Writers of Pro Football Prospectus 2008

29 Dec 2010

Bill Polian Thinks We Are Idiots

Sigh. You know how I often say that FO isn't trying to get rid of scouting? That having better stats doesn't mean scouting isn't important? You know how I often say that as good as our stats are, they are even better if you can explain them in context? You know how I often say that our individual defensive stats don't necessarily tell you which players are good or bad, that they are better for explaining the "shape" of the defense and where the plays happen?

Apparently, Bill Polian hasn't read those comments.

Worse, if you read a lot of Polian's comments about advanced stats, you can see a general theme: "Fans in general are morons who know nothing about football." He belittles any attempt where people outside the game try to understand how the NFL works. Does Bill Polian understand that he is in the entertainment business? Does he understand that anything which increases interest in football is good for the National Football League?

Posted by: Aaron Schatz on 29 Dec 2010

251 comments, Last at 08 Jan 2011, 5:49pm by Anonymous1

Comments

1
by loneweasel (not verified) :: Wed, 12/29/2010 - 11:56am

It's wonderful that you say those things.

But on the other hand, FO is pretty good at sabotaging its own credibility by dogmatically pushing garbage like the 370 curse and the Lewin projection, etc.

4
by Abdul Nonymous (not verified) :: Wed, 12/29/2010 - 12:07pm

Both of those things are quite accurate in the general, even if flawed in the specific.

Obviously runningbacks don't explode on their 370th carry. It is true, the power of that number has been implicitly overstated more than a few times. At the same time, its premise is incredibly important; that runningbacks tend to wear down badly once they get to a fairly high carry level. The importance of this theory is magnified by the fact that many NFL teams seem not to realize this.

Lewin projections are obviously not perfect by a long shot, and any attempt to make them a definitive argument-ender is foolish indeed. However, they do help make good observations about players, and most of all work to discourage teams from drafting unpolished QBs with cannon arms, who on the whole have unusually poor performances in the NFL.

Both theories can be (and at times, have been) overstated in their specific importance. But both theories contribute a greater understanding to the analysis of professional football as a whole, and do not merit your scorn.

8
by DeltaWhiskey :: Wed, 12/29/2010 - 12:18pm

"At the same time, its premise is incredibly important; that runningbacks tend to wear down badly once they get to a fairly high carry level."

Perhaps within a season, but there is insufficient evidence of a hangover effect to the next season.

"Lewin projections are obviously not perfect by a long shot, and any attempt to make them a definitive argument-ender is foolish indeed. However, they do help make good observations about players, and most of all work to discourage teams from drafting unpolished QBs with cannon arms, who on the whole have unusually poor performances in the NFL"

The Lewin Projection, if useful at all, is only good after the draft due to the round one/two specification.

53
by Arkaein :: Wed, 12/29/2010 - 1:07pm

It's actually useful before the draft in that proper application would discourage teams from drafting QBs with low completion percentages or few games started.

If GMs used it Jamarcus Russell and Kyle Boller would not have been high draft picks.

It can't tell you which QBs should be drafted on its own, that's true. But it has done a good job demonstrating that GMs have historically overvalued certain measurables and undervalued accuracy and either experience or the ability to to do extended scouting over multiple seasons.

214
by loneweasel (not verified) :: Thu, 12/30/2010 - 3:24pm

On the other hand, it would tell you not to draft Jay Cutler, Matt Ryan, Matthew Stafford or Josh Freeman, just from recent years. Either the system is broken or it was a historical artifact due to unique circumstances of the late 90's-early 2000's.

219
by Arkaein :: Thu, 12/30/2010 - 4:38pm

I suppose I haven't seen or paid as much attention to more recent projections.

I think it's quite plausible that the system successfully identified two major flaws in NFL QB drafting, and that more recently GMs have either rectified those flaws or improved their scouting in other ways which have rendered the Lewin system less useful.

221
by tuluse :: Thu, 12/30/2010 - 6:02pm

Jay Cutler

My projection system suggests that Cutler will perform somewhere along the lines of Drew Brees or Byron Leftwich (an average of about 2.5 points above replacement per game over the course of his career), and this does not even take into account his running ability. This would make him a solid top-5 choice

I can't find Matt Ryan's Lewin projection, but given that he had a lot of starts, I'm guessing he had a high projection too. Since you're making things up in the first place, I feel comfortable with that guess.

No one knows if Stafford should have been drafted high (playing like Shaun Hill is not a good sign though).

222
by Sander :: Thu, 12/30/2010 - 8:04pm

Lewin Forecast actually said that Freeman would be the best QB from that draft, though it didn't predict him to be as good as he is.

11
by loneweasel (not verified) :: Wed, 12/29/2010 - 12:19pm

Regression to the mean. Historically NFL teams undervalued college completion %. There, those statements are far more accurate, useful and succinct than the two FO is still pushing.

The curse and Lewin projection should have been fun curiosities people could look back on with nalstagia, something like Bill Jame's pitcher game scores. But FO is touting them as useful tools TODAY, which is more than a bit self-congratulatory and dishonest.

56
by Abdul Nonymous (not verified) :: Wed, 12/29/2010 - 1:10pm

Do you get angry at Coke because it claims to have some intrinsic virtue over Pepsi? Do you become irate when you see those cute Coke(tm) bears mugging for the camera - does it seem self-congratulatory and dishonest? After all, these commercials don't exist because Coke has some intrinsic superiority over the competition; they exist to increase consumption of the product.

Football Outsiders is, within reason, a business. It sells a product, and its ability to produce that product well is dependant on how well the product sells. Accordingly, it is somewhat naive to think that FO should be restricting itself to absolute truth at the expense of their own brand surivival and propitiation. Imagine if FO had an article with this wording: "Larry Johnson just got his 412th carry of the season in Week 17. He may wear down - some backs do - but no promises. Anything can happen. It may just be a regression to the mean, who knows. I mean, we did some research (nothing amazing mind you) and found that a lot of carries *can* be bad, generally, for runningbacks. But no conclusions. Just keeping it real."

Who would read that? I wouldn't. Honesty is laudable, but somewhat impractical in this context. FO has no incentive to downplay their own achievements and value, and plenty of incentive to do the opposite. It's just good business. Furthermore, most sportswriters are crazy assertive whether or not they should be and their popularity is usually tied to their confidence, not their accuracy. (*cough*PeterKing*cough*)

I would much prefer FO to make overambitious assertions that encourage their growth and development, and trust myself to evaluate each claim within its context. But that's me.

64
by loneweasel (not verified) :: Wed, 12/29/2010 - 1:24pm

If their goal is sensationalist snakeoil salesmanship, which I do not completely believe, then Polian's criticism is entirely justified. Isn't it?

Or they could grab eyeballs by cranking out better, more accurate data and presenting them better. Hard as it is to believe, one could actually write about regression to the mean (or any other statistical concept) with both clarity and humor. It would take more chops than inventing a dishonest curse, of course, but it's better business. There are baseball writers who could do it. There are also baseball writers who are dishonest and sensationalist. The latter get laughed at after a while. The former gets creds.

To persist with a dishonest multiple-endpoints trick like the curse would be akin to a weather forecasting agency that knowingly uses a bad model instead of better ones. Their choice which way they go, I guess.

67
by drobviousso :: Wed, 12/29/2010 - 1:36pm

So all it takes to be better, is to be better? Why didn't I think of that.

70
by loneweasel (not verified) :: Wed, 12/29/2010 - 1:39pm

The first step is to stop knowingly and doggedly do bad. You might have missed that.

69
by Abdul Nonymous (not verified) :: Wed, 12/29/2010 - 1:37pm

I'm pretty sure that you have misunderstood my point. Alas.

77
by DeltaWhiskey :: Wed, 12/29/2010 - 2:09pm

The key point you seem to be making is that deception is okay if it makes a buck?

86
by Abdul Nonymous (not verified) :: Wed, 12/29/2010 - 2:27pm

Is Coke deceiving people when they talk about how great their soda is? Is Peter King deceiving people when he takes positions that he believes are true (even if very probably false?) Are you deceiving people when you make it sound like FO posts something on the Curse of 370 and then cackles up their sleeve at the sheeple reading it?

People generalize when they communicate. Do you want a multi-paragraph breakdown of the exact accuracy of the high-workloads=subsequent-off-year premise *every time* the concept is referenced? I suspect not. FO writers do what we all do; when they want to make a reference to the premise, they do, and move on. They don't go into all the weaknesses (and strengths) of the premise every time they reference it because it is simply inefficient to do so. In the process they necessarily overplay its accuracy. Imperfect? Of course. But this does not constitute a deception of any kind, any more than the above examples do.

91
by loneweasel (not verified) :: Wed, 12/29/2010 - 2:36pm

People who don't believe Coke's propaganda might still buy it and enjoy it because it's a cheap sugary beverage.

On the other hand, FO has nothing to sell other than information. If people don't believe their information, then there is no reason for FO to exist.

Simple distinction.

100
by Abdul Nonymous (not verified) :: Wed, 12/29/2010 - 3:15pm

You are drawing a false distinction. Both are companies that sell something that people like to consume. Both companies, at times, release products that perhaps should not have been released (new coke anyone?) If a sufficient quantity of their products are deemed defective, they will go out of business, having lost the popular goodwill necessary to maintain sales.

