23/2008 - 7:11pm

> You’re almost always going to have outliers in a statistical analysis.

I think the biggest problem I've had with statistical analysis of sports in general (and the same problem I've had with baseball sabermetrics, although to a lesser extent) is the implied precision ascribed to the system. Typically I've found that this assumed precision is more a misapplication by the users of the system, not of its creators, who generally have a better understanding of the underlying probabilities involved.

With football teams operating in a complex system but only over 16 regular season games, we're talking about large margins of error being inherent, and that should be well understood going in. I'm not even sure that the Giants are a major "outlier", as opposed to a nominal one which has only been bumped up a deviation or so over the past four games.

32
by Will Allen (not verified) :: Wed, 01/23/2008 - 7:19pm

It amazes me that Green Bay made it into overtime.

33
by Will Allen (not verified) :: Wed, 01/23/2008 - 7:30pm

Just to clear up any misperceptions regarding what the FO statistical model projected for the Giants before the season, here it is, from PFP 2007.

Mean projection: 7 wins
0-4 wins: 14%
5-6 wins: 29%
7-8 wins: 31%
9-10 wins: 19%
11-16 wins: 6%

34
by Sergio (not verified) :: Wed, 01/23/2008 - 7:32pm

I see what's happening. Aaron's trying to rile up the Giants fans in order to invoke the FOMBC...

re: 24

"I think that the only way to determine exactly how much the Giants have improved is to do a play by play look at each of their drives (both on offense and defense)."

Is my sarcasm meter broken?

35
by GlennW (not verified) :: Wed, 01/23/2008 - 7:35pm

> It amazes me that Green Bay made it into overtime.

This is an example of DVOA being a superior measure of performance to the final score (but then again simply your eyes would have told you this). Green Bay hit on one exceptionally long TD play, had another TD set up by a long kick return and a major penalty, and had their final FG created from a recovered fumble on a Giants' INT return. Meanwhile in between fumbles the Giants were missing makeable FGs. The Packers were absolutely atrocious in the second half especially...

36
by sippican (not verified) :: Wed, 01/23/2008 - 7:39pm

I'm a Pats fan. That has nothing to do with the following assessment:

I cannot see any way the Giants win.

37
by GlennW (not verified) :: Wed, 01/23/2008 - 7:45pm

> I cannot see any way the Giants win.

Well, the evidence to the contrary here would be the 38-35 result in a game in which the Giants committed the only turnover as well as the serious breakdown in coverage on the long Moss TD. I'm not convinced that we'll see something like this game again, but at least we did once, so we know it's remotely possible. The counter would be "well, that was a regular season game and the Patriots weren't really fully prepared or motivated", but that's not the kind of game I saw.

38
by The McNabb Bowl Game Anomaly (not verified) :: Wed, 01/23/2008 - 7:52pm

Re: 33

So according to the official FO predictions, the Outsiders said the Giants had a better chance of winning 9-10 games than winning 0-4 games, and a better chance of winning 7-8 games than winning 5-6 games.

I would never have thought that from reading the comments Giants fans have posted here about the FO predictions.

39
by Scott de B. (not verified) :: Wed, 01/23/2008 - 7:54pm

I can see the Giants winning. I thought the Patriots could win in 2001, also. I don't think their chances are great, but they can win. I gave the Patriots a 35% chance of winning SB 36. I'd give the Giants a 25% chance of winning SB 42.

40
by Pat (not verified) :: Wed, 01/23/2008 - 7:58pm

#38: Yeah, it's amazing how so many Giants fans ascribe a preseason comment by one of the writers to all of the writers. After all, one writer can be biased for or against a team, right? DVOA is the only 'unbiased' prediction the site has. Everything else is subjective.

The "one writer" was Bill Barnwell.

He's a Giants fan.

41
by ChrisFromNJ (not verified) :: Wed, 01/23/2008 - 7:58pm

#38:

It wasn't so much the actual numbers as Aaron's preseason prediction that the Giants would undergo a total meltdown and get the #1 pick in the draft. That's really where most of this crap is coming from, I fear.

42
by Dr. Wayne Pitcher (not verified) :: Wed, 01/23/2008 - 8:03pm

I agree that the Giants can win. This does not seem to be a mismatch on the order of Bears-Patriots or 49ers-Chargers. The close game in week 17 seems to bear that out, and the general perception that the Giants are not huge underdogs is seen in the 12-point line for the superbowl (remember, the line on the Niners-Charges superbowl was 26 and the Niners pounded the Chargers in the regular season!).

Does anyone have the DVOAs for the Patriots-Giants game in week 17? I can't seem to find them (or rather, I can't find the Pats' DVOA), and those results may be helpful in this disucssion.

43
by GlennW (not verified) :: Wed, 01/23/2008 - 8:07pm

> Does anyone have the DVOAs for the Patriots-Giants game in week 17? I can’t seem to find them (or rather, I can’t find the Pats’ DVOA), and those results may be helpful in this disucssion.

I'd be more interested in straight VOA, just to quantify how the game was played head-to-head, as that's all that matters now. I'm assuming the Giants outfared the Patriots in DVOA, given the system-perceived quality of opponent.

44
by Aatrouss (not verified) :: Wed, 01/23/2008 - 8:34pm

The reason for the Giants secondary playing better than the regular season is mosly due to Corey Webster, he was a second round pick and it usually takes a couple of years for dbs to "get it" it just so happens that he "got it" starting in week 16 against Buffalo.
Same for Eli, if you compare Eli and Rivers, it took them 4 years to "get it", and they are both playing better. I read somewhere that the qb class of 04 is ahead of other qb classes in term of playoff wins. So 4 years seems to be the floor for reaching potential.
Somebody says that The Giants are a mediocre team with a nediocre qb playing above the rim, why can't we turn it around, they are a good team with a good qb who were underachieving during the season. This is a team that went to the playoffs 3 straight years and without massive injuries could have beaten the Eagles in the first round playoffs last year, they lost by a last minute fg, and guess who took them the length of the field to tie the game? You got it, their mediocre qb.
The question is how do you model that?
Your modeling has to account for reaching potential, a cb needs a couple of years, a receiver needs a couple of years, a qb needs 4 years. And that is especially true for young teams and the Giants are the third youngest team in the NFL.
Young teams are more prone to being inconsistent than veteran teams, and that maybe what your model is missing.

My 2 cents.

45
by MarkB (not verified) :: Wed, 01/23/2008 - 8:48pm

"Yes, that’s correct. The Patriots have a higher DVOA rating than the Giants even if we only measure the past four weeks. It’s really not statistically significant, ..."

Whoa there. If you set a level of statistical significance, and your test doesn't reach it, then you cannot say that one value is higher. That's the whole point of the test and the significance value. If you live by the test, you need to die by it. If you actually did a statistical test of the hypothesis that the Patriots' 44% is larger than the Giants' 43%, and it came back negative, than 44 is not larger than 43 - they are effectively the same.

46
by Kyle (not verified) :: Wed, 01/23/2008 - 8:55pm

Re #38

You need to return to the preseason subjective predictions from FO.com to understand the context of more recent criticisms from Giants fans here.

Giants fans, by and large, cried foul at the FO.com writing staff becaues they abandoned their own statistics in favor of subjective analysis. When ranking the units in the NFL, they threw out their statistics in favor of media talking points like "Petitigout left, the Giants have no LT, the #2 OL from last year is now the #25 OL in the NFL".

Instead of looking at their DVOA-fueled predictions for wins this season -- which pegged the Giants as, for all intents and purposes, an 8-8 football team -- the writing staff predicted a god awful season. The head writer declared them the worst team in the NFL. Many other writers foresaw a 5-6 win season max.

Giants fans, contrary to the perception some "DVOA is Religion, It Never Fails" fanatics created, were not irrational homers in their critique of FO.com's preseason predictions and rankings.

Giants fans asked a simple question:
Why are the Football Outsiders ignoring their own statistics solely in the case of the NYFG?

47
by NewsToTom (not verified) :: Wed, 01/23/2008 - 9:29pm

Re #46
I don't know; where are all the New Orleans Saints fans complaining about the Saints ending up mediocre when everybody on FO apparently thought they'd outperform their projection and be good again?

48
by Eddo (not verified) :: Wed, 01/23/2008 - 9:31pm

46 (Kyle) -
Giants fans asked a simple question:
Why are the Football Outsiders ignoring their own statistics solely in the case of the NYFG?
But they didn't ignore them solely in the case of the Giants. DVOA picked the Buccaneers to win their division, but the writers were skeptical. DVOA picked the Packers to get a first round bye, and some writers were skeptical. In fact, every preseason the writers have a "which team will outperform/underperform their projection" segment, where they are encouraged to look beyond DVOA.
Remember, DVOA is not the be-all end-all. FO constantly points out that it needs to be used as a tool and that there are some things DVOA cannot accurately measure. Therefore, when making their individual preseason predictions, the Outsiders are never going to just blindly use DVOA - or else they wouldn't need to write any articles, they would just publish the numbers and be done.

49
by Pat F. (not verified) :: Wed, 01/23/2008 - 9:33pm

Re: 46
I am going to direct you, and any Giants fans who feel that "the Football Outsiders [are] ignoring their own statistics solely in the case of the NYFG," to the actual 2007 staff predictions (linked). It contains an entire section of staff members "ignoring their own statistics." Teams for which the FO staff ignored their own statistics when making predictions:
Dallas
San Diego
Cincinnati
Houston
Indianapolis
San Francisco
St. Louis
Minnesota
New York Giants
Green Bay
Jacksonville
Oakland
Washington
Atlanta

That is 14 teams, or about half the NFL. By the logic of your post, fans of any of these teams ought to be up in arms.

Even if we single out teams the staff predicted to fall short of their statistical projection, more picked Green Bay (4) than the Giants (3). There is not, nor has there ever been, any anti-Giants bias amongst the staff.

Just for fun, I'm going to print the Green Bay comments:
Ben Riley: Green Bay. The projection system must have access to some of da kine stuff, because there’s no way Green Bay is winning 9.5 games. Or 7.5 games.

Michael David Smith: Green Bay. I have a bad feeling that Brett Favre is going to be really bad this year.

Tim Gerheim: Green Bay. I don’t buy the defense being strong enough to carry this offense toward 10 wins. The running game , “led” by Vernand Morency, smacks of the 2005 Cardinals.

And unquestionably the most hilarious one (in retrospect of course)...
Vince Verhei: Green Bay. Do you realize that according to the projection system, the Packers are not only going to win the division, but get the number TWO seed, a first-round bye and a home playoff game? Yeah, none of that is happening.

50
by patriotsgirl (not verified) :: Wed, 01/23/2008 - 9:36pm

46: Giants fans, contrary to the perception some “DVOA is Religion, It Never Fails” fanatics created, were not irrational homers in their critique of FO.com’s preseason predictions and rankings.
Predictions based on personal predeliction are often silly and completely subjective, and sometimes flat-out wrong. That's kind of the point of a prediction; if we just wanted to see what the numbers projected, the FO folks could just post some data and walk away. I suspect it'd make their lives easier.

That aside, it's not "just" the Giants who got singled out. Out of 16 writers who contributed to the category "Team Most Likely To Fall Short Of Its PFP Projection (besides Tampa Bay)", 3 (Schatz, Giants fan Barnwell, and Rose) picked the Giants.

Four picked Jacksonville and four picked Green Bay. (One picked Washington - and, of course, Tampa Bay was slighted in the headline.)

And did you see the FO staff's criticism of Del Rio? Which I agreed with - and was also completely wrong.

Link in my name.

51
by patriotsgirl (not verified) :: Wed, 01/23/2008 - 9:39pm

48 and 49 beat me - too slow!

52
by Glenn (not verified) :: Wed, 01/23/2008 - 10:09pm

The Giants have all the "heart and human spirit"? Please stop disrespecting Rodney Harrison!

53
by Pat (not verified) :: Wed, 01/23/2008 - 10:23pm

Why are the Football Outsiders ignoring their own statistics solely in the case of the NYFG?

OK, ignoring the fact that you've already been shown to be wrong regarding "solely." Very wrong, in fact. Anyway. Why were they ignoring the statistics in those comments regarding the Giants?

Because the question was "which team is most likely to fall short of its PFP projection." Any answer to that question is going to ignore their own statistics. It has to. The question is "which team are the statistics wrong about?"

A lot of the response to that was based on the fact that the "team" was going to fall apart, possibly due to Coughlin. That, to me, is where the mistake is: I think sportswriters and fans read waaaay too much into the 'locker room drama' that sportswriters invent for storylines.

The interesting thing about this is that it's easy to look at the Giants season and see how they could've ended up one of the worst teams in football. In the latter half of the season, Eli was pretty darned bad, Burress was frequently hurt - without the pass rush and defense which frequently seemed "lost" early in the season, this team could've easily been in the 2-4 win range.

Really, a lot of it comes down to the fact that Spagnuolo is a very good DC. I'll admit to being way wrong on that one (dunno how he sucked so bad as a linebackers coach, then).

54
by MC2 (not verified) :: Wed, 01/23/2008 - 10:36pm

God, after all this time, you nerds still don't get it, do you? Nobody is ever going to take your statistical mumbo-jumbo seriously, unless you start adding in a dummy variable for "heart", so get to it. And while you're at it, cook up some dummy variables to account for all the other important factors that you currently ignore, like "moxie", "swagger", and most important of all, "clutchness".

Any true fan knows there's a reason for every upset, and if guys like Skip Bayless and the "Around the Horn" gang can figure them out (often just hours after the upset took place), then you damn well better include them in your model, if you ever want to have a 100% successful prediction rate. Until then, you're just spittin' in the wind.

55
by Toast Patterson (not verified) :: Wed, 01/23/2008 - 10:53pm

I don't think the Giants run shows that their is anything "flawed" with DVOA. (And I hope no one thinks DVOA is foolproof either because that would be insane).

The Giants have shown that teams do not remain static throughout the season and that teams and players can get better or worse as the year progresses.

For the Giants, the horrible WAS game was a real tipping point. Following that game, they have done a better job balancing the run & pass, alleviating some of the pressure (both physical and mental) on Eli. Also, after the WAS debacle the Giants appear to have simplified the option routes in their passing game and Eli's completion % has gone up.

Also, following the WAS game they started incorporating Bradshaw and Smith more in the offense. Bradshaw is a very good runner whose style meshes well with Jacobs'. And Smith may finally be the legitimate 3rd receiver that the Giants have been looking for for years.

The D has played better lately too. This has coincided with old man Sam Madison getting hurt and Corey Webster getting his starting role back. Webster looked like a bust earlier in the year but for whatever reason he has been solid since he replaced Madison.

All of these factors have contributed to help make the Giants a better team lately.

I think they will need to play flawlessly and have to hope that the Pats will slip up a couple of times for the Giants to win. But that doesn't mean they won't be able to compete like the '85 Pats or the '00 Giants.

I don't think that right now the whole season DVOA tells the story for either of these two teams. The Giants are playing at a higher level than they were in weeks 10-15 when they looked really cruddy. Likewise the Pats were otherworldly in weeks 1-11 but they have looked merely very good since Colvin went down against the Iggles.

All told, these teams, today, are a lot closer than many Pats fans would like to admit. It should be a pretty good game in 11 days.

56
by Kyle (not verified) :: Wed, 01/23/2008 - 11:08pm

Meh. Apologies for hyperbole in using the word "solely". It was a shortsighted and poor word choice.

Amusingly, however, that word was the only part of my post argued. Granted it proves an irresponsibly strong statement incorrect, but the rest of the post's intent and message was either ignored or implicitly agreed with by pointing to other teams and saying "hey look, you're wrong... because they ignored DVOA regarding the Saints/Bucs/Cowboys/Etc".

To reiterate my point, it was that Giants fans did not yell at the FO.com crew for a bias against the team or to criticize their statistics, but rather to encourage them to actually use their statistics instead of falling into mainstream media traps.

57
by JasonK (not verified) :: Wed, 01/23/2008 - 11:10pm

Really, a lot of it comes down to the fact that Spagnuolo is a very good DC.

Yes, Spagnuolo is a big contributing factor (and plenty of Giants fans will be cracking a cold one to celebrate if ATL & WAS both make their HC frontrunners official this week). But the Giants also got unexpectedly strong performances from lots of players:

-Strahan returning from his holdout and being (IMO) the second-best lineman on the team (after Tuck);
-Sam Madison defying Father Time and having a damn solid (and mostly injury-free) season;
-Aaron Ross developing into a solid starter much more quickly than anticipated;
-Getting good contributions from the depth at DB (Dockery, Mc1/4ths, Webster, Michael Johnson);
-Toomer being more effective than anticipated coming off his ACL surgery; and, perhaps most importantly,
-Dave Diehl being a better OLT than Luke Petitgout was.

Pre-season, there was no rational reason for a computer or a person to expect that any of the above would be true.

As for DVOA and the Giants, generally, count me among those who think it's a hell of a lot of fun to be the Cinderella team in the Finals. DVOA's regular season observations jive pretty well with my subjective observations of the team. They were mediocre. But they're not playing that way right now, and that's what makes this whole thing so much fun-- that's why you root like heck for you team to grab that last playoff spot! Because, whether you're a person or a statistical projection system, you never really know when things are going to change. Predictable sports are boring.

58
by Chris (not verified) :: Wed, 01/23/2008 - 11:43pm

57- There was no computer that picked the Giants to be 10-6, but I did :)

I think I get a big old " told ya so" for Michael Vick, and another big " I told ya so" for the 10-6 Giants.

Of course, I was wrong with the Eagles/Pats super bowl, but I stand to make thousands on my NFL futures with the Pats.

59
by JV (not verified) :: Wed, 01/23/2008 - 11:47pm

The week 17 game can be spun a lot of ways for both teams. The Giants played tough and gained momentum from that while the Patriots had 7 scoring drives to the Giants 4 (not including the KO return TD) as well as a 36/24 TOP advantage. Is it too simple to say that red zone performance and special teams will govern who wins the SB and how close the final outcome is? If #17 thinks the Patriots are "staggering" to the finish line, then his expectations seem a tad unreasonable. Brady threw two picks in the second half against SD, but to hold the ball for 22 minutes in the half, including the last 9:13 is as dominant a football performance as you will ever see. The Patriots also held the ball for > 32 minutes against Jacksonville and significantly outrushed them, essentially beating them at their own game. Other than Brady's picks against SD, I think they've got a pretty good winning formula going. It would not surprise me if this team is a little burned out from having to match everybody's best effort week after week. I would be surprised if the they did not have very high intensity for the Super Bowl knowing that it's finally over after that.

60
by footballprofessor (not verified) :: Thu, 01/24/2008 - 12:49am

Folks, folks, DVOA doesn't deserve this sort of bashing. Yes, the Giants have outplayed their projected abilities thus far, but don't forget that each of the three teams they played in the postseason were already collapsing! Tampa, Dallas AND Green Bay were vulnerable in the very week the Giants played them.

Click my name for the article.

61
by Jarrod Bunch (not verified) :: Thu, 01/24/2008 - 12:53am

You guys are the most biased, hypocritical so-called football analysts in existence.

"Yes, that’s correct. The Patriots have a higher DVOA rating than the Giants even if we only measure the past four weeks."

Really? 44% vs. 43%. By any valid measure of statistical analysis (as FO claims to have as a competency), this is an insignificant difference. How about writing, then, something such as:

"yes, that's correct. The Patriots and Giants have played at the same high level during the past 4 weeks."

But that's ok, since the Giants really only have such a high rating because of their difficult schedule:

"...but the difference becomes more substantial if we take out the opponent adjustments, which are still heavily rewarding the Giants for the very act of staying in the game with the Patriots."

Isn't that the entire point of DVOA? To adjust accurately for opponent adjustments so we can compare the metric for 2 different teams? Or do you only refer to opponent adjustments when you're trying to argue that the Giants are a terrible team?

62
by seth (not verified) :: Thu, 01/24/2008 - 1:28am

jason k(57)-
one quibble-the one "surprise" about the 2007 giants that i think is not true is the play of david diehl. diehl stepped in at left tackle for petitgout last year and acqitted himself reasonably well. because he was a guard moving over to left tackle there was an automatic reaction thinking that he was going to be overmatched, but the tape and his play from 2006 never indicated that. luke petitgout was a decent left tackle, but certainly no star, and you have to give the giants credit for recognizing that the younger, healthier, and less-expensive diehl was a solid replacement.

63
by AHBM (not verified) :: Thu, 01/24/2008 - 1:30am

Toast Patterson (#55) put into words just what I was thinking as I read the article. There is one thing that DVOA and every other possible statistical method are simply incapable of predicting, and that is when and how teams will change in quality. Teams get hot, they get key players back in the mix, they get cold, the QB gets a little shaky and starts thinking that selling insurance wouldn't be so bad after all... I would be *highly* suspect of any system that claimed to be able to tell when these things would happen. The NFL is small enough, and the season short enough, that not everything regresses to the mean at the end.

DVOA, as I see it, doesn't exist to account for changes in performance, but rather to show that it happens and to what degree. We can see that the G-Men are hot right now precisely because the statistics changed in a measurable way, and it can't be adjusted away by tweaking the formula. I fail to see how this could be characterized as anything but a good thing.