What you seem to be saying (to continue the comparison) is that FO released a product that they knew or should have known was defective/sub-par. Which is not an unreasonable position. But I can't really imagine anyone (assuming a similar situation with Coke) taking the position "how can you stand Coke as a company - they knew New Coke was going to suck and they released it anyway! Unacceptable." It would seem that such an act primarily hurts the company and not the consumers.

At any rate, the two models actually seem fairly similar except for one point: that FO's market niche is built on the assumption that their products are better (more accurate) than everyone else's. But again, not sure how this is such a big deal on the consumer end.

107
by DeltaWhiskey :: Wed, 12/29/2010 - 3:29pm

Except Coke stopped selling New Coke. Curse of 370 is still for sale.

Taking your analogy a little farther. Coke put out New Coke, it bombed, had Coke tried to put out New Coke v1.1, I think the consumer would have been wary of the product. With Curse of 370, we are required to be wary of the latest FO product because: 1. We know they are capable of putting out crap, and 2. They are willing to continue "selling" the crap.

111
by Abdul Nonymous (not verified) :: Wed, 12/29/2010 - 3:34pm

Which is not an unreasonable position.

Incidentally, did you know that Castro (who was a long-time coke drinker) railed against New Coke as a sign of the decadence of capitalist society?

Hilarious.

112
by DeltaWhiskey :: Wed, 12/29/2010 - 3:35pm

Cuba Libres with New Coke taste like shit.

198
by Mikey Benny :: Thu, 12/30/2010 - 2:31pm

Dupe.

197
by Mikey Benny :: Thu, 12/30/2010 - 2:31pm

loneweasel,

You care way too much about things that don't matter.

212
by loneweasel (not verified) :: Thu, 12/30/2010 - 3:14pm

wrong, I don't care about you at all.

226
by Cid (not verified) :: Fri, 12/31/2010 - 2:51am

There's nothing faulty about the information though. What you seem to be actually objecting to is the short hand reference to the longer and more thorough treatments of these subjects.

87
by RichC (not verified) :: Wed, 12/29/2010 - 2:28pm

The whole point of FO was to debunk conventional wisdom that was flat out wrong. They're now pushing their own "wisdom" that is just as wrong.

You have no problem with that?

96
by Abdul Nonymous (not verified) :: Wed, 12/29/2010 - 2:53pm

I guess I just can't get that excited about it. FO isn't perfect, any more than any other pundit. I have looked into the whole 370 thing at your implied urgings, and it does look fairly shoddy as a conclusion. But its occasional reference simply doesn't make me indignant the same way it seems to be making others. FO is a source of information and analysis, all of which is imperfect, some of which is quite flawed. And as I trust my ability to discern the difference between good and bad analysis, the occasional misstatement is simply not that big a deal. But that's me.

98
by DeltaWhiskey :: Wed, 12/29/2010 - 3:08pm

"But its occasional reference simply doesn't make me indignant the same way it seems to be making others."

It's not the occassional reference, it's the continued reference in the face of independently generated evidence to the contrary that is problematic.

99
by loneweasel (not verified) :: Wed, 12/29/2010 - 3:12pm

Exactly. It's not about the error. It's repeating the same error and total resistance to acknowledge or correct the error.

102
by Abdul Nonymous (not verified) :: Wed, 12/29/2010 - 3:22pm

I can understand that.

224
by Bowl Game Anomaly :: Fri, 12/31/2010 - 12:32am

Actually, the good research (conducted not by FO but by P-F-R) suggests that backs wear down from heavy use during individual games (carries in the high 20s and up) rather than over the course of a season, and IIRC more specifically in wins than losses due to the nature of the playcalls. It comes down to something like "A high number of carries over the season (like 370) probably implies some brutal individual game workloads which tend to lead to injury and loss of effectiveness down the road."

179
by RickD :: Thu, 12/30/2010 - 1:18pm

Did you read the comment linked to?
Nice job of hijacking the comments section, BTW. Polian makes an absurd, unwarranted, wholesale attack on the validity of statistics at all. And your argument is to pick a few arguments that you don't like and thereby offer credibility to Polian's blanket assertion?

Bad form.

190
by DeltaWhiskey :: Thu, 12/30/2010 - 1:49pm

I'm glad you posted this response, b/c as I came back to re-read I was thinking of the Curse and Lewin system again, but esp the Curse, but I didn't want re-hijack the thread.

The marked refutation by others demonstrating the questionable validity of the Curse combined with the continued promulgation of the Curse at FO provides plenty of fodder for Polian and anyone else who wants to dismiss FO and any other advanced stat work. FO would be well advised to revisit the Curse in some manner or other and either provide better evidence to support it and refute the naysayers, or admit that perhaps they were premature in their assertions or flat out wrong. The diligence with which Aaron et al. work towards improving DVOA and DYAR is admirable, keeping the Curse as part of "Football Outsiders Basics (a.k.a. "Pregame Show")" to some degree undermines these efforts and does not serve them well.

207
by loneweasel (not verified) :: Thu, 12/30/2010 - 3:07pm

You don't know how the credibility thing works, do you?

One professor caught plagiarizing or one department caught falsifying data impacts the reputation of the whole university. If the university then doesn't take any actions, then they lose all credibility.

I'm not just accusing FO of making a mistake. We are way past that. I'm stating that their actions constitute knowingly and dishonestly perpetuating their mistakes. If they are really interested in truth and better analysis, then they should answer their critics and go from their. If their goal is to be a GIGO entertainment shop for ESPN, then they can ccontinue their current ways by not even acknowledge the issue.

Bad form? So what?

213
by tuluse :: Thu, 12/30/2010 - 3:17pm

I don't think you know what dogmatically means. I can't even remember the last time an FO writer has used the curse of 370 in serious tone. Most of the time they've shifted to talking about just general overuse.

I don't see how the Lewin projection system figures into any of this. In the explanation they end with "However, the Lewin system has mixed successes (Kevin Kolb) with failures (Brady Quinn, Matt Leinart, Brian Brohm) in recent years, and we're likely to revisit it in the near future."

241
by fillylabinga (not verified) :: Sun, 01/02/2011 - 10:45am

Furthermore, can we all please agree that both the Lewin and 370 Curse things have *no relation to the core approach/value of Football Outsiders*--that being a situationally-adjusted analysis of the success/failure of individual plays? Maybe FO is taking an unnecessary risk by folding them under the FO banner (they are certainly inviting criticism by doing so), but these theories are only tangentially-related at best. This thread is blowing things way out of proportion.

243
by DeltaWhiskey :: Mon, 01/03/2011 - 10:02am

"the core approach/value of Football Outsiders*--that being a situationally-adjusted analysis of the success/failure of individual plays"

Agreed, but (there's always a but) how do we know that DVOA/DYAR don't suffer similar shortcomings? What do we know about the validity of the DVOA/DYAR system?

244
by tuluse :: Mon, 01/03/2011 - 1:49pm

Well DVOA correlates well to winning, so we have proof of correlation, if not causation.

245
by DeltaWhiskey :: Mon, 01/03/2011 - 3:53pm

Yes it does and you are correct, but the improvement in that correlation with each "improvement" of DVOA has not been shown to be a statistically significant improvement to my knowledge.

Don't ask me why I continue to beat this dead horse...I guess it's just what I do.

246
by greybeard :: Mon, 01/03/2011 - 7:34pm

Win-loss records correlate to winning better than DVOA.

247
by tuluse :: Mon, 01/03/2011 - 7:49pm

But not to win-loss record the next year, where DVOA is king over win%, pt dif or any other measure I know of.

248
by greybeard :: Tue, 01/04/2011 - 3:05am

It may be slightly better than the win loss record from previous year but not in a meaningful way. It was like .21 correlation in 2009 which is pretty meaningless. DVOA of current year correlates slightly better with the DVOA of the year before (.25 I believe), which is still meaningless and I am sure nobody cares to know what their teams DVOA will be next year but everybody would be interested in finding out what their win-loss record would be.

I would love to see an article from FO how the DVOA correlates to win-loss record next year and what it means in practice.

Until then check out http://www.advancednflstats.com/2010/07/pre-season-predictions-are-still....

249
by greybeard :: Tue, 01/04/2011 - 3:08am

BTW, my comment about win loss record correlating better with wins was for the same year. It was a joke on that you did not mention in your post that you were talking about the win-loss record for the next year and not the current year.

250
by tuluse :: Tue, 01/04/2011 - 3:16am

I was talking about current year, and I got your joke. I took it seriously as to make the point that DVOA has predictive qualities.

As for advancednflstats, I don't like that site. The author seems to have an axe to grind at FO's expense, and his tone often seems spiteful. No thank you.

Also, 2009 was by far FO's worst year for predictions so talk about "cherry picking data" Mr Burke.

216
by Noah of Arkadia :: Thu, 12/30/2010 - 3:40pm

Is there an organization I can join to stop this? Or should I simply write my congressman about it?

217
by Nathan :: Thu, 12/30/2010 - 4:17pm

nice

2
by Jim D (not verified) :: Wed, 12/29/2010 - 12:02pm

It seemed as if Polian was saying stats need to be looked at in context, which is exactly what you say in the intro here. You may be overreacting.

I thought it was interesting later in the article how he talked about their plans for pumping in extra loud crowd noise this year in a home playoff game.

12
by Sander :: Wed, 12/29/2010 - 12:20pm

While you can read that into his statements, I don't think that's what his words actually say. His words say that all (public) statistical analysis is useless because there's no context or possibility to create context.

I think the disconnect here is between what Polian sees as useful in-house, and what FBO sees as useful toward the public. I had the same feeling about his statements at the MIT/Sloan conference. Basically, he says that all the statistical analysis sites like Football Outsiders do is useless to the Colts in general. That makes perfect sense, as it'd be pretty disappointing if the Colts couldn't do a lot more than Football Outsiders can. But in dismissing that analysis as useless for NFL teams, he also denies their use to the public - and that's the problem.

22
by loneweasel (not verified) :: Wed, 12/29/2010 - 12:28pm

The answer to that, of course, is not whining. Improve your stats instead, so your publically available data could be useful to professional franchises.

Look at baseball. Outside statistical analysis has and is helping MLB teams.

Polian isn't saying that outside stats are inherently and irrevocably useless. He's saying that they are right now.

23
by Aaron Schatz :: Wed, 12/29/2010 - 12:32pm

We consult for NFL teams. And a number of other teams use our publicly available data. The Colts just aren't one of them.

32
by Sander :: Wed, 12/29/2010 - 12:42pm

I've heard you say that before and it honestly surprises me. I know most teams analyse every game with all-22 film in depth and do their own statistical analysis. They really can't do a better job than you can? And I say that with all due respect, I love what you do but you don't have nearly as much information, film or expertise as the teams themselves.

34
by Aaron Schatz :: Wed, 12/29/2010 - 12:45pm

Who says they use our stats and ignore the all-22 film? They use them together.

41
by Sander :: Wed, 12/29/2010 - 12:50pm

I didn't mean to imply that. What I'm wondering is what your statistical analysis adds to what they can't already do themselves. I love what you guys have done, but I can't fathom that teams can't (and don't) do this same analysis in-house.

Unless they're looking for alternate viewpoints to break possible groupthink, which would make sense.

45
by Aaron Schatz :: Wed, 12/29/2010 - 12:52pm

They can't do that stuff for all 32 teams! Our information on league-wide trends and the other 31 teams is often very useful. They are rarely looking for us to tell them things about their own team.

52
by Sander :: Wed, 12/29/2010 - 1:03pm

Well, that makes sense.

I guess I'm expecting NFL teams to have a full complement of statisticians and game analysis people on staff, but that's probably not a realistic expectation.

62
by Trust Doesn't Rust (not verified) :: Wed, 12/29/2010 - 1:21pm

I think you think being a statistician is more glamorous than it is. These are the types of jobs that, in-house at an NFL franchise, an assistant or even intern might be tasked to do ("hey boy, watch this random buffalo-carolina game and calculate how many passes x cornerback gets thrown to, who he's covering on each pass, and how many passes are incomplete due to good coverage.") Why wouldn't an NFL team outsource this if it could, especially if it could receive a large amount of information in a short period of time?

132
by The Other Ben Johnson (not verified) :: Wed, 12/29/2010 - 6:42pm

Being a statistician isn't glamorous? Unbreak my heart, Trust Doesn't Rust.

49
by MJK :: Wed, 12/29/2010 - 12:55pm

Teams have access to better data than FO does, but not all teams might have good statisticians that know how to glean useful information out of data. I work at a research laboratory, where you would think everyone is pretty smart (and they are), but there is a surprising lack of understanding of statistics in just about every group except that statistician group. I'm not a statistician, but I appreciate the value that they bring, and every project that I'm a part of that I've managed to convince them to bring in a statistician, the results have been good.

A lot of time people have good data, but for whatever reason, they aren't making full use of it. A methodology like DVOA, but tuned to the better data that teams have, could become a powerful weapon if used correctly.

71
by drobviousso :: Wed, 12/29/2010 - 1:41pm

"... but there is a surprising lack of understanding of statistics in just about every group except that statistician group."

This is true in many disciplines. Stats, like econ and psych, is totally in practice from what you learn in the first 2 or 3 college classes you take on the subject. Unfortunately, for many very smart people, 2 or 3 is the most they'll ever take, maybe supplemented with some poker know how.

227
by Cid (not verified) :: Fri, 12/31/2010 - 3:00am

It might be cheaper to rent the analysis than to build it yourself.

232
by Raiderjoe :: Fri, 12/31/2010 - 7:29am

If ytou build it they will come.

Build football fiekd in backyard (if have big enouigh backyard l;ike if live in rural settng) bronko Nagurski, mel hein, al blozis, and somr others might come out of vegetation and plsy football for you. So get your pork rinds and beer or soda pop ifb not 21 yet

60
by tbwhite :: Wed, 12/29/2010 - 1:19pm

I'm sure the NY Giants probably do a better job analyzing the NY Giants than FO can do. But I'm also willing to bet that the Giants don't wish to spend the resources to analyze EVERY NFL team like they do their own, perhaps the next team or two on their schedule and that's probably about it. So, what FO can't provide to an NFL team in depth, they can make up for in scope and context by providing data on the entire league.

50
by Jim D (not verified) :: Wed, 12/29/2010 - 1:00pm

Not sure this is the best way to go about landing them as a client :)

180
by RickD :: Thu, 12/30/2010 - 1:22pm

I think it's a good way to defend a product against the criticisms of somebody who obviously doesn't want them as a client.

All sycophantism, all the time, really isn't the only business model.

40
by Dave :: Wed, 12/29/2010 - 12:50pm

They are, and always will be, useless for him.

I agree about the disconnect between use for fans and use for teams. I just think he's overdoing things by speaking in absolutes. He's fond of taking any angle that will make him seem smarter and/or more indispensable. That's what's happening here. Instead of saying "they're better than the fantasy stats, but there's still so much more to what we do that can't be captured" he has to go overboard and say that everything's worthless and dumb. It's just how he is, for better and for worse.

120
by SFC B (not verified) :: Wed, 12/29/2010 - 5:25pm

Baseball also makes EVERYTHING available to the public with accurate play-by-play data going back... what... 70 years?

I think it's silly, even possibly rude, to be mad at a site like FO not being able to provide the same level of information as a site like BP, when their respective sports are decades apart in accomodating the desires of their more hardcore fans.

MLB has Pitch FX data available, what, the day after a game? The NFL can't even be bothered to EVER release All-22 film.

FO can only improve their stats so much given that the sport they cover refuses to offer better stats.

125
by Dave :: Wed, 12/29/2010 - 5:57pm

Actually, PitchFX data is available immediately following every at bat.

3
by drobviousso :: Wed, 12/29/2010 - 12:04pm

Didn't he, or someone posing as him, post on here after the Sloan/MIT video where he bashes statistics? Was that ever verified?

The fact of the matter is that it doesn't really matter what he says publicly about it. If it work, it will get used at some point, especially as more "quality control" people turn into head coaches, since QC coaches are more likely to be computer and math literate.

The culture of secrecy is probably good for each individual team, but bad for the league as a whole.

5
by jklps :: Wed, 12/29/2010 - 12:09pm

Not saying BP isn't a good GM, but if he is so right about everything, how come a team with a legendary QB has only won a single Super Bowl while having a huge crutch in not ever having to think about drafting a QB?

16
by loneweasel (not verified) :: Wed, 12/29/2010 - 12:21pm

Fun fact Pats fans might not understand: Football didn't start in the 2000's.

19
by Nathan :: Wed, 12/29/2010 - 12:25pm

Fun fact you may not understand: Peyton Manning never played for the Baltimore Colts.

28
by jklps :: Wed, 12/29/2010 - 12:35pm

I'm not a Pats fan, or a fan of any team that's won the Super Bowl in the last 15 years.

I'm just saying it's an amazing luxury for the Colts, Pats, and now Saints to be so set at the QB position...and yet, the Colts, with Peyton Manning who is amazing(as is Brady, Brees) have only won a single Super Bowl.

36
by Snack Flag (not verified) :: Wed, 12/29/2010 - 12:47pm

I think he's pointing out that Bill Polian has been around longer than just Peyton Manning's career and developed those early nineties Bills teams that did pretty well. They never won a SB, but some of that was luck and their resume is impressive regardless of the losses.

38
by Sander :: Wed, 12/29/2010 - 12:48pm

Also did a good job with the expansion Panthers, getting them to the NFC Championship game in just two years.

39
by loneweasel (not verified) :: Wed, 12/29/2010 - 12:48pm

Carolina too.

The man's a giant dbag. But to doubt his GMing acumen is pretty out there.

43
by Aaron Schatz :: Wed, 12/29/2010 - 12:51pm

By the way, I agree. He's one of the greatest GMs in history.

146
by Anonymous Jones :: Thu, 12/30/2010 - 12:32am

Only fools would disagree that he is a good GM.

The "championships won" stat is from a *very* small sample size. You have to look at a larger sample size (and then using context at every turn) to determine whether there is any GM acumen on display. Hard to argue that Polian hasn't exhibited a lot of that.

158
by DeltaWhiskey :: Thu, 12/30/2010 - 9:04am

Esp considering that eight teams have won 31 out of the 44 (70%) Super Bowls played.

PIT 6
DAL 5
SF 5
NE 3
WAS 3
LA/OAK 3
GB 3
NYG 3

And the remaining 13 games are divided amongst 10 teams

182
by RickD :: Thu, 12/30/2010 - 1:28pm

That list of teams shows that the history of football is marked by a small number of experts who have been able to build teams that have been able to win multiple Super Bowls.