64
by seth (not verified) :: Thu, 01/24/2008 - 1:48am

aaron-
i'm a very loyal fan of this site and all your work, and i'm basically on the side of those who say that DVOA is working well as is... but one thing concerns me, and i don't have an answer for it. a couple friends keep bugging me about this: all year it seemed like the vegas lines had the giants rated as an above-average team. by DVAO analysis, and my own analysis, which is primarily video review and information from FO, the giants appeared to be a .500 team, a team that was "meant" to go 8-8, maybe 9-7.... yet they kept getting bet like they were a 10-6 level team- a team that was really significantly above a .500 team. in addition, this was during the season, before anyone knew of the emergence of ahmad bradshaw or corey webster, before steve smith returned from injury (a bigger factor than realized, i believe), before (at times) the return of gibril wilson (not much discussed, but the defense really missed him- he went out week 12 at vikings and returned for week 16 at the bills)) and before the "new and improved" eli burst on the scene in week 17 (he did have 5 fumbles in week 16 at buffalo, remember). so i wonder- were we missing something? did the emergence of a few young players really make the difference, and thus was the "crowd" accidentally proved correct? or maybe the giants took adavantage of playing mistake-prone qbs in the last 2 rounds of the playoffs, and limited their own mistakes, which can always just happen?

65
by Carlos (not verified) :: Thu, 01/24/2008 - 2:19am

was the “crowd” accidentally proved correct?

Not necessarily. Vegas sets the line to try and offset winner and loser dollars. There are more NYers and hence more Giants fans than probably any but a few other franchises. Vegas undoubtedly gets slightly more pro-Giants action then team "merits" every year.

I say slightly, b/c if it was anything more than that, then the market would very quickly correct with all the smart money taking the NYG opponents every week.

66
by t.d. (not verified) :: Thu, 01/24/2008 - 3:14am

In each of the past two years, someone on this site has argued that what the eventual champion was doing was unsustainable. Obviously, the Pats have to be heavily favored, but this is a formidable Giants team that got this far for a reason

67
by Jon (not verified) :: Thu, 01/24/2008 - 6:52am

A lot of the comments here are making me cringe. FO isn't out to "get" the Giants. And I don't believe a majority of NYG fans who are familiar with this site think that. Fans of any New York team get a bad rap for being obnoxious, and I'd like to not perpetuate that stereotype if we can avoid that. Readers are a lot more likely to post comments if they're upset. Hey, it brings the page views.

That being said, I was a little dismayed reading FO's subjective commentary parrotting the conventional wisdom on the team this year. The Giants just had a ridiculous schedule in 2006, which combined with a lot of injuries and bad luck. I thought they'd end up at 10-6, with the only caveat being the strength of the NFC East being one potential obstacle.

FO's metrics say the Giants should lose the game. So do prognostications by almost any standard, so I don't see why FO should be jumped on about that point. The team is a statistical anomaly that has no meaningful analytic value besides being a reminder that anomalies do occur. Usually regular season performance is a good predictor of playoff success. The fact that the team has improved during the past few weeks in spite of its challenges should be reason enough to celebrate for NY fans.

57, are you serious about the Toomer comment? He's only been playing at a high level during the playoffs, besides his critical drop vs. the Packers. Simplifying the offense plays a part, but he simply has lost a step. Considering how bad Whitfield was last year (Petitgout didn't play all that much), I absolutely thought Diehl would upgrade the LT position this season.

68
by JasonK (not verified) :: Thu, 01/24/2008 - 9:25am

#62:

Most of the sports media's opinion of Diehl before the season came from the Giants' one nationally-televised preseason game, wherein Terrell Suggs made him look silly. I certainly give the Giants credit for having the foresight to make the Diehl move, but I'll stand by the assessment that the general public who don't see him in practice every week had no reason to expect him to be an improvement over Luke P. (Over Whitfield, sure, but not over Petitgout, who did start for over half of the season.) Yeah, he did OK in the 2 games he started in late '06, but they were giving him lots of help from the backs and TEs in those games.

(Personally, although I didn't think the sky would fall in like some analysts did, I did think that he would need more help from backs & TEs than he has.)

#67:

Amani Toomer in FO stats:
'04: 4.9 DPAR (59th), -7.1% DVOA (59th)
'05: 11.4 DPAR (34th), 0.8% DVOA (48th)
'06: 8.6 DPAR (50th), 9.8% DVOA (31st)
'07: 16.4 DPAR (29th), 8.9% DVOA (32nd)

Although there were some drops that stick out in fan memories (the home game v. Washington), he's pretty much the same guy he was before the injury in '06. That's a heck of a lot better than I expected. (The pre-'06 improvement was probably caused by QB improvement and the acquisition of Burress.)

69
by Independent George (not verified) :: Thu, 01/24/2008 - 10:28am

Oh crap. FOMBC strikes again. We're doomed.

70
by mmm... sacrilicious (not verified) :: Thu, 01/24/2008 - 10:42am

I'll bet a lot of people who don't publish websites 'picked' the Giants to go 10-6 before the season.

71
by Tom (not verified) :: Thu, 01/24/2008 - 10:47am

The only thing that I saw that struck me about this supposed "possible DVOA flaw" is that you noted that there was a team that was exactly the same as this Giants team, who also went to the superbowl completely unexpectedly after a late season surge. Of course in statistics it's really a matter of trying to predict what is most likely to happen. Even this year's 49ers COULD put above a 40% DVOA 4 games in a row, but it is extremely statistically improbable. It is completely reasonable to believe that the Giants as simply an outlier, and thats what I assumed it was, but I began to think that maybe something was mising when I realized this wasnt the first time.

The fact that this has happened twice shows that it is more possible that SOMETHING on teams such as these needs to be fixed. But of course it could just be two statistical outliers of similar types, and the formula is correct. I highly doubt those two teams were both actually as good as their last few weeks showed, so it isnt a question of "did they outperform their actual skill?" its a matter of how much. Maybe it is less, maybe it isnt. I only took intro to statistics in college, so I only know the basics, so dont ask me. Im sure you know more than I do about DVOA :-P.

72
by Paulo Sanchotene, Brazil (not verified) :: Thu, 01/24/2008 - 10:47am

# 18: Actually, Giants are 2-0 on neutral ground as they played a home game against the Jets.

73
by nat (not verified) :: Thu, 01/24/2008 - 11:02am

61 Jarrod Bunch:

The problem Aaron is dealing with is not the opponent adjustments, but the fact that he included a head-to-head game in his week 17-20 DVOA averages. That creates a logical loop, one that doesn't matter much for a 16 game season, but does matter in a 3-4 game stretch. Like so:

If you assume the Patriots are +52% DVOA over the Giants, then they are only +1% DVOA.

See the violence inherent in the system?

His solution (show VOA instead) wasn't ideal. Better would be to use the VOA from the Pats-Giants game averaged with the DVOA from the playoff games. In other words, assume the teams are equal, and see what the DVOA average shows.

Or, he could have just used the playoff games, and left week 17 out of the mix. Over the playoffs, the Giants have a 50% DVOA and the Patriots have a 57% DVOA. A statistically "significant" advantage, but not a certain blowout or even a large favorite.

Personally, I suspect Aaron knew what he was doing by including week 17, and was being intentionally obtuse to give us something to talk about for the next week.

74
by Dennis (not verified) :: Thu, 01/24/2008 - 11:17am

Re 72: How is that a neutral field? The Giants were in their home stadium with their fans. Yes, it's also the Jets' home stadium, but the tickets went to Giants fans. Of course Jets fans had a chance to buy them on the secondary market, but that's no different than when the Giants play any other team.

75
by Bill Moore (not verified) :: Thu, 01/24/2008 - 11:29am

As an aside, what I love about the comments that reference "heart" and "will to win" is the assumption that the opponents of the heart-led teams have no "heart." These are highly-skilled professional football players whose main priorities are money and championships (in one order or another - and often the two are interlocked). Can anyone really make a legitimate argument that Green Bay or Jacksonville or the 2001 Rams or anyone else didn't have the "will to win?"

"Heart" is one of those post-hoc factors that gets applied when trying to explain the margin of error in analysis of this sort.

Remember, History is written by the victors.

76
by Brooklyn Bengal (not verified) :: Thu, 01/24/2008 - 11:49am

What's strange about this post is that Aaron ends with stating that the Giants "were a mediocre, overrated team from Week 1 through Week 16." Um...really? Mediocre, sure. Overrated? How were they overrated? They were 10-6 beating up on lesser opponents. People called them an underdog against TB in the wild-card round. The Giants were not overrated in the least bit. All NY media expected another one and out year, and I can't imagine the sportswriters in other cities saying much different.

The 2007 Giants don't prove that DVOA is broken...they just highlight the importance of using DVOA's objective measurements to inform your own subjective opinion. Objectively, the Giants didn't beat one winning team during the regular season. Subjectively, seeing them pull out wins over TB and Dallas in the post-season wasn't very surprising. It's a little surprising, even among Giants fans, that the Giants have managed to play so well for four weeks in a row. Anybody who's watched this team knows consistency is a problem. Have the Jints turned a corner? Or is this a lucky streak? We'll have to wait and see...

77
by Brooklyn Bengal (not verified) :: Thu, 01/24/2008 - 11:57am

RE: 14

Sorry, but Washington was completely unimpressive to me all season. My gf is a HUUUGE Skins fan, so I keep up on them pretty well, but it was tough for me to believe they'd even make it past the Seahawks.

Still, this whole thread of the Giants being underrated/properly rated by DVOA seems to have brought out a lot of bitterness in the FO staff. No need to get into bickering matches with us posters, just stand by DVOA as having accurately gauged NYG's performance, not their potential.

78
by Rich Conley (not verified) :: Thu, 01/24/2008 - 11:59am

"Anybody who does not answer “Washington” is lying."

I don't know if I would have picked the giants in that 1 of 4, but I sure as hell wouldn't have picked washington.

79
by BDC (not verified) :: Thu, 01/24/2008 - 12:28pm

74: "How is that a neutral field? The Giants were in their home stadium with their fans. Yes, it’s also the Jets’ home stadium, but the tickets went to Giants fans. Of course Jets fans had a chance to buy them on the secondary market, but that’s no different than when the Giants play any other team."

Well, while I agree with you to a certain extent, surely you can see how it is at least a little different then when they play "any other team". If they are playing say, Dallas, while it is possible there will be some Dallas fans at the game, it is unlikely a large number of them will fly all the way to NJ to see the game. Whereas when playing the Jets, there may very well be a large number of Jets fans in the immediate area who can get to the game.

80
by GlennW (not verified) :: Thu, 01/24/2008 - 12:37pm

> If #17 thinks the Patriots are “staggering” to the finish line, then his expectations seem a tad unreasonable.

Sure, that's why I qualified the statement by saying "relatively speaking", as the Patriots have still probably performed as the best team in the league over the past two months. I was just saying that they're not the same team as they were earlier, especially defensively, and I don't know if the difference has even fully showed up in Weighted DVOA (as I don't know exactly where the cutoff point is).

I'm consistently seeing a lot of open space over the middle in the Patriots' pass defense. In the Super Bowl maybe the Pats will continue with their improvement in redzone defense to offset this, and maybe they won't. Their redzone performance has been excellent in the playoffs, but was near the bottom of the league in the regular season, and was poor in the Giants game specifically. It's the one area where I see the Patriots as especially vulnerable, so we'll soon see...