Pittsburgh and Dallas are exceptional in having two separate periods where they've won multiple Super Bowls.

171
by nat :: Thu, 12/30/2010 - 12:18pm

While I think Polian has been a very effective GM, isn't your opinion (and mine) that he has exhibited a lot of acumen based entirely on small sample sizes and anecdotal evidence?

Each season there's only one draft, one champion in a conference, one Super Bowl, one QB to hire, draft, keep or cut.... If picking and keeping a QB (which comes up maybe 5 times in a GM's career) is to be considered, then certainly the ability to orchestrate championship seasons is a more reliable measure of a GM's overall skill.

To put it another way, if Polian had drafted Ryan Leaf instead of Peyton Manning, would we even be having this discussion? If the Colts hadn't had a top pick that year, would Polian still be considered great? Isn't that a sample size of one?

183
by RickD :: Thu, 12/30/2010 - 1:34pm

A good example of a misapplication of the "small sample sizes" argument.

Would anybody argue that the argument that Leonardo da Vinci is a good paint is poorly supported, since the existence of the Mona Lisa is a "small sample"?

Not to mention that your argument completely ignores his work at Buffalo and Carolina. Nobody in the history of football has had the success in multiple cities that Polian has had. Polian has built teams that have gone to multiple Super Bowls with two separate franchises. Nobody else has done that.

"If Polian had drafted Ryan Leaf instead of Peyton Manning, would we even be having this discussion?"

If da Vinci had given la Joconde a moustache, would we be having this discussion?

194
by Noah of Arkadia :: Thu, 12/30/2010 - 2:14pm

If you hadn't written this post, would we even be having this discussion?

46
by jklps :: Wed, 12/29/2010 - 12:53pm

I'm not doubting his GMing acumen. I'm just saying he isn't always right, for example in this case.

And any GM with a brain knows you need to figure out your QB first, which he has obviously had in Buffalo and Indy.

47
by Dave :: Wed, 12/29/2010 - 12:53pm

Well, there have been those pesky Patriots and Steelers, each of whom are also pretty well set at QB and play in the same conference.

There's some bad luck, some injury luck (which, ultimately, I blame on the staff), some weak/timid coaching, some other very good teams, and a million other reasons too.

Winning "one single super bowl" is actually quite an accomplishment, if you think about it. Most teams don't win any. The only measure by which the Polian/Manning Colts are a failure is the "are they as good as the Patriots" one.

74
by Will Allen :: Wed, 12/29/2010 - 1:59pm

Change a Zebra's call and a couple of longish field goals, and the Pats may well have fewer Super Bowl Championships than the Colts.

Judging GMs by the won-loss outcome of a miniscule sample size of closely decided games is really foolish, just like it is for qbs and coaches.

101
by Jim D (not verified) :: Wed, 12/29/2010 - 3:15pm

It's not Polian's fault that the Colts only have one title..It's Manning's. Most of his career Peyton only cared about putting up gaudy numbers, while Brady just wants to win.

104
by Nathan :: Wed, 12/29/2010 - 3:23pm

Actually what happened was Peyton was a certified Loser. He could Never Win The Big Game until the Winner Fairy visited him in his sleep and whacked him with the Winner Stick, after which he Could.

108
by loneweasel (not verified) :: Wed, 12/29/2010 - 3:30pm

And after the 2005 Superbowl, a magical troll of Losing visited Tom Brady. Afterwards he lost all Desire of Winning and only wanted to compile Gaudy Stats.

The magical troll's name? Josh McDaniels.

202
by Mikey Benny :: Thu, 12/30/2010 - 2:44pm

Nice one :)

105
by CoachDave :: Wed, 12/29/2010 - 3:28pm

You are trying too hard.

106
by CoachDave :: Wed, 12/29/2010 - 3:28pm

(Duplicate post)

184
by RickD :: Thu, 12/30/2010 - 1:38pm


Change a Zebra's call and a couple of longish field goals, and the Pats may well have fewer Super Bowl Championships than the Colts.

What a profoundly silly argument!
Change a few refs' calls and the Pats beat the Colts in the AFC Championship game in 2006-7, not to mention the Super Bowl against the Giants. So, modulo the referees, the Patriots won 5 Super Bowls in 7 years! (Yes, I'm ignoring the Bears.)

187
by tuluse :: Thu, 12/30/2010 - 1:47pm

All you have done is support Will's statement. Whether a team wins a championship or not can come down to tiny things which don't really say anything about whether the people in charge are good at their jobs or not.

196
by Will Allen :: Thu, 12/30/2010 - 2:30pm

It was such a profoundly silly point that you decided to agree with it!

I'll try again. Very close football games are determined to a great degree by random events that are non-predictable. Vinitieri has been a nice kicker; that doesn't mean, as we have seen in other games, that he can be expected to make every long kick. An official can see a very small arm movement, or he can not see it, and thus make a different call. A football can have sufficient friction with a helmet to stay pinned there after the receiver falls to the ground, or it can slip. These games are essentially ties, where luck determines the outcome. It is really foolish, when evaluating the performance of any coach, player, or executive, to look at the won-loss outcome of a very tiny number of closely decided playoff games, to gain insight, especially when the coach, player, or executive has a total sample size of a couple hundred or more games to be evaluated with.

116
by JonFrum :: Wed, 12/29/2010 - 4:32pm

Really?

(Says someone who remembers the glory days of Babe Parelli)

181
by RickD :: Thu, 12/30/2010 - 1:23pm

What a remarkably inane comment.

6
by ODBvernon :: Wed, 12/29/2010 - 12:12pm

I didn't read his statement as being clearly negative or dismissive of advanced stats. Especially since he basically allows for the fact that the Colts are using advanced metrics themselves: "there are some [metrics] we use where you do match it up with a front or a coverage. We do some pretty sophisticated analysis of it, but you have to match it with fronts and coverages."

Given his choice of targets as being representative of the stats he's dismissing, I highly doubt he's truly bashing advanced statistical analysis - more just being the cantankerous, crotchety shit that he is (and this coming from a Colts fan who adores him because I remember pre-Polian all too well). Who knows, but he has probably recieved some questions about, or criticisms of, Wayne and Garcon and their low catch rates/high target percentages and responded emotionally defending the quality of his guys and their seasons.

It really seems to me that your counterpoint to his statement is just as reductive as what you are critiquing in the first place. He clearly does not throw advanced stats out the window - although you are right that he likely does think most football fans are idiots, since most of them surely are. Just not the tiny minority who truly understand and embrace real advanced stats like the small FO-community.

18
by Spielman :: Wed, 12/29/2010 - 12:24pm

"'There are two statistics that count on defense: points surrendered and turnovers.' That's all that counts. Everything else is meaningless."

It takes a helluva lot of rationalizing and redressing to explain that that doesn't mean what it seems to mean.

29
by andrew :: Wed, 12/29/2010 - 12:39pm

By that logic, why should turnovers matter anyway, its just points surrendered. If a team forces no turnovers and surrenders 0 points, is that not superior to one who forces two and surrenders 52?

33
by Sander :: Wed, 12/29/2010 - 12:43pm

Turnovers are useful because it gets the offense the ball back and hence increases the chance to score.

Of course, 3-and-outs do the same thing.

35
by Nathan :: Wed, 12/29/2010 - 12:45pm

with much worse field position, generally

37
by Sander :: Wed, 12/29/2010 - 12:47pm

Well that depends. Bend-don't-break defenses will usually give up a lot of yards in pursuit of a turnover and I doubt the resulting field position is much better than a team that forces a lot of 3-and-outs.

51
by andrew :: Wed, 12/29/2010 - 1:00pm

I was being sarcastic. :P

7
by Nathan :: Wed, 12/29/2010 - 12:17pm

I guess an offense that chalks up 1500 yards a game, always gets to the 1 yard line but can't punch it in is completely worthless and must be totally blown up and rebuilt. Cause any stat other than points scored is meaningless.

9
by jklps :: Wed, 12/29/2010 - 12:18pm

Plus, you should NEVER consider your opponent when looking into these things....

21
by B :: Wed, 12/29/2010 - 12:28pm

Well, they should definitely get a new field-goal kicker, if they aren't scoring any points in that situation.

24
by Ted (not verified) :: Wed, 12/29/2010 - 12:32pm

You exaggerate to make your point, but I'd go equally as far to say that that offense you describe is worthless. Racking up tons of yards is awesome, but if it doesn't translate to points, then who cares. Same as defense.....if you give up 1000 yards a game, but only 3 points, then ultimately you're probably going to win most of those games.

25
by Spielman :: Wed, 12/29/2010 - 12:34pm

Field goals exist.

27
by Aaron Schatz :: Wed, 12/29/2010 - 12:35pm

Shhhh... don't tell Clint Stitser.

185
by RickD :: Thu, 12/30/2010 - 1:41pm

FGs exist but they are not successes. Teams built on FGs don't win a lot of games.

234
by Spielman :: Fri, 12/31/2010 - 9:42am

We're talking about a theoretical offense that scores no TDs, but moves to the opponent's one yard line every time it gets the ball. I'm going to posit that an offense that attempts 18 yard field goals on every possession would win quite a few games.

26
by Nathan :: Wed, 12/29/2010 - 12:35pm

It is certainly worthless as far as winning games. However, by looking at a few statistics (goal to go offense) I'm pretty sure you find you don't need to blow up the entire offense to fix the problem. You could probably start by adding a good fullback, short yardage back and maybe a blocking tight end or two. See what I mean?

10
by Regnad Kcin (not verified) :: Wed, 12/29/2010 - 12:19pm

C'mon give the guy a break; he does have that awful speech impediment. Can't pronounce "Patriots" correctly.

Unless "patri-AUGHTS" is supposed to suggest his name should be prounounced "bill po-LYIN'"

13
by Purds :: Wed, 12/29/2010 - 12:21pm

Do you really think Polian is specifically addressing FO, and even if so, did you really expect him to read your comments?

And, to this: "Does Bill Polian understand that he is in the entertainment business? Does he understand that anything which increases interest in football is good for the National Football League?"

No, Polian doesn't, and no, he personally is not in the entertainment business. He's in the winning business. Nice guys in the NFL get fired as quickly as jerks in the NFL if they lose. Now, I would like to have a nice guy lead my team instead of having a jerk lead them, but from the coach/GM's personal perspective, the reality is they need to win, no matter the deportment.

Your quotation is exactly what links Polian and Bellichick. They're both unquestioned personal jerks, their fans love them because they produce unquestionable winners, and opposing fans love to hate them because they're jerks. Move on.

20
by Aaron Schatz :: Wed, 12/29/2010 - 12:28pm

I've spoken to Polian in the past as well as two of his sons. We had a long talk about this after his speech at the MIT conference last year. Yes, he is talking about us. Most NFL people, I have no idea if they pay any attention to FO. But I know for a fact that Polian does because I've talked to him about it.

You also have to understand that Belichick never says anything obnoxious about fans. Oh, sure, he's an asshole to the media, but he never says anything bad about the fans.

54
by PantsB (not verified) :: Wed, 12/29/2010 - 1:08pm

Belichick is also an "asshole" to the media as part of a conscious strategy. He gameplans how he's going to be an asshole to the media because it benefits his team by denying any useful information to the opposition and increasing the likelihood that his team can play the "us against the world" card when the media gets whiny about it. In reality, the worst he really ever does to the media is not help them do their job by being brusque.

On the other hand, Polian is just a prick:

"And as backup quarterback Doug Flutie rolled out with pressure on his heels, Polian muttered, 'Break his leg.'"

"While discussing the matter with a Jets employee, Polian grabbed him by the lapels and threw him against a wall."

79
by Will Allen :: Wed, 12/29/2010 - 2:11pm

Is there anything more pathetic than an old frail fart getting physical with somebody? The guy is great at what he does, but it would have been even more great if the guy he put his hands on had kneed him in the nuts, and then followed up with a sharp kick to the ribs. Nothing that threatens his life, of course, but a few weeks of agony would have been instructive for the old fool, regarding the wisdom of keeping his hands to himself.

80
by loneweasel (not verified) :: Wed, 12/29/2010 - 2:13pm

Tough guy on the internet, yo.

89
by Will Allen :: Wed, 12/29/2010 - 2:33pm

Don't know for sure what you meant by this, but on the off chance it was meant to be criticism of posturing as someone who was willing to engage in violence, read again. The point is that people who engage in violence should have good reason to do so, lest they get soundly beaten as a result of their foolishness. Being mad at the employee of another football team likely doesn't qualify as a good reason to engage in violence, thus I would have been entertained by the prospect of near-elderly man, who should surely know better at his advanced maturity, receiving a lesson regarding the costs of inadvisable violence.

Internet illiteracy, yo.

93
by loneweasel (not verified) :: Wed, 12/29/2010 - 2:39pm

it's an answer to your original question.

The only people more pathetic than violent geezers are internet tough guys who threaten said geezers through the internet. That means you.

And people who tuck their shirt on the golf course, of course.

94
by Will Allen :: Wed, 12/29/2010 - 2:44pm

Lying about what other people have written is pretty pathetic as well, and is something you engage in with frequency.

205
by Mikey Benny :: Thu, 12/30/2010 - 2:50pm

Will, ignore him. Loneweasel, I already told you: go back to your bridge.

211
by loneweasel (not verified) :: Thu, 12/30/2010 - 3:13pm

Is that where your mom is waiting?

215
by Noah of Arkadia :: Thu, 12/30/2010 - 3:34pm

Wow! Is that what we've come to?

115
by Alaska Jack :: Wed, 12/29/2010 - 4:30pm

I don't get it. Where did Will threaten Polian?

1. Guy A threatens Guy B with violence
2. Guy A gets his comeuppance.
3. Will is gratified.

Seems pretty straightforward to me. Please clarify.

- AJ

123
by Will Allen :: Wed, 12/29/2010 - 5:48pm

Don't bother. He appears to have some sort of bizarre psychological obsession which results in him gaining satisfaction from lying about what I've written. It's happened before.

186
by RickD :: Thu, 12/30/2010 - 1:43pm

+1

228
by Cid (not verified) :: Fri, 12/31/2010 - 3:19am

Saying "I wish that guy had kneed him in the nuts" isn't a threat.

A threat would be "If Bill Polian ever tried that with me, I'd knee him in the nuts!"

For a such a smart guy, you're pretty dumb.

97
by Purds :: Wed, 12/29/2010 - 2:57pm

And you KNOW this is a conscious strategy as opposed to his personality because ... ? Dont be a "In Bill We Trust" guy here -- that excuse can work for anything he does, effectively saying that because he wins, he's perfect. Come up with a better reason, or you'll hear Colts fans argue the same and sound as inane--Polian does that on purpose to make people think he's not as shrewd as he is, and then the other GMs won't question his motives in trades as much.

127
by Dave :: Wed, 12/29/2010 - 6:00pm

Ha. Trades. Those happen all the time!

130
by MJK :: Wed, 12/29/2010 - 6:20pm

The poster is perhaps exaggerating a little, or inferring a little too much, but David Halberstam's book "The Education of a Coach", about Belichick's life, does talk about how Belichick decided that the biggest mistake he made in Cleveland was how he handled the media, and so when he took over the Patriots he sat down and formally developed a strategy that he would follow that would allow him to handle the media more to his own advantage. So it's not far fetched to say that Belichick plans out his press conferences (he does), and that he tried to plan to include things that will advantage his team, or at the very least provide minimum advantage to other teams.

Of course, nothing helps with the local media more than winning a few SB's...

140
by Purds :: Wed, 12/29/2010 - 8:20pm

Fine, so BB plans out everything with the media and is perfect, as far as the earlier poster believes. That means that when the season ends for his team and he can't utter more than 5 words after a dramatic AFCC loss, he's doing that for strategy? Or, when he publicly shuns former assistant coaches who have the audacity to take HC jobs that he doesn't like, that's strategy? Don't forget, this is the guy who took a former mentor's groomed HC job for one day, then split. That's all strategy, not personality?

143
by Anonymous2 (not verified) :: Wed, 12/29/2010 - 9:20pm

"Or, when he publicly shuns former assistant coaches who have the audacity to take HC jobs that he doesn't like, that's strategy? Don't forget, this is the guy who took a former mentor's groomed HC job for one day, then split. That's all strategy, not personality?"

Neither of these have anything remotely to do with how BB handles the media.

If you think BB doesn't handle the media, than you're not paying attention.

He doesn't give away possibly important information the way that some other coaches do. That's the whole point.

And for this, people like you spend hours (apparently) doing psychological profiles on the guy.

145
by PatsFan :: Wed, 12/29/2010 - 10:09pm

Yeah. Saying publically how he was glad McDaniels got the Denver job and then when it all fall apart publically it was sad how things turned out for McDaniels and that he got a somewhat raw deal is "shunning".

As I noted in another thread, you have a very odd idea of what common English words mean.

Or perhaps you were referring to Mangini, who tried to recruit players and staff to the Jets on board the team flight home from the Broncos loss and who allegedly stole laptops. Can't imagine why BB might want to "shun" him. Guess it's only because BB is a psychopathic bastard. Nope. There's no other reason besides that for him to give Mangini the cold shoulder. None at all.

147
by Purds :: Thu, 12/30/2010 - 12:43am

Again, speculation that must be proof because it makes BB the good guy. You're right, BB's a charmer. So quitting after one day on the job is fine? And, I am sure he wouldn't have tried to take any personnel with him.

He's a guy who has cost NE two first-round draft picks because of his actions over the years. Yep, those are the traits of a upstanding guy.

Anyhow, it's good to see that the cool aid still flows liberally in NE. No one's accusing BB of being a bad coach, just a personal jerk, but as he's a good coach, you're right that all that other crap must be made up. He'd never force a player to do full-contact drills while recovering from a concussion. Not him. He'd never cheat. He'd never turn his back on a former mentor, just do what he did (look for a different situation while in one -- hmmm., sounds like what you accused Mangini of doing). Nope, none of that is true, cause BB wins, baby!

149
by Andrew Potter :: Thu, 12/30/2010 - 1:15am

Way to argue with something he didn't say.

159
by Dingle-Doodah (not verified) :: Thu, 12/30/2010 - 10:09am

"You're right, BB's a charmer. So quitting after one day on the job is fine?"

Except that he never quit because he didn't accept the HC position in the first place. There's a difference between "I quit" and "Nah, I'll pass".