81
by Rich Conley (not verified) :: Thu, 01/24/2008 - 12:42pm

"Their redzone performance has been excellent in the playoffs, but was near the bottom of the league in the regular season"

Except when they played the Colts...and the Steelers. You know, the teams they played in the regular season who affected playoff seeding.

Its funny how that works out, in the games that matter (good AFC opponents and playoff games), their redzone D has been great. I have a hard time believing that its a coincidence.

82
by Kurt (not verified) :: Thu, 01/24/2008 - 1:00pm

Objectively, the Giants didn’t beat one winning team during the regular season.

Now we're down to *zero* such wins? Jeez. Objectively, the Redskins were 9-7, however much they impressed you sitting on your girlfriend's couch. The Eagle wins obviously should count too, but if you insist on excluding 8-8 teams. Having one win against a winning team is completely unexceptional.

Here are all the 9-11 win teams, and the number of wins they had against winning teams:

Pitt: 3
Cleveland: 1
Jax: 4
Tenn: 2*
San Diego: 2
NYG: 1
Wash: 2*
Tampa: 2
Seattle: 1

* - week 17 games involving Brad Johnson or Jim Sorgi

Yes, the Giants are a surprise Super Bowl team, but this year 9 of the 12 playoff teams would have been surprises if they had made it. I don't see any reason to believe the Giants are any bigger a surprise than Tampa or Seattle or Washington would have been.

83
by Kurt (not verified) :: Thu, 01/24/2008 - 1:03pm

BB, I do agree with you about the "overrated" bit. I don't see how they could be considered overrated, unless anyone who liked the Giants more than Aaron did was by definition overrating them.

84
by Gerry (not verified) :: Thu, 01/24/2008 - 1:13pm

Aaron,

If you see this question and have the time, how about sharing the following information:

How many 40+ DVOA games each team in the NFL has had over the past two seasons, playoffs included. Obviously NE is going to top that list, and I would imagine that Indy will be second. I am also guessing the Giants are not in the top 3.

But I would be surprised if they aren't in the top 10-- which would in my opinion validate what I have been saying, that the Giants are capable of playing at a level that not many teams are capable of playing at. What has made the Giants the Giants is that they play at a mediocre level (or below) significantly more frequently than other teams that share the ability to play at that kind of a high level.

85
by Gerry (not verified) :: Thu, 01/24/2008 - 1:20pm

"-Dave Diehl being a better OLT than Luke Petitgout was.

Pre-season, there was no rational reason for a computer or a person to expect that any of the above would be true. "

Had you left off that last one, I wouldn't have been able to argue with you, JasonK. Preseason, I expected Diehl to be better than Petitgout. I did not and do not consider Petitgout to be that fantastic when penalties are not included, and I knew that there would be a marked decrease in penalties without him.

The other points you raised, I agree were not things that were likely at all.

86
by Gerry (not verified) :: Thu, 01/24/2008 - 1:24pm

"

Not necessarily. Vegas sets the line to try and offset winner and loser dollars. There are more NYers and hence more Giants fans than probably any but a few other franchises. Vegas undoubtedly gets slightly more pro-Giants action then team “merits” every year.

I say slightly, b/c if it was anything more than that, then the market would very quickly correct with all the smart money taking the NYG opponents every week."

If this was true, we would see at least one of two things:

1) The Giants would, in most year, have a below 500 record against the spread, and/or
2) The final spread in Giants games would have moved from the opening spread significantly more often than it does for other teams.

I am pretty sure that neither is borne out by the facts.

87
by Gerry (not verified) :: Thu, 01/24/2008 - 1:38pm

"Objectively, the Giants didn’t beat one winning team during the regular season."

As #82 said, they beat one.

And now, playoffs included, they have beaten 4.

Let's not ignore that latter fact. While it may be interesting to say what we could have or should have expected at the start of the playoffs, the more important question now is what can we and should we expect in the Super Bowl.

I see a good team that is on a hot streak going against a great team that is on a bit of a downstreak (for them). And the hot streak and cold streak are relative-- the Pats cold streak has still had them playing just as impressively as the Giants have been playing during their hot streak.

I expect the Pats to win. However, the only result which would absolutely shock me would be the Giants blowing out the Pats. A Pats blowout win, a Pats close win, and a Giants close win are all things that would fall into the range of things that would not shock me.

88
by GlennW (not verified) :: Thu, 01/24/2008 - 1:53pm

> Its funny how that works out, in the games that matter (good AFC opponents and playoff games), their redzone D has been great.

I was under the impression that in pursuit of the perfect season, all the games mattered. Including the Week 17 game against the Giants, in which the Patriots played all of their healthy starters, and played hard till the end.

I think the precise redzone breakdown is a coincidence, but the larger point is that there have been a lot more such redzone situations to defend later in the season. That's what somewhat concerns me, how relatively easy it has been to march down the field on the Patriots. After that, the defense has stiffened in the playoffs, but I think there's a good amount of luck and other factors involved there too (like Dennis Northcutt not being able to catch, Tomlinson being unavailable etc.). The defense is of at least a minor concern imo.

89
by MJK (not verified) :: Thu, 01/24/2008 - 2:05pm

Two thoughts, one on each team:

On the Giants: I know DVOA wasn't crazy about their regular season, and I know conventional wisdom saw very real things like all of Eli's mistakes, but it's possible we were all overlooking something in the schedule.

Just for fun, I track team performance using a Maximum Likliehood Estimation system (linked in my name--sorry it's only updated through 1/12--computer crashes have kept me from getting it more current). It only looks at W/L, not play-by-play performance, injuries, home team, etc., and hence is a lot less sophisticated than DVOA (plus it tends to be inaccurate for teams with very few losses or wins--don't trust New England's high rating), but one thing it does do that DVOA doesn't do (at least, I don't think DVOA does) is that it iterates until convergence. I.e. it keeps iteratively adjusting opponent strengths until ratings across the board don't change anymore. I believe DVOA just performs one or two iterations (it didn't do any until the Colts easy schedule broke it a year or two ago). Aaron, please correct me if I'm wrong...

Anyway, MLE thinks that the Giants faced a significantly harder schedule than many other teams this year, and hence were significantly better than a lot of teams that DVOA had pegged higher. In general, this is because it thinks both the NFCE, and the NFCN (which the Giants played) were two very good divisions, with four above-average teams in each. Maybe these two divisions really were significantly better, across the board, than some of the other divisions out there, but the limited connectivity of the schedule means that you have to really iterate the opponent adjustments to see that. And so maybe the Giants really weren't that bad in the regular season...MLE seemed to think so.

Don't know how much I trust this, but it does make me wonder if maybe DVOA would be improved by iterating the opponent adjustments until convergence, instead of just once or twice, if Aaron isn't already doing this. I wonder if this would have pegged the Giants higher and avoided some of this controversy...

On the Patriots:
I am amused by people talking about them "panting down the stretch". This is a team, that even in their post bye "slump", had only two close games out of nine. The other seven consisted of two blowouts and five games that weren't really that close--the game was largely put away, barring collosal bad luck, by the 4th quarter (i.e. a 2-score lead by sometime midway in the 4th).

I think the absolute dominance of the Patriots in the first half of the season artifically raised people's expectations. If the Patriots had had the exact same nine game stretch since their bye, but had, say, gone 6-3 before it, with say only one blowout and a few close games (i.e. a normal team's performance), no one would be talking about the team slumping now. In fact, the five convincing wins since their Baltimore squeaker and one closer game in Week 17 to close the season out would have people talking about how they were "peaking at the right time".

It's all relative.

90
by MJK (not verified) :: Thu, 01/24/2008 - 2:06pm

Sorry, typo in the link I posted above. Click my name here for the working one.

91
by NOLA_SaintsFan (not verified) :: Thu, 01/24/2008 - 2:11pm

Has anyone gone back and tried to use DVOA as a predictor of the outcomes of past games over the life of the DVOA statistics and compared the accuracy of DVOA to the other statistical systems that are available out there, as well as the predictions of the respected analysts in the sports world? I would expect that DVOA would do quite well overall and prove to be one of the best tools for predicting the outcomes of games given the above alternatives. If that does prove to be the case, can we stop saying that it needs to be "fixed" to account for this year's giants team as it is already one of the most accurate systems out there.

92
by David Mazzotta (not verified) :: Thu, 01/24/2008 - 2:27pm

#84 - all of that is probably true, but Aaron's question still remains unanswered: How do you see that coming?

How do you measure the difference between a good team playing badly, and a plain old bad team?

93
by Gerry (not verified) :: Thu, 01/24/2008 - 2:34pm

David,

I threw around some ideas on the Audibles thread, but probably the answer is "no clue."

However, to me at this point it is less interesting to ask "could we have seen this coming" and more interesting to ask "is this a good team that was playing poorly or a poor team that is playing well", since that seems more applicable to the Super Bowl which is coming.

After the Super Bowl, there is the whole offseason to brainstorm about ways that we might have been able to see this coming-- if this is something other than a fluke.

94
by MJK (not verified) :: Thu, 01/24/2008 - 2:36pm

To clarify:

I am in no way suggesting DVOA needs to be "fixed". I think it's the most advanced metric out there for evaluating football performance.

I'm just curious as to whether it does iterate to convergence when taking opponent adjustments into accoutn, and, if not, if modifying it to do so (although I confess ignorance as to how difficult that might be) might increase its predictive power overall. Not just for the Giants.

I was just citing the Giants and MLE as an example of how a system that is inferior in most respects to DVOA, but possibly superior in one small respect, came to a different conclusion than DVOA, and was wondering why that might be.

95
by coldbikemessenger (not verified) :: Thu, 01/24/2008 - 2:37pm

GlennW
I have a qiestion.
One thing the chargers did that drove me up the wall was that almost constant first and 10 run in the red zone.
Do a lot of teams do this with the Pats?
Constantly facing 2nd and third down seems to work in the Pats favor on defense.
Conversely the Pats often seem like a pass first team near the goal line.
They showed some graphic for brady's td to int ratio.
It was insane!
Anyway,
Have you noticed this all year, pats very aggressive opponents not so much?

96
by GlennW (not verified) :: Thu, 01/24/2008 - 2:47pm

> It’s all relative.

Absolutely. My comments were relative to the standard of The Greatest Team of All Time, and a team a couple posters in this thread claimed can't be beaten in the Super Bowl. Obviously if the Patriots had started the season 6-3, we wouldn't be talking in these absolute terms.

> Have you noticed this all year, pats very aggressive opponents not so much?

I think absolutely you have to pass on the Patriots on longer 1st-and-goal, 2nd-and-goal situations. And that's essentially what the Giants did the first time around-- four redzone TD passes.

97
by MJK (not verified) :: Thu, 01/24/2008 - 3:06pm

Have you noticed this all year, pats very aggressive opponents not so much?

To a degree.