162
by Spielman :: Thu, 12/30/2010 - 10:45am

There's also a gigantic difference between simply saying "Nah, I'll pass", and what Belichick actually did, which couldn't have been better calculated to embarrass the Jets organization.

242
by fillylabinga (not verified) :: Sun, 01/02/2011 - 12:06pm

I agree with the general point, but technically didn't he in fact quit? Didn't the famous note read "I resign as HC of the NYJ"?

193
by LnGrrrR (not verified) :: Thu, 12/30/2010 - 2:04pm

Randy Moss certainly excoriated the man, right?

What about the numerous players who have said how horrible he is? Oh wait, you mean it was only really Ted Johnson?

BB is not a saint, but he's not some evil mastermind either. He has his faults.

236
by JL (not verified) :: Fri, 12/31/2010 - 1:20pm

I don't think anyone said that Belichick was perfect and immune to emotional reactions, just that a lot of his brusqueness with the media was part of a deliberate strategy. So was turning down the Jets job with an unstable ownership situation and Parcells leaning over his shoulder - a classic no-win situation.
Mangini, whom Belichick had first hired as a QC coach, and had mentored his entire career, took the Jets job, against Belichick's advice, and then tried to raid his staff on the way out the door. If Belichick was a little nasty to him after that, to me that is not evidence that he is an "asshole."
From what I have seen, more of Polian's apparent anti-social behavior is a result of his being out of control and unconcerned with what anyone thinks, whereas at least a meaningful portion of BB's seems to be part of a larger plan.
That having been said, Belichick has certainly done some unsavory things and is in all likelihood not a candidate for sainthood.

139
by Purds :: Wed, 12/29/2010 - 8:17pm

I stand corrected about Polian referring to you. On the BB thing, I elaborate in a lower comment.

166
by Noah of Arkadia :: Thu, 12/30/2010 - 11:42am

Often bad publicity is good publicity. Plus, if Polian thinks you're idiots, why does he even read the site? I bet he doesn't read Glazer, does he?

14
by Jetspete :: Wed, 12/29/2010 - 12:21pm

for whatever reason i cant open the link. nonetheless, it would be tough to argue that Polian isn't one of the top 5 executives in NFL History

188
by RickD :: Thu, 12/30/2010 - 1:47pm

Polian is definitely one of the top 5 executives in NFL history. He's also a jerk. And his argument in the comment linked to is an exercise in oversimplification, followed by a pathetic backtrack based on the realization that he's overstated his case.
It is not reasonable to argue that the only statistics on defense that matter are points scored and turnovers. Yardage matters. 3rd down conversion rate matters. Sacks matter. And using all of the statistics available allows one to manage a defense a lot better than simply considering only two of them.
That's why statistics exist: because people have found them to be useful.

15
by Sander :: Wed, 12/29/2010 - 12:21pm

It's interesting that statistically, he'll take an offense that gets a turnover at their own 10-yard-line (and doesn't give up field goals) over one that gets stops at the opponent's 20-yard-line.

17
by Spielman :: Wed, 12/29/2010 - 12:21pm

"We're now talking about receivers being “targeted” as opposed to how many balls they catch. All that tells you is: a, that the ball was thrown his way – it doesn't tell you why. It doesn't tell you what kind of a ball it was – whether it was good, bad or indifferent."

Whether a batter gets a hit or not doesn't tell you why, either. It could be a bloop that fell in because the outfielder got a bad jump, or a ball that was absolutely crushed on one hop to the left fielder. Yet somehow, we find value in batting average and OBP.

84
by Will Allen :: Wed, 12/29/2010 - 2:19pm

Well, to be fair, whether Joe Mauer gets a hit or not doesn't have much to do with how well Alex Rodriguez or the Twins number 8 hitter is performing. The problem I have with even most advanced individual stats in football is that unless those numbers are matched up with a helluva lot of charting data, you really can't draw any strong conclusions from them alone, in the manner you can with baseball.

This isn't a criticism of the work done here, but merely frank recognition of how much more difficult it is to produce useful quantitative analysis for football, compared to other games.

109
by Spielman :: Wed, 12/29/2010 - 3:31pm

The point was just that a statistic doesn't have to be perfect to have value. Obviously football statistics are far more problematic, but the same principle applies. Throwing out data because it isn't perfect, or because it could theoretically be improved, generally isn't a very smart idea.

92
by RichC (not verified) :: Wed, 12/29/2010 - 2:38pm

Well, we keep track of Line Drive %, GB %, FB %, etc. So yeah, the stats do tell you what kind it was.

103
by Spielman :: Wed, 12/29/2010 - 3:23pm

Batting average and OBP don't, which is why I mentioned them specifically. The fact that more advanced statistics exist that do doesn't change that, and doesn't mean that OBP isn't going to correlate with scoring runs, just because it doesn't contain within itself a complete picture of the offensive value of a player or team.

30
by nat :: Wed, 12/29/2010 - 12:40pm

He's complaining about standalone statistics. Which is a bit of a strawman, since no one I know quotes football stats without some nod to context. What he really seems to be saying is "my stats that my team uses internally are good" while "any other stats that the fans might value are useless". So, yes, the thread title pretty much captures the gist.

He's wrong, of course. If I have a stat that says my defense is generating sacks, hurries and batted balls at a much lower rate than average (per pass play, adjusted for opponent), I don't need much context for that to be useful in identifying a general problem with my defense. I still have to diagnose why I have the problem and decide what adjustments to personnel or scheme to make. That's what game film and coaches are for.

Weeks later, I can look at that same stat to figure out if my adjustments seem to be working.

31
by The Anti-Dave (not verified) :: Wed, 12/29/2010 - 12:41pm

Bill Polian once filed a formal complaint with the league because he was afraid a mascot was causing too much disruption to his team.

Formal complaints about a man dressed as a giant plush kitty tell you about all you need to know about Bill Polian.

55
by Dave :: Wed, 12/29/2010 - 1:09pm

This is a really dumb reason to dislike the guy.

Don't get me wrong; there are plenty of legitimate ones. But every single team makes post-game complaints to the league every week. In fact, the guy most notorious for it has one of the best public reps in the media. He files a laundry list of minor and annoying (and often false) complaints at every stadium he visits.

Polian catches hell for things like this, though, because it's more notable when he does it, due to his profile, and [mostly] because of how he has been known to treat people.

I don't blame anyone, of course. He's a jerk. (My kind of jerk, but a jerk nonetheless.) But I'd rather you just called him an asshole than making character judgments based on something as meaningless and common as this.

Edit: Your username being attached to this specific post is pretty funny.

90
by The Anti-Dave (not verified) :: Wed, 12/29/2010 - 2:34pm

I just tend to find the mascot story more entertaining than the legitimate reasons.

Yes, I do like the Dave/Anti-Dave dynamic.

168
by The Anti-Anti-Dave (not verified) :: Thu, 12/30/2010 - 12:05pm

You're both wrong, you know

42
by MJK :: Wed, 12/29/2010 - 12:50pm

OK, some thoughts on this.

First, it is definitely true that there is a TON of information that coaches and teams have that we (the public) don't, starting with the actual rulebook, including the "All-22" multi-angle coaches film for all the games, exact knowledge of which players were on the field for every play, andand, probably most importantly, knowledge of what play or defense was called for a given situation.

If good statisticians, like some of the folks at FO or that go to the Sloan conference, had access to all this data, they could probably revolutionize how teams do things in the NFL. In fact, some teams probably have hired statisticians that do exactly that. So obviously, Polian's implications that publically developed statistical models are missing something because they don't have all the data is true. Could FO come up with better tools and models if they had all the data Polian has? Of course. But that does not mean that their models are useless.

Some people have already pointed out that FO's information is perhaps more useful to fans than it is to teams, and that makes sense. FO knows offenses tend to be more consistent than defenses from year to year, but doesn't know why. So fans of a team with a good offense can be confident for the next year, but that doesn't necessarily mean a team with a good offense should take that information to mean that they don't need to do anything to improve their offense in the offseason (maybe part of the reason why offenses are more consistent is that teams are good at improving offensive talent every offseason).

However, there's a bigger point that Polian is missing. Statisticians for years have been trying to figure out how to infer useful information in situations where they have incomplete data. There are ways to factoring in the "big picture" and letting inaccuracies like Polian alludes to "filter out" in the noise, in an average sense. This is where the power of statistics really lies.

Perhaps Polian is just striking out against some fans that seem to think statistics are the be-all and end-all (as I think the FF culture is encouraging). But his "I'm smarter than you all could ever by at the game, so you should just stop trying" is a little condescending. I respect Polian for building very good teams and innovating the game, but I actually think he's a pretty horrible person.

48
by PatsFan :: Wed, 12/29/2010 - 12:55pm

a TON of information that coaches and teams have that we (the public) don't, starting with the actual rulebook,

http://www.amazon.com/Official-Playing-National-Football-League/dp/16007...

(and comparing the pages Amazon makes available for viewing to my PDF of the 2009 Official Rulebook that someone got me last year off the media-only NFL site, it's pretty clear this book is the real thing and not just some digest of rules.)

78
by Raiderjoe :: Wed, 12/29/2010 - 2:09pm

Have it. It is real book wuth the rules. Not jusr digest of them

113
by Jovins :: Wed, 12/29/2010 - 4:00pm

I've always thought that offensive efficiency being highly consistent has everything to do with quarterbacking.

Bad teams don't usually magically get good quarterbacks; good quarterbacks don't magically turn bad. So, most of the time, a good quarterback will lead a good offense, and then do the same year in year out. Even more likely is that a bad quarterback leads a bad offense, and that team sucks on offense year in and year out.

However, there's no real way to isolate the effect, because a lot of new quarterbacks are rookies, and rookie quarterbacks not named Dan Marino (or Ben Roethlisberger if you're a Steelers fan) aren't top-tier quarterbacks.

136
by zlionsfan :: Wed, 12/29/2010 - 8:06pm

I think you've said pretty much what I would have said about Polian's comments.

I feel like this is the same routine that the MLB people performed, an extension of the philosophy that says "You can't possibly understand any of this if you were never part of it." We (with few exceptions) never played or coached the game, so what we do isn't meaningful, no matter what it is ... and establishment people repeat it over and over again until it's obvious they're wrong, sometimes beyond that point. (I think this is the context in which Morgan's name is mentioned, referring to his comments about sabermetrics.)

The problem is that they're not completely wrong. There are some things we can't understand from the outside, sure; on point, we can't draw solid conclusions about coverages and such because we don't know for sure what's supposed to happen. However, there are times when it's helpful to have an outsider's perspective. Sometimes being in the middle of the situation prevents you from seeing things, particularly in businesses like professional sports, where the degree to which risks are avoided is astounding at times. Sometimes a different way of looking at things is valuable. (It must be, to a certain extent, or FO wouldn't be providing data to NFL teams.)

Absolutes really make it hard to make valid points. He and Simms would have been much better off saying something like "Statistics without context can be very misleading, and in a team sport like football, it's very difficult to evaluate the performance of an individual player." (The problem with Simms saying that is that CBS regularly asks him to violate both principles, and all his on-field experience doesn't change the fact that he knows just as much about the current on-field playcalling as we do.)

44
by srsbzns (not verified) :: Wed, 12/29/2010 - 12:51pm

Obviously, Bill Polian has discovered the secret Pats bias of this site and is trying to discredit you.

110
by CoachDave :: Wed, 12/29/2010 - 3:33pm

That's a secret? Since when?

124
by SFC B (not verified) :: Wed, 12/29/2010 - 5:50pm

Do you, Dave, and Anti-Dave combine into some sort of gestalt uber-commenter? Like a Devastator Dave or Davtron?

128
by Eddo :: Wed, 12/29/2010 - 6:01pm

I'm not sure about that - if Dave and Anti-Dave combined, they'd cancel each other out; if Coach Dave was too close when that happened, he'd be destroyed, as well.

What I'm saying is, Daves, don't try to form Davtron.

57
by Trust Doesn't Rust (not verified) :: Wed, 12/29/2010 - 1:11pm

Ironically, Polian is a major stats guy and he doesn't even realize it, as he seemingly constructs his team more on type than he does on raw scouting. He knows the exact type of speedy coverage linebackers he needs, the exact type of run-stopping SS he needs, exact type of stretch RB, etc. which is why he's been able to so consistently rotate talent in and out and never miss a beat. I'd almost guarantee that Polian studies combine stats more than anybody to identify the attributes he's long-since identified as fitting his plug and play approach to personnel.

Of course, you could also look at his failure to develop any new pro bowlers since 2004 (when Bob Sanders was drafted) as evidence that he needs to reevaluate his methods.

58
by Dave :: Wed, 12/29/2010 - 1:15pm

That's kind of a gross oversimplification, don't you think?

Speedy and good in coverage isn't exactly something captured by a statistic. And suggesting that he relies on anything over scouting is exactly the type of assumption that leads him to make exaggerated statements like the one in question.

68
by Trust Doesn't Rust (not verified) :: Wed, 12/29/2010 - 1:36pm

I'd bet Polian has very specific measurables in mind for the guys he drafts and picks up in free agency, specifically on defense (and probably running back too). How else to account for the startling unanimity among the guys he brings in (think of the endless cookie-cutter corners and linebackers he's had over the years)? I'm sure he has a specific scouting measurable in mind as well, don't get me wrong, but the point is that his approach is way more similar to FO's than he'd care to admit.

Think of it like this: Imagine if FO came up with a metric for drafting linebackers in rounds 3-undrafted. This metric says that linebackers drafted late are more likely to be successful if they have sub 4.6 speed, weigh 235-240, are 5'10-6'0, and had to at least be a 2 year starter in college. It's called "The Polian Forecast."

75
by Dave :: Wed, 12/29/2010 - 2:03pm

If FO came up with that metric I'd also say it's way oversimplified and not all that impressive.

There are hundreds of LBs that meet those measurements, by the way. They can't all be plugged in and play.

82
by Trust Doesn't Rust (not verified) :: Wed, 12/29/2010 - 2:15pm

I don't think there's hundreds of undersized linebackers that fit that profile, but if you want to knock it down further throw in a college stat like passes defensed. Either way, it's very notable that Polian's linebackers, to name just one group, have almost the EXACT SAME measurables year after year. That's not a coincidence. He's not just sitting there scouting college guys who are good in coverage and can move good laterally, and it just so happens that they all turn out to have almost exactly the same height, weight, and speed.

161
by Dingle-Doodah (not verified) :: Thu, 12/30/2010 - 10:38am

.... and brittleness

170
by Noah of Arkadia :: Thu, 12/30/2010 - 12:14pm

I'm very skeptical of this. Players aren't made in factories.

191
by RickD :: Thu, 12/30/2010 - 1:53pm

Players aren't made in factories but different people have different ideas of what they are looking for. Polian tends to prefer speed at certain positions where Belichick prefers size.

114
by DZ (not verified) :: Wed, 12/29/2010 - 4:10pm

No new Pro Bowlers since 04? Uh Bethea and Addai both made Pro Bowls and were drafted after that.

156
by Bobman :: Thu, 12/30/2010 - 4:53am

If you're gonna start mucking this up with mere facts....

I'm going back above where people discuss with certainty and deep understanding what Polian and Belichick are thinking.

59
by roguerouge :: Wed, 12/29/2010 - 1:17pm

Yet another cranky old man, albeit a successful one; his opinion is just as valid as Joe Morgan's. Baseball was full of these blowhards until Alderson, Bean, Maddon and Epstein; now, it's just the game broadcasters.

Also, I wouldn't put it past Polian to be flat out lying about all of this, because advanced stats are a competitive advantage his team might want to make sure others don't get.

63
by Dave :: Wed, 12/29/2010 - 1:24pm

You're going to an extreme that is the opposite of his, and that makes you just as wrong. Joe Morgan can be emphatically proven wrong with facts. There's no single statistic that can tell you everything about a play, a series, a game, a team (etc) as a skilled pro's eyes can in football, and there never will be.

Aaron isn't pretending he can, of course, which is why Polian's little rant was insulting. But he's always going to be right that no statistic can accurately account for every variable. This is part of why football is so great.

122
by roguerouge :: Wed, 12/29/2010 - 5:42pm

"There's no single statistic that can tell you everything about a play, a series, a game, a team (etc) as a skilled pro's eyes can in football, and there never will be."

Speaking of fallacies, you can't prove a negative.

Still, I'll offer an argument by analogy. People said that about baseball too, especially when it came to things like quantifying a pitcher's stuff or measuring defensive ability. And yet, hidebound baseball, unlike football, devoted resources to PitchFx to measure break and velocity of every pitch and has several independent companies tracking the angle and placement of every batted ball. Now, people understand that you can get ever closer to quantifying two aspects of the game that were previously thought unmeasurable.

Similarly, if the cranky old men of football would devote the tiny resources of opening up all-22 film, the game would be the better for it. Opening up the raw data to the crowd for public analysis has reformed many vital industries. Of course, that would open them up to the scrutiny of informed consumers with the evidence to prove them wrong. It's in Polian's interest to make sure that he retains the competitive advantage he does by denigrating new advantages and hoping to prevent them from transforming the industry during his twilight years. I think it's rather sad.

61
by roguerouge :: Wed, 12/29/2010 - 1:20pm

An educated consumer is a dangerous consumer. Polian may be part of the people denying fans All-22 access for so many years, which allows him to be "right" about how ill-informed they are. If he actually cared about this issue and the game, rather than just winning, he'd use this kind of interview to pressure the league for more and better access to game footage.

65
by Dave :: Wed, 12/29/2010 - 1:25pm

I agree with this. It's a lot harder for the average fan to get truly educated. Partly because the game is more complicated, partly because so many are content to just rely on fantasy stats, but also partly for the reason you give.

66
by tbwhite :: Wed, 12/29/2010 - 1:36pm

When I read it I assumed that when he said standalone he meant individual statistics. He specifically cites targeted passes as a misleading stat. That's true, the scorer can't always know who the QB really intended the pass for. I think what he is missing is that over time with a large enough sample size those things even out, which is why even though baseball stats don't really distinguish between a hard liner and a Texas Leaguer they are still meaningful.

In general, if his point is that individual stats aren't useful in football, then that's a stronger point. On a team level DVOA can take into account context, but at the individual level it's harder to make the argument. I don't think he's right, but I do concede that it's tougher to wrap your head around individual stats for say offensive lineman vs some sort of team level stat of offensive prowess.