On long 1st-and-goals, the Pats USUALLY (say about 65% of the time) will run some kind of pass play where someone is in the endzone, so if the defense screws up Brady can capitalize, but usually the intent of these plays is to ge the ball down inside the 3--the first read will often be the quick hitch to a WR where the CB is backed up, or a throw in the flat to the RB, or a quick slant to a WR at around the 3. Sometimes they'll be more aggressive. But a fair percentage of the time, say 35%, especially very early in the season and very recently (not so much in between), they'll hand off. These runs often get stuffed, but sometimes (maybe 30% of the times they are called) they pick up a few yards and give 2nd and short.

In contrast to other teams, the Pats NEVER seem to throw the fade. (Which is good, because I hate the fade, on artistic grounds. Absolutely no style or finnesse--it's essentially a coin flip). I don't know why not--with Moss's leaping ability, you'd think it would be a natural play, and with their short CB's, you'd be right in thinking other teams would try it against them, so you'd think they'd run it in practice to give the D practice against it, but...

As far as other teams go, I haven't noticed an especial propensity for opponents to run on long 1st-and-goals. One thing I have noticed, though, that may contribute to the Pats stiffening red zone defense of late, is that the weakness of the Pats defense is the ILB's covering passes about 10-15 yards deep in the middle. Once you get to 1st-and-goal, this hole is gone, so the LB's are (1) better in coverage, and (2) can crowd the line to stop the run more safely.

98
by sippican (not verified) :: Thu, 01/24/2008 - 3:15pm

Glenn- To be extra clear, I did not say the Patriots were unbeatable. I said I cannot see any way the Giants beat them in two weeks. Not the same thing.

The "blueprints" offered every week consist of vague to-do lists of ways to lose to the Pats by fewer points. They've all been a waste of time.

The Colts/Pats would have been a show, but it was not to be. Norv is worse than Marty. The Jags played terrific and lost. I was rooting for the Packers to win because I thought it would make for a more lively Super Bowl. This one looks like a drubbing.

99
by Rich Conley (not verified) :: Thu, 01/24/2008 - 3:15pm

"I was under the impression that in pursuit of the perfect season, all the games mattered. Including the Week 17 game against the Giants, in which the Patriots played all of their healthy starters, and played hard till the end."

If that game had been really important, Troy Brown wouldn't have been returning kicks. Chad Jackson wouldn't have been active. Kaczur (who had practiced that week) would have played. Merriweather wouldn't have played most of the game. The game was important, but the Patriots didn't exactly bring out the big guns.

100
by Rick A. (not verified) :: Thu, 01/24/2008 - 3:17pm

Let's face the facts. A great percentage of you following DVOA are looking for an "edge" and your criticisms of the best system of evaluation I've ever seen in a lifetime of playing and studying football are emotionally based--after the fact of you being unable to come up with that edge. In sports today no one will ever come up with an edge because, in all pro sports--not just football, the players do not necessarily give 100%, no matter what they say, until the last stages of the season. Haven't you noticed that "Wild Card" teams keep winning in all the sports ? You just do not see what a team is capable of until they collectively decide to fully put out and, hence, you can't possibly know in advance what is truly likely to happen. Much is being said about the "improvement" of Eli as the big factor here. I can clearly see that's not the biggest deal here. He's been very good all along--how many NFL QB's can consistently win on the road ? Throughout the season his receivers were selfish, they dropped balls to the tune of #2 in drops, they consistenly ran the wrong patterns and were lazy about it too--DVOA doesn't measure that one--and the defense, in spite of all it's talent, visibly paced itself more than other teams do-often creating late deficits too hard to overcome. DVOA is conceptual, and it's accurate--but human beings are on the field, my friends. Don't be surprised if New York's best game is still yet to come.

101
by GlennW (not verified) :: Thu, 01/24/2008 - 3:23pm

> To be extra clear, I did not say the Patriots were unbeatable. I said I cannot see any way the Giants beat them in two weeks. Not the same thing.

That sure sounds like a statement that the Patriots are unbeatable *in this Super Bowl*, which is what I said and specifically what I was attempting to analyze: this game, against the Giants, not a hypothetical rematch with the Colts or anything.

102
by Joseph Adler (not verified) :: Thu, 01/24/2008 - 3:40pm

A couple thoughts: first, I wonder if there is a way to build a "IDVOA"; an "Injury and Defense adjusted DVOA" rating. One of the reasons why each of the remaining teams won so many games late in the season was due to opponent injuries. (One example is the Chargers without a healthy Tomlinson, Gates, and Rivers, though I'm not sure how big a difference that would have made. A second example would be the Sanders-less Colts.)

Secondly, I wonder if there is enough data to adjust specific player additions and subtractions. The best example I can think of are the changes in TE on the Giants; I suspect that Jeremy Shockey is at best a "replacement level" player (to borrow the Baseball Prospectus term), and probably a lot worse.

103
by Dan (not verified) :: Thu, 01/24/2008 - 3:43pm

One important thing to look into is injuries. Injuries and coaching. Last year, the giants had Tim Lewis as their defensive coordinator. His predecessor was Johnny Lynn. Although Mike Tanier wrote that we lost our best coach when management (thanksfully!) forced coughlin to get rid of lewis---but in reality, coming from someone who watches more plays over and over and more games---we got rid of our WORST coach (besides hufnagel). Lewis installed a read and react defense which was just pathetically bad with our personnel. In general, i'm a firm believer that the read and react defense cannot work with almost any personel actually. Yet even so, we went through our DVOA burst weeks 5-8...WHY? Because the defense was COMPLETELY jelling. Then what happened???--INJURIES! injuries mean everything. For some reason I cant remember if Osi was hurt too, though i think so, but besides him STrahan (our most important defensive player) and Lavar got injured, factor that with a declining pierce and a PATHETICALLY bad carlos emmons and you will run into problems with your run defense. Anyway, this giants team is completely different. We attack on defense, and ball control on offense. This leads to LESS time for the defense on the field and MORE time for the offense...which leads to a fresher defense and a high time of possession in the games we win.

104
by Scott de B. (not verified) :: Thu, 01/24/2008 - 3:51pm

The fact that this has happened twice shows that it is more possible that SOMETHING on teams such as these needs to be fixed. But of course it could just be two statistical outliers of similar types, and the formula is correct.

The 2004 Panthers went 7-9, which supports the 'fluky' characterization. We'll have to see how the Giants do in 2008.

105
by Gerry (not verified) :: Thu, 01/24/2008 - 4:12pm

Scott de B,

But the 2004 Panthers had not made the playoffs the prior two years, and had not had a run of 4-straight 40+% DVOA games in the 2003 season.

The Giants have made the playoffs three years running now, and last year they had a 4 game stretch where they rattled off DVOA of 40+%. So where the Panthers came out of nowhere, the Giants at least had been hovering around and shown stretches of brilliance (admittedly, surrounded by periods of mediocrity or worse).

106
by GlennW (not verified) :: Thu, 01/24/2008 - 4:24pm

Season-to-season correlation isn't very good in the NFL (hence FO's system spits out these crazy yet not-so-crazy-after-all preseason predictions), nor should that be the objective here anyway. Next season is next season; the question at hand is/was, what will the Giants (or any other team) do in the playoffs this season? And I still say that anyone who had $10k at his disposal (and is honest) wasn't betting on the Giants to make the Super Bowl at anything less than 10-1 odds-- Giants fans, supporters, and eternal believers included.

107
by Pat (not verified) :: Thu, 01/24/2008 - 4:50pm

I am amused by people talking about them “panting down the stretch”. This is a team, that even in their post bye “slump”, had only two close games out of nine.

This must be some strange way of counting where "two" is equal to "three." The Eagles, Ravens, and Giants games were all close by any definition of close: the Patriots were either tied or behind in the fourth quarter in all three games. That's as close as you can get.

I don't think it's an exaggeration to say that the Chargers game was "close," either. Yes, it got to a two-score game in the fourth quarter, but the fourth quarter of that game was so short that really, the portion of time where the Patriots were "extremely likely to win" was basically just one drive.

And I don't really see your point - a team can slump from a team that puts the game away from most opponents in the first half to a team that puts the game away from most opponents midway through the fourth quarter.

108
by Pat (not verified) :: Thu, 01/24/2008 - 4:54pm

#103: Not exactly: it's important to realize that the playoffs and the regular season are two totally different things. The game doesn't change, but the opponents do. It's entirely possible for a team to be well-built to match up in the playoffs, but not to match up well in the regular season.

That being said, the simplest way of understanding that ('Secret Sauce') had the Giants ranked near-last, but it's entirely possible that what the Giants and Panthers had in spades isn't easily measurable.

109
by Julio (not verified) :: Thu, 01/24/2008 - 6:18pm

Do you think it's possible that there is a Ewing effect going on with Shockey gone? The Giants scored 38 points against Buffalo the week after Shockey got injured, then 35 the next week against NE. Maybe DVOA can't measure the psychological effect of losing a prima donna?

110
by MJK (not verified) :: Thu, 01/24/2008 - 6:47pm

Re 107:

I count the Ravens and the Eagles game as close. I didn't consider any game where the losing team has to recover an onside kick in the closing minutes in order to have a shot at winning as "close". You could make an argument that it was, depending on how you define close, but we can disagree on that.

111
by not perplexed (not verified) :: Thu, 01/24/2008 - 7:10pm

I think you’re pushing it a little with that definition of “close,” MJK. New England only had two (Steelers and Dolphins) of their last eight games basically locked up going to the 4th quarter and they were still probably playing the best football of any team in the league over that period. You could argue their edge over the next best teams was smaller than it had been through their first ten games but the edge was still there. The two biggest factors in closer games were weather and other teams raising the level of their play to compete with the Patriots, which led to some great games down the stretch.

Using this expanded definition for these six close games, two were played in bad weather (Ravens and Jets, although that’s somewhat unfair to how well the Ravens played), two featured the Pats playing at a high level with the opposition raising its own level of play (Eagles and Giants, which they’ve subsequently managed to maintain) and two were playoff wins over a couple of the strongest teams in the league (Jags and Chargers).

Yes, the defense isn’t great. Yes, they’re beatable. But the only time their overall performance was less than outstanding was when they played in bad weather.

112
by GlennW (not verified) :: Thu, 01/24/2008 - 7:22pm

> Yes, the defense isn’t great.

Maybe to partially quantify my earlier comments about the Patriots defense, FO had New England's Weighted Defensive DVOA at #18 in the league going into their first playoff game, and at a marginally below-average +1.0% for Weeks 17-20 in this latest posting (I'd still be interested in the full-season and Weighted DVOA numbers for the four teams still playing last weekend, but that format has been dropped). I assume the latter number includes credit for whatever success the Patriots have been having in the redzone in the playoffs.

113
by PatsFan (not verified) :: Thu, 01/24/2008 - 7:56pm

Re: #100

Swagger. You forgot to mention swagger.

114
by Pat (not verified) :: Thu, 01/24/2008 - 8:10pm

#110: How in the hell would the Giants game not count as "close"? The Patriots didn't get a two-score lead until two-thirds of the way through the fourth quarter.

For 92% of the game, the margin was within one score.
For 78% of the game, the Patriots were losing.

The fact that the Giants needed an onside kick at the end to win is immaterial. Almost every team who loses needs a miracle at the end of the game to win (or tie). The Giants lost that game by not holding the Patriots to a field goal on Maroney's drive.

The Giants didn't actually "need" an onside kick, either, until after the 2-minute warning - they still had two timeouts left (plus the two-minute warning beforehand). A field goal would've brought them within one score, and the Patriots would not have been able to kneel out the clock if they recovered the kickoff.

I don't see how needing to recover an onside kick is less likely than, for instance, if a team gains the lead with 10 seconds left, kicks off to the other team, and the other team is forced to go 80+ yards in 1 play. I've seen more onside kicks recovered than I have miracle plays like that work.

115
by vis (not verified) :: Thu, 01/24/2008 - 11:08pm

104- Though Gerry's answer is an important part, I think there's more. To replay my comment that got lost in the Audibles-cum-Greatest-of-all-time-thread:

On the CAR/NYG comparison…

It should be pointed out that the regression in the following year by Delhomme (and thus, CAR) was largely attributed to a change in success rate on close and ill-advised passes. Where in 03, Delhomme was throwing into double-coverage and making completions, in 04 and since, Delhomme’s same types of passes were getting defensed (or picked).

In other words, Delhomme’s (and ultimately, the Panthers’) success then failure had much to do with a better-than-expected rate of return on the same types of attempts, followed by a regression to the mean.

Eli’s increased success seems primarily from an improvement in decision-making and timing (or, a “maturing” if you prefer), a change that is presumably more resistant to regression.

Or so Giants fan’s would hope.

116
by vis (not verified) :: Thu, 01/24/2008 - 11:23pm

If we look at a 3- or 4- year spread, surely there's a better comparison for the Giants than the Panthers. Carolina came from nowhere and returned to nowhere, with one good season in between. NY has been a "wildcard-caliber" team for several years, and finally pushed through.

The missing element in evaluation of current-year DVOA is prior-year DVOA. While there is certainly turnover in personnel, year to year, teams tend to follow some overall arc through down and up generations. I think considering prior year performances (with decreasing weight) would at least partially combat the "sample size" problem continually complained about.

A composite arc-DVOA score would be fairly simple algorithm to develop and calculate.

117
by FoggyProfessor (not verified) :: Fri, 01/25/2008 - 12:48am

Aaron, two theories

1) You belong to the Jets cabal.

2) You're to stubborn to let go. If you keep "tweeking" until it fits, you will surely find something less useful than what you have. Create a hypothesis, test it, and move on. You've decided that performance against an average is a valuable yardstick. Good hypothesis. Now determine its predictive power. Good. Now move on. Perfomance relative to the top quartile? Bottom quartile? Test it. Good. Move on. Average playoff team? test it. Keep going. It's what an "outsider" should be doing.

118
by terry (not verified) :: Fri, 01/25/2008 - 3:09am

dvoa only shows past results the giants
are getting offence from differnt players bradshaw smith boss and corey
webster and the offence(field position and points) has helped the defence

119
by Rick (not verified) :: Fri, 01/25/2008 - 4:06am

Prior to the NE/Phi Super Bowl, Aaron, you stated that you thought it had all the hallmarks of a NE blowout. It turned out to be a very close game.
This Giants team is nowhere near the level of that Philly team, while this NE team is much better than that NE team. So do you see this as having all the hallmarks of a NE blowout?

Frankly, I do. I'm not saying the Giants don't have a shot, they do. A very legitimate shot.
But they are playing WAY over their norm, so the question becomes when do they regress to the mean? Assuming it's not in the Super Bowl, then they could pull it off (especially if they continue their 0 turnover streak AND Eli racks up another 100+ QB Rating).

But this is the most mediocre team to ever make the Super Bowl, I'm guessing. Perhaps you can sort the stats and see what DVOA says, but I'm thinking this is a worse blowout than the Giants/Ravens was.

Then again, NE tends to win Super Bowls by 3 points. So who knows.

120
by BDC (not verified) :: Fri, 01/25/2008 - 11:05am

91:

"Has anyone gone back and tried to use DVOA as a predictor of the outcomes of past games over the life of the DVOA statistics and compared the accuracy of DVOA to the other statistical systems that are available out there, as well as the predictions of the respected analysts in the sports world? I would expect that DVOA would do quite well overall and prove to be one of the best tools for predicting the outcomes of games given the above alternatives. If that does prove to be the case, can we stop saying that it needs to be “fixed” to account for this year’s giants team as it is already one of the most accurate systems out there."

Actually, yes, I did this earlier this year. Sometime in December, and posted the results on here. It turned out that using straight DVOA (or Dave, earlier in the year) resulted in a "straight up" win rate of around 68%. This was virtually identical to the "straight up" win rate Vegas had, which was also around 68%. Keep in mind, I only looked at this season, and I only looked at games up until that point (week 14 I think it was, but I might be mistaken).

What that means exactly, I leave to you.

121
by FredFarmer (not verified) :: Fri, 01/25/2008 - 1:36pm

No 91.

DVOA can be used as a very good predictor of future games.

There's a thread on a UK site that has gone 107-76-9 for 58% against the Vegas line in real time this season and is 65% since week nine.This implies that DVOA improves with age and more plays.

http://www.punters-paradise.com/forums/showthread.php?t=8278&page=6

Re the whole debate.

Not really sure what all the fuss is about re the Giants getting to the Superbowl.DVOA had them at around 7 point dogs in all three games which gave them about a 30% chance of winning in each game.Three straight wins would have been the outcome about once every 38 years.

Every possible outcome has a probability of occurring even if it's vanishingly small....and the Giants winning three straight was hardly that.

The problem you have in trying to tweak DVOA to account for a set of outcomes at the extreme of the predictive curve is that in doing so you you may make the predictive content work less well for the majority of other teams.

Exponential smoothing's probably the way to go.

FF

122
by im_no_playa (not verified) :: Fri, 01/25/2008 - 3:16pm

continuing 102...

I think he's on to something about Shockey going out. ASSUMING a simple replacement of Shockey by Boss, look at the individual TE numbers on both and you see how much better Boss has played. I think it is correct that it is possible to break down DPAR & PAR to per pass (as I believe the *PAR values are cumulative), which gives Boss a 0.414 to Shockey 0.091 per pass DPAR, and Boss 0.400 to Shockey 0.095 per pass PAR.

So the system would have had to predict that Boss would be much better than Shockey (or at least change the offensive system to have a net gain), and obv that is impossible

So what this may come down to is Coughlin having to play Shockey because he is high $$, but the NYG are better w/o him. Oh, and I should mention that I don't like this Shockey fellow.

123
by Johnson Autobaton (not verified) :: Fri, 01/25/2008 - 4:17pm

Bottom line is DVOA can't predict new players to step in and play better. The Giants have undergone one of the biggest in-season changes of any team out there. Ahmad Bradshaw has given them a big lift at RB. Gerris Wilkinson played lights out at LB (one of their biggest weaknesses) against the Patriots. Hixon has turned around an anemic kick return game. Webster has gone from buried on the depth chart to playing very well at CB. Ross has emerged at CB. Steve Smith has started to shape up as a slot receiver. Plax has seemed to get healthier. Shockey went down, Boss steps up. New D-coordinator comes in with a new system the defensive players have to adjust to. The list goes on and on...bottom line is this Giants team isn't the team DVOA or the FO staff was projecting before the season..and you can't expect either to have the foresight to do so...it's not even the same team from the first half of the year. Doesn't mean there's anything wrong with DVOA...it's just the type of thing that makes sports fun to watch..changes in personnel, strategy, confidence, injuries.

124
by Rick (not verified) :: Fri, 01/25/2008 - 7:22pm

#123:
Nothing can predict new players stepping in and doing well. The Eagles' first loss to the Giants is more indicative of what happens with new players (WINSTON JUSTICE vs. Osi Umenyiora???) than what the Giants are experiencing with someone like Boss.

The Giants are much better with Boss than Shockey. But Boss isn't the only one who is stepping it up, he certainly is NOT the singular difference maker of these games.

What that says, though, is that the possibility of letdown and reversion to mean is far more probable than a continuation of the high level of play they are experiencing.

I'm not saying the Giants don't have a shot. Certainly they do, and as someone pointed out, they've had a 30% chance of winning each of 3 games. While they've had a 30% chance of winning each, the odds of them winning all 3 in succession is even smaller...making their odds of winning 4 very small.

Again, not impossible, just very small. Which is why a Giants win would be so very improbable.

How many upsets have occurred in Super Bowls over the years?
Jets
1st Pats
Buccaneers
Denver over Green Bay
Giants over Buffalo
Chiefs over Minnesota

In 41 Super Bowls, the underdog has won 13 times. That's less than a 30% chance for the Giants in this game.

Overall, the odds are just very low.

I'm JUST talking odds. Reality has proven the odds wrong about 28% of the time.

But this Giants team? Its inconsistency leads me to believe it has the most remote chance in history.

125
by Pat (not verified) :: Fri, 01/25/2008 - 7:43pm

The Eagles’ first loss to the Giants is more indicative of what happens with new players (WINSTON JUSTICE vs. Osi Umenyiora???)

Being fair, that game was more about "what happens when facing a strong defensive pass rush and a complete cluster*!(! happens with pass protection." I hate Justice getting blamed entirely when the problem was really Justice overmatched, Runyan struggling, Westbrook being out (he's a fantastic blocker), and the tight ends struggling as well (LJ Smith was still out at that point).

New players stepping in don't usually cause that big a problem because teams can usually adapt. The Eagles would've still had problems if William Thomas had been in. No, not "6 sacks" bad, but it still would've been bad.

That's why they didn't pull Justice to shuffle the line up : because it really wouldn't've solved the problem entirely. Justice later played in the Buffalo game which ended the season, and for the most part did fine.

126
by vis (not verified) :: Fri, 01/25/2008 - 8:46pm

124 Rick, if I remember my statistics courses correctly, you're making the old coin-flip series fallacy.

i.e., if I flip a coin 3 times and it lands heads 3 times, what is the chance it lands heads again on the 4th flip? 10/90? 25/75? nope. still 50/50. the overall likelihood of flipping a coin heads four times in a row only applies before the string of flips has begun. each individual flip's odds remain the same.

so to say that the Giants have won three-straight, that makes the odds of a fourth straight win smaller is absolutely incorrect. before the playoffs started, the odds of the Giants winning the superbowl (four flips) was very small. but we only have one game ahead of us (one flip), the odds of success are independent to this game.