95
by RichC (not verified) :: Wed, 12/29/2010 - 2:53pm

" I think what he is missing is that over time with a large enough sample size those things even out, which is why even though baseball stats don't really distinguish between a hard liner and a Texas Leaguer they are still meaningful.
"

There are two problmes with this:

1) In most cases, no, they don't even out because we never get large enough sample sizes.

2) Baseball stats actually do distinguish those things. Look up SIERRA.

126
by SFC B (not verified) :: Wed, 12/29/2010 - 5:58pm

I think it is safe to say that a SIERRA-like stat is still a long way off for football. Not to denigrate the work of FO et al, but football is still working at a AVG/HR/RBI level. Until the NFL decides to give their stat-minded fans some of what they want football stats are just working from a much shallower pool of information.

72
by Staubach12 :: Wed, 12/29/2010 - 1:56pm

I think this Polian quotation perfectly sums up his take on stats:

"Speak English, please. I have a very hard time understanding all the mathematics of it. I have a lot of work to do and a lot of decisions to make, and if you can't make it understandable for me, I have no use for it."

  • http://sports.espn.go.com/boston/news/story?id=4972331
  • 172
    by Noah of Arkadia :: Thu, 12/30/2010 - 12:24pm

    That makes more sense. But then he would be the "idiot" who doesn't understand stats, and he can't have that.

    73
    by Staubach12 :: Wed, 12/29/2010 - 1:58pm

    (double post)

    76
    by Basilicus :: Wed, 12/29/2010 - 2:07pm

    I really want to screen the movie Pi for Bill Polian. Watch his head explode.

    85
    by Nathan :: Wed, 12/29/2010 - 2:22pm

    INT. BATHROOM

    Polian scrubs the DVOA database, then places a drill to his scalp... the motor WHIRS.

    CUT TO:

    EXT. PLAYGROUND - DAY

    Polian watches the branches sway. Just then, Jenna surprises him.

    JENNA
    Can we do one, Bill,
    can we?

    Polian shrugs, not able to say no.

    JENNA
    What is Pierre Garcon's DVOA?

    Polian thinks. Jenna types on her calculator.

    JENNA
    I got it! I got it! What's
    the answer?

    POLIAN
    I don't know. I really don't
    know. What is it, Jenna?

    JENNA
    -35.6%

    POLIAN
    Oh.

    FADE OUT.

    81
    by Raiderjoe :: Wed, 12/29/2010 - 2:13pm

    Stats tgat matter---
    Pints for
    Points agaunt
    Wins
    Losses
    Ties

    Stars that sort of matter
    turnovers
    Yards per pass attempt
    Opp yards per pass attempt

    Stats that dont mayter too much-
    Others

    83
    by drobviousso :: Wed, 12/29/2010 - 2:18pm

    I would argue that, given this article, running success rate (and defensive running success rate) "sort of matter" as well.
    http://www.advancednflstats.com/2010/10/how-coaches-think-run-success-ra...

    What do you think?

    117
    by Alaska Jack :: Wed, 12/29/2010 - 4:40pm

    I think points per possession (for and against) would be more meaningful than simple points for/against.

    Especially if it was somehow weighted against field position. Like points divided by yards, or something like that.

    On the other hand, I am emphatically not good with numbers.

    - AJ

    203
    by LnGrrrR (not verified) :: Thu, 12/30/2010 - 2:48pm

    Sounds like the "Scorability" stat on ColdHardFootballFacts.com

    230
    by ygold (not verified) :: Fri, 12/31/2010 - 3:37am

    Stats that definitely don't matter-
    Spelling errors. I love it.

    88
    by speedegg :: Wed, 12/29/2010 - 2:29pm

    Sometimes I wonder if Polian is scared that FO is on to his system and it will be made public...

    118
    by jmaron :: Wed, 12/29/2010 - 4:43pm

    You seem a tad thin skinned Mr. Schatz.

    It seems to me that Mr. Polian thinks you guys aren't nearly as smart as he is. That is quite different from saying he thinks you are idiots.

    119
    by Nathan :: Wed, 12/29/2010 - 4:46pm

    Standalone statistics such as exist in baseball – sabermetrics, if you will – have no validity in the NFL. Zero.

    He's saying they're dedicating their professional lives and a great deal of recreational time to an endeavor that has absolutely zero validity.

    Pretty sure that qualifies.

    121
    by jmaron :: Wed, 12/29/2010 - 5:38pm

    Suggesting that what one is doing is pointless is not calling someone an idiot. I happen to think that many very smart people do and believe very pointless things.

    Technical Analysts come to mind.

    157
    by Bobman :: Thu, 12/30/2010 - 4:58am

    Technical analysis... God I haven't heard that term in years. You mean I can't tell how a company will perform based on it's historical stock price chart? Egad!

    163
    by jmaron :: Thu, 12/30/2010 - 11:18am

    I like what Buffet said about it -

    "I knew it technical analysis didn't work when I turned the chart upside down and got the same answer"

    But back to the topic at hand - the more I think about it the more I think Polian comes off as arrogant and a little scared. Of course some of these stats FO have some meaning - certainly the contextual challenges are there but catch rate does tell you a little and it could be easily refined so that charters could come up with something the equivalent of errors. Catch rate for catchable balls.

    I apologize to Aaron - I think Polian was rude and arrogant and more importantly - wrong.

    177
    by Tom Gower :: Thu, 12/30/2010 - 12:57pm

    Just as a reminder, the Colts only won the Super Bowl in 2006 because the moon was in the AFC house.

    208
    by Sander :: Thu, 12/30/2010 - 3:07pm

    Reading that back it's amazing how many people thought there was some merit to the case Football Professor was making.

    129
    by Bigpurpleguy. (not verified) :: Wed, 12/29/2010 - 6:11pm

    "Standalone statistics such as exist in baseball – sabermetrics, if you will – have no validity in the NFL. Zero."

    Is he actually saying he doesn't use ANY statistics to make decisions?

    This in itself is, um, insane.

    If he thinks the QB isn't completing enough passes, whether or not he looks at the actual percentage, he is using his own brain's perception of that percentage to make a decision, isn't he?

    Or am I misunderstanding this guy?

    137
    by greybeard :: Wed, 12/29/2010 - 8:07pm

    He is saying without the context for each situation, such as the play that was called, the front that was used, the role of the play that was called in overall game stategy, etc, the stats for individual players are meaningless. He gives an example of a defensive player with just two tacklesin a game but each one cauisng the drives to end as not having a great day wrt stats but the reality being he opposite. He also does not seem to be in the opinion of "these things work themselves out in the long run".

    141
    by Michael K (not verified) :: Wed, 12/29/2010 - 8:59pm

    And yet, based on the defeats stat, that guy would be accurately valued. His success rate would be phenomenal. And if the criticism Polian were making here was more along the lines of the example you created - you say he says it, but I don't see that anywhere in the linked page - that would be a perfectly valid critique. But that's not what he's saying at all; he specifically saying that attempts to figure out the value of that specific player are worthless, and that's absurd and absolutely similar to the crap that way too many old baseball guys spit out about never knowing the value of pitchers.

    148
    by greybeard :: Thu, 12/30/2010 - 1:02am

    Here is what he says (the real link should be: http://www.colts.com/sub.cfm?page=article7&news_id=05db7667-bd19-43a8-8a...);
    "Q: The statistics showed Colts defensive end Dwight Freeney with perhaps two tackles – another example of his statistics don't always tell the story . . .
    A: I think it was (CBS broadcaster and former New York Giants quarterback) Phil Simms on a broadcast I heard who said, 'Statistics are virtually meaningless in football unless you can match them up with videotape and say this is what this statistic specifically means against this defense in this situation against this offense in this situation. Standalone statistics such as exist in baseball – sabermetrics, if you will – have no validity in the NFL. Zero. I was glad to hear Phil say it publicly. It's not said enough. There is a growing – albeit small, but noisy – cadre of people who try to convince you that standalone statistics are really meaningful. They're not. As (Colts linebackers coach) Mike Murphy is fond of saying, 'There are two statistics that count on defense: points surrendered and turnovers.' That's all that counts. Everything else is meaningless. There are some we use where you do match it up with a front or a coverage. We do some pretty sophisticated analysis of it, but you have to match it with fronts and coverages. You have to know what you're looking at. You have to understand why a particular statistic has validity. I'll give you an example. We're now talking about receivers being “targeted” as opposed to how many balls they catch. All that tells you is: a, that the ball was thrown his way – it doesn't tell you why. It doesn't tell you what kind of a ball it was – whether it was good, bad or indifferent. It tells you whether he caught it. It creates a totally meaningless picture. I guess fantasy people look at it. As I've said many times, I have no idea what that's about. I'm sure it's meaningful to them, but in terms of winning games, most of them are not meaningful. When you talk about (defensive tackle) Fili Moala's contribution, I can think of two huge tackles for losses, which created second-and-double digits, which flipped the field in both cases. Dwight Freeney was a force all day. Two people were on him all day, and he was a force all day. He was coming hard, causing disruption. He was forcing the quarterback to move and forcing the quarterback to get rid of the ball. The number means absolutely zero. When you look at the tape, you see Dwight Freeney played a heck of a game. Fili Moala might have played his best game as a Colt. (Defensive tackle Antonio) “Mookie” Johnson certainly did. I can't remember – and