127
by Johnson Autobaton (not verified) :: Fri, 01/25/2008 - 10:23pm

#124

I understand and agree...thing is, new players aren't always a bad thing...sometimes they are a good thing. In the Giants case I believe so...now I'm not saying the Giants are going to beat the Pats or that anyone should've seen this coming. All I'm saying is that looking at numbers from earlier in the season, while a good measure and probably the best we have, is never going to be a perfect indicator of what the team has now. To say that having new players is reason to believe the Giants will eventually regress to the mean is wrong I;d say..maybe they will, but to assume that is the case is certainly not correct.

128
by LamKram (not verified) :: Fri, 01/25/2008 - 10:27pm

#121: Excellent points. DVOA and other models only predict the most likely outcomes, but there is always an element of chance. 30% x 30% x 30% = 2.7% sounds about right for the odds the Giants would make it to the Superbowl. But of course, there were 4 wildcard teams. The odds that ANY wildcard team will make it to the Superbowl are probably pretty good. It has to be somewhat better than 2.7% x 4 because the wildcard team wont always face the top 2 seeds going down the stretch. Without going thru the math, I'm guessing there's about a 20% - 25% chance that at least one wildcard team will make it to the SB each year. This was one of those years, and that team happened to be the Giants.

That said, I find it amazing that people are talking about the Giants having momentum, while the Patriots are limping toward the finish line. The last time I checked, the Giants won their 3 playoff games by 10 pts, 4 pts, and 3 pts (in OT). The last two games came down to the last play. The Patriots, meanwhile, won their two playoff games by 11 pts and 9 pts, and both games were essentially over halfway thru the 4th quarter. I guess relative to expectations, the Patriots underperformed and the Giants overperformed. But the Pats are still performing at a higher level than the Giants.

129
by B (not verified) :: Sat, 01/26/2008 - 2:02am

If the players for the Giants are better now than they were at the beginning of the year, what does that say about the Giants coaching staff? One one hand, they did a very good job building a team with quality backups, but on the other hand, their backups in many cases were better than the starters, so why weren't the better players starting from the beginning?

130
by BDC (not verified) :: Sat, 01/26/2008 - 5:38am

129:

Probably for the same reason Brady wasn't starting ahead of Bledsoe back in 2001. It's all well and good to see a guy perform well in practice, but it is not the same as how they might perform in a game. And often times, a team doesn't want to risk losing a game or games just to find out how a guy might perform "in a real game".

131
by JoshuaPerry (not verified) :: Sat, 01/26/2008 - 6:22am

I can't resist, Manning is better than Brady in playoff games! Ha! It is pertinent!

132
by BDC (not verified) :: Sat, 01/26/2008 - 7:38am

God, please don't tell me I started *that* argument again! I didn't mean to, I swear.

133
by Rick (not verified) :: Sat, 01/26/2008 - 4:14pm

#126 vis - I'm aware of the fallacy, but that's not the odds I'm referring to.
The odds aren't what is the likelyhood of the next flip being heads, it's what are the odds of getting 4 heads in a row? THAT is NOT 50/50, it's the odds over the course of iteration.
The odds of the Giants winning each individual game WAS 30% (I think it's closer to 20% with the SB). So there are 2 calculations - one for each individual game in its own right, then one for the sum total of ALL games.

Think of it this way: What were the odds of the Patriots going 18-0 to this point? For each game, there were probably 60%+ odds they would win. But the odds for 18-0 were minuscule, at best, with those odds diminishing as they won each game.

Comparatively, from that standpoint, the odds of the Pats going 19-0 are much lower than the Giants going 4-0, simply on a standard statistical model. But when you factor in level of play, consistency, and power ratings, the odds shift in the Pats favor dramatically.

Again, there is a chance for the Giants, and I don't write these things to demean their outstanding accomplishment. But I am less impressed by a team that has beaten 3 teams, 1 of which was questionable at best, and 2 which had stumbled in the final 4 weeks. The Pats, on the other hand, beat 2 teams that were actually improving with time. Of all the teams I expected to have a shot at beating them, the Jags and the Chargers were the 2 based on their improvement.

One can say the Giants have shown similar improvement over the last 4 games, which is why they have a shot. I would argue that there is a spectre of inconsistency looming in the background. December was still a weak month for the Giants, and they won to stay in the playoff hunt...but only just.

The Green Bay game, from my standpoint, was beginning to show cracks in their improvement. They failed on 2 attempts to put the game away. Defense wins championships, however, and their defense stepped up at the end of Dallas and Green Bay. I have a hard time believing they've got it in them for one more.

Plaxico's jawing isn't going to help, either.

134
by Rick (not verified) :: Sat, 01/26/2008 - 4:16pm

#127:
I agree with the gist of your comment. But new players *can* be good. They aren't automatically good. And they will typically have letdowns, at some point. The question is when?

135
by pacificsands (not verified) :: Sat, 01/26/2008 - 5:54pm

What do the 2000 ravens, the 2001 patriots, the 2005 steelers, and the 2006 colts all have in common?

They entered the playoffs without a bye week (That is, not seeded #1 or #2), caught on fire in the playoffs, and won the super bowl. Did DVOA really see any of these four teams coming? Are the 2007 Giants really such an anomaly?

136
by Nicky P (not verified) :: Sat, 01/26/2008 - 8:16pm

135 - the 2001 Patriots had the #2 seed and a bye.

137
by vis (not verified) :: Sat, 01/26/2008 - 11:38pm

133 agree completely, Rick, thanks for clarifying.

138
by Rick (not verified) :: Sun, 01/27/2008 - 1:21am

135:
An anomaly is something that is out of the ordinary, a deviation from the norm. So yes...the Giants are an anomaly. The fact that other teams have won as a wild card, or without a bye week, is odd, not normal...and the fact that few have done it shows that the odds of completing the task are very high.

Given the Giants' inconsistent play throughout the season, I'd say those odds are extremely high, despite the recent display of excellent football.

Remember, I don't think people are saying it's a guarantee that the Pats will win. It's just very highly probable.

Add to that fact is the Giants' record after a bye week: 4-15. And Coughlin's record after a bye week: 6-7
Not exactly resounding support for continued high performance there.

Look, nobody is debating that the Giants have played VERY WELL for the last 3 weeks. But for anyone to run out stats that would support a clear case for a Giants' victory would be VERY VERY HARD. There is nothing there, except the record of the previous 3 weeks, and the overall performance of the previous 4. But the Patriots have matched or exceeded those performances over the course of the year.

My personal opinion is that the Giants are a mediocre team that gelled at a very good point in time - when they were entering the playoffs, and that allowed them to overcome long odds against 1 mediocre and 2 excellent teams, all of whom were not at the peak of their game.

Winning changes everything, though, and if they overcome the odds 1 more time, then they will have earned the right to say they were the best team this season, and they'd have the record and wins to prove it.

139
by Rick (not verified) :: Sun, 01/27/2008 - 1:37am

#125 Pat:
I agree, there was much more wrong with that game aside from Winston Justice. But even if the ONLY change was Winston Justice (Runyan healthy, Westbrook healthy, etc.), I think Osi would've had his way with him. As you said, maybe not 6 sacks, but possibly 2 or 3...which would've still been bad (though it would have kept Osi from having overinflated #s for the year).

Justice did play well against Buffalo. and that's the kind of game you want to introduce a guy in....not a game like the Giants where they have a well known and powerful rush. Of course, the Eagles game singlehandedly created the aura of that squad. Their 53 (NFL leading) sacks, of course, were racked up against an overall group of teams with losing records (102-106). Toss out 1/2 the sacks from the Eagles game, and they are tied with......The New England Patriots! Who racked theirs up against teams with overall winning records (108-100).

That the Giants began showing strength and consistency as they brought in new players is a credit to their resilience as a team, and a credit to the quality backups they have. But they are backups for a reason, and one would presume for good reasons. That's not a certainty in the professional game, of course (Wally Pip would agree), but it's usually the case (Jared Lorenzen?).

140
by Rick (not verified) :: Sun, 01/27/2008 - 1:47am

Aaron:
I remember that Joe Montana's staff once removed several plays that, upon review, showed the highest probability of him throwing interceptions. As a result, his overall performance jumped dramatically the following year.

Could it be that there is something similar going on with Eli? Perhaps a few of the more complex plays were tossed out, and this allowed them to "cut down on the mistakes"? If that's true, then DVOA is NOT invalidated, but the coaching staff found a way to reduce error and improve overall output.
And that's the essence of DVOA, isn't it? Using the stats to find ways to improve?
I think DVOA is very useful, but as with any other tool such as this, you can't expect it to answer EVERYTHING perfectly. Anomalous streaks occur within all statistical data. In fact, it goes right along with the old Vegas/Wall Street saw of "don't buck the trend". When you're winning and being lucky, that tends to clump. Consistency is finding out what works and maximizing those streaks.

The Giants did not fix their problems until very late in the year, apparently. Some part of it could be new players (Boss, etc.), some of it could be playbook alteration, some of it could be health (Burress)...but all in all, the stats don't lie. The Giants WERE mediocre all season long. And it isn't like the teams they beat in the regular season were out there on the field lighting it up.

141
by Rick A. (not verified) :: Sun, 01/27/2008 - 1:36pm

Again, people, you're putting so many factors into discussion seeking to explain how in the world did this "mediocre" NYG squad come to this ??
You want to talk about the sub players that came in, about Eli's development, about playbook changes that you imagine happened--the fact is, most of you are emotional from having lost going against them. In reality they're yet another "Wild Card" that "got hot at the right time" (something that happens now regularly in all sports) and the overview reason is not that the coaches had the wrong talent evaluation going in the first half of the year or any other such imagined explanations--the reason is that they started putting out 100% only late in the season. You didn't see it before, you made incorrect judgements about their capabilities, and you lost. You'd better watch out betting against them again too. Don't blame DVOA for your own habits.

142
by Rick (not verified) :: Sun, 01/27/2008 - 2:59pm

Rick A #141:

I mentioned getting hot at the right time as a point. And I haven't "lost" against them.
Either way, the Giants WERE mediocre, as Aaron points out, all season. Getting hot at the right time doesn't mean they weren't. It just means things suddenly came together....and it still means there is a hint of mediocrity.
Other items DO come into play, like altered gameplans, injuries, etc. But getting hot at the right time isn't usually a recipe for success. It CAN be, but isn't always. Only 4 of 9 wild card teams that made the Super Bowl have won it. The odds are not in the Giants favor.

Which means if they do win it, it is richly deserved.

143
by Dr Bob (not verified) :: Sun, 01/27/2008 - 8:53pm

The Giants haven't actually been that much better in the playoffs than they were in the regular season if you take out the fact that they turned the ball over ZERO times on offense in 3 playoffs games - which is more random good fortune than anything else since fumbles are nearly random and Manning is historically interception prone. If the Giants don't turn the ball over against the Patriots then we have a game, but if they have the standard 2 turnovers then it could be a rout.

144
by Tom D (not verified) :: Mon, 01/28/2008 - 6:30am

Re 143:

The biggest misconception of FO. "Fumbles are nearly random." That is not true. Fumble recoveries are random. Fumbles forcing fumbles and holding on to the ball are repeatable skills.

145
by Jon Coit (not verified) :: Mon, 01/28/2008 - 11:52am

121: NOW the truth comes out!!

Eli is the sum of a remainder of an unbalanced equation inherent in the programming of the matrix!

He is THE ONE!

146
by Rick (not verified) :: Mon, 01/28/2008 - 2:08pm

#143:
That's partially correct. However, the fact that Eli has NOT thrown any INTs IS "the Giants playing better" in some very large measure.
Also, the WRs have dropped many fewer passes, which is ALSO "the Giants playing better".

These are also points that I feel could, eventually, lead to the Giants reverting to the mean.
If they can keep the game close, and not turn it over, I see them having a shot as long as the score remains under 46 total points.
If they don't, it could be the Patriots in a rout.

Frankly, I see the latter as far more probable. Either way, it will be a very surprising outcome because there will be a 19-0 team, or one of the largest upsets in SB history.

I wonder, of the 4 other wild card winners, which was as bad or worse than this Giants team. I have a feeling none, considering the weak overall schedule they played.

147
by Rick A. (not verified) :: Tue, 01/29/2008 - 2:07am

I'm sorry to appear in this three times now--but I just cannot believe the number of you people arguing against the reality of the Giants being in this position. Face it--they're the best TEAM in the NFC !! They MAY not have the talent of the Cowboys or the Pack--but when is the last time anyone ever won 10 consecutive road games ?? Answer--never !! That does not happen as a fluke. What's the matter with you guys ??? This is as good of a TEAM as we've seen in a while in the weaker NFC. Those of you carrying on this dialogue, including those of you with FO, are all arguing against reality itself. It seems totally apparent none of you ever played the game. There is no such thing in football as a protracted flukey streak--especially one that happens on the ROAD--and even more especially, when it is followed up BY MORE OF THE SAME IN THE PLAYOFFS. What you should be saying right now is that if New England stops the Giants streak then that will prove that they truly deserved an undefeated season and are, indeed, a truly great team. Instead you're talking about the Giants as being one of the worst SuperBowl teams ever--completely, utterly, a load of nonsense. What will you possibly say if they should knock off the Pats ??

148
by Rick (not verified) :: Wed, 01/30/2008 - 1:56pm

#147 -
The issue isn't that we're arguing against the reality of them being in the Super Bowl. We're trying to figure out what has occurred that allowed them to make it.

How does a CLEARLY mediocre team pull off something as astounding as this?

The road wins are very interesting, in the post season. Not so much in the regular season. Only Washington had a winning record, after all. So their competition on the road was worse than mediocre.

In the playoffs, it's much more astounding what they've done, and on the road, to boot!

IF they win (which I admit is a possibility and have never argued against that possibility), then they will have well and truly accomplished something as amazing as the 18-0 current record of the Pats. 10 road wins, and one neutral site. Excellent performance, and one that is really unseen before.

The question remains - how have they pulled it off? And that is what alot of the questioning revolves around.

Today is a new article about what Gilbride has done in simplifying the playbook. Which is something, I feel, has gone a LONG way in helping them in the post season.

149
by Rick A. (not verified) :: Wed, 01/30/2008 - 4:30pm

#148-
You're a thoughtful guy. You're a student of football. Football is a sport-there is only limited science in analyzing it and that will forever be. DVOA is the best thing ever in that arena--but sports, like life itself, is unpredictable at best. All the things you and the others have talked about enter into the equation of how it happened that New York made it--but in the biggest picture what I've written about in this forum--is the main reality--and that is that people looking to predict in advance what is going to happen based on the past performance can only be partially effective at best. The Giants, like so many other teams in various other sports, did not put out at their max until late in the season. Period. People such as yourself say they are "CLEARLY" a mediocre team. Clear to who ?? Pundits, people who never played, broadcasters, gamblers ?? Mediocre teams do not accomplish what they accomplished. You talk about the sub 500 competition they faced. That's the way the NFL is--that doesn't diminish anything in any way. Many's the year the winning teams, in all the various sports, played the majority of their games against losing teams. The Giants did not have the weak schedule, they were a playoff team last year. Part of why their competition ended up sub 500 is that they could not beat the Giants. How about the Patriots ?? They played 5 sub 500 teams on the road. You guys will never find a key to figuring out in advance what is going to happen. Intangibles cannot be quantified--and, most obviously, if a team does not even show you what they're capable of until late then how can you possibly derive any tipoffs as to what they are capable of doing when they truly step to the plate ?? The New York Giants have done something that is well beyond fluke status. It APPEARS that they do not have the talent that the Patriots have after Robert Kraft purchased alot of free agents with great talent to blend with an already stellar cast. And it can be said that the NFC is weaker than the AFC and the Giants emerged from that. My absolutely final words on this are to repeat that no one should be surprised if somehow or another they pull it off anyway. Sport is a great thing guys, underdogs can and do win and plenty of times the world just cannot fathom how it happened. All the pressure is on the Patriots--we'll see how good they really are.

150
by Gerry (not verified) :: Thu, 01/31/2008 - 2:11pm

"Fumbles forcing fumbles and holding on to the ball are repeatable skills"

But they are skills which are more learn-able, for lack of a better word, than other skills are.

Tiki is probably the most obvious example. He was a fumbler. Coughlin taught him how not to be one. The change was dramatic and sudden.

151
by Rick (not verified) :: Thu, 01/31/2008 - 7:07pm

#149...
Since we share the same name, I hope we share the same temperament, LOL. I'm not getting overly excited here.

But really, this team WAS mediocre for most of the season. You admitted yourself they didn't start playing their best ball til recently, so you seem to have accepted that fact.

While some people are trying to formulate a winner out of the available information, what I see this site and most of the people on it doing is to try and figure out how the Giants "turned it around", as they CLEARLY did.

As for the records of the teams they played, only 1 game would alter the record of any of these teams....the game they played against the Giants. Toss out the NFC East, as none of them had a losing record. You've got Atlanta, Miami, Detroit, Buffalo and Chicago...take out the Giants and they still have a HORRENDOUS total record.
I won't discuss the Pats' playing 5 losing teams on the road because they went undefeated overall, and that speaks for itself....and they beat most of these teams by significant margins.

My question about the road wins is this - if the Giants are SO GOOD and we just didn't know it, then why couldn't they win at home? Is it because they played their tougher games at home (they did, clearly)? But some of those games should've been wins, like Minnesota or Washington. Even Dallas, at home, shouldn't have been such a blowout.

Clearly the team improved as they entered the playoffs. They beat 1 team that had trounced them twice, then beat another team that thoroughly thrashed them at the beginning of the year. But both these teams were past their peak as the Giants were peaking....and the Giants barely hung on to win both.

I'd argue that while yes, the Giants are playing MUCH BETTER ball than they have been, they are not playing the kind of outstanding football that is long term indicative. They are playing good game to game ball. And that is fluky.

However, in the NFL, that's pretty much all you need sometimes, and it could carry them into the Super Bowl and lead them to win it.

As you say, statistics can only carry you so far, and then they have to play the game. Nobody here is trying to "win the game" before it's played. We're just trying to figure out the angles of attack. What will the Giants' have to do to win? What can't they do? Clearly, neither team can just show up and expect to win. But in the SB, emotion rarely makes the difference. Plenty of underdogs have come into this game with all the emotion in the world, and almost all of them have left losers. The rare few that have left winners managed to have other intangibles that played a factor THROUGHOUT THE YEAR and were generally overlooked. It is only through hindsight that we're able to analyze games like the Giants over Buffalo or Jets over Baltimore and determine why they pulled off the upset. Going into either game, if you put money on the underdog you were simply betting with your heart, and we see more money lost that way than not. It's why Vegas remains in business.

So I'd say that most of your discussion points are accurate but not precise. Everything you pointed out is true. But it doesn't eliminate the fact that this team played a fair to middling regular season and scratched into the playoffs, only to suddenly set the world afire.

152
by Kurt (not verified) :: Thu, 01/31/2008 - 10:14pm

As for the records of the teams they played, only 1 game would alter the record of any of these teams….the game they played against the Giants. Toss out the NFC East, as none of them had a losing record. You’ve got Atlanta, Miami, Detroit, Buffalo and Chicago…take out the Giants and they still have a HORRENDOUS total record.

Miami and Atlanta are the only horrendous teams in that bunch, and even Atlanta was 3-4 at home (excluding the Giants game). Add in the 3 NFC teams (which the Giants did play, after all) and you have 8 road games against teams who in the aggregate were over .500 at home when not playing the Giants.

As for why the Giants were so much better on the road, who knows. They had two losses against non-elite teams, and both happened to be at home. I don't know how much meaningful data can be extracted from two games.

Clearly the team improved as they entered the playoffs. They beat 1 team that had trounced them twice, then beat another team that thoroughly thrashed them at the beginning of the year. But both these teams were past their peak as the Giants were peaking….and the Giants barely hung on to win both.

What is the evidence that Green Bay was past their peak, other than the Giant game itself?

153
by Rick (not verified) :: Thu, 02/07/2008 - 1:41am

The "proof" that GB was past its peak was a trouncing by mediocre Chicago squad. Their final 3 games were against cakewalk teams, theoretically. Lousy St. Louis on the road, which they won handily. Then Chicago, to stay in the home field race...which they lost handily. Then home against a slumping Detroit.

They dismantled Seattle at home, but only after showing some shaky legs early on.

Then the Giants. They had several chances to put the Giants away, and didn't.

I think alot of people on this thread read posts like this and assumed it was meant to say the Giants didn't have a chance against the Pats. If you reread all the previous posts, I know I have said they did have a chance, but several things had to happen. As it occurred, these things DID happen.

I wouldn't say that the Pats LOST the game, because the Giants clearly controlled the tempo of the game and made their plays when they had to. But their defense WON the game more than Eli or their offense did. In their 3 games running up to the SB, there was more balance between the offense and defense for the Giants. Eli was performing at a high level. His performance in the SB wasn't as good as the previous 3 games, but the defense stepped it up big time, and that made a huge difference.

All things considered, the game was a MAJOR upset and anyone who said they KNEW the Giants would win was simply speaking from their heart and not the facts.

To be fair, I had told all my betting friends (I live in NYC) that the BEST bet was to take the underdog straight up. Why? Because if you're a Giants fan, nothing would be worse than winning the spread bet but losing the game. More importantly, as I pointed out to them, if the Giants were to cover the spread, they would in all likelihood win, because for the Pats to cover, they'd have to control the game from the start and impose their will on the Giants.

Giants fans, unwilling to accept that could happen, seemed more willing to bet the spread thinking they'd lose but earn a little cash. I said, perhaps, but if they actually win you get 2.5X the cash...and if you think they can cover, then you have to believe they can win.
My logic made alot of my Giant fan friends some serious coin.

My bet? I didn't place it...because I had no stake in the game...was to parlay the under and the Pats covering. I'm glad I didn't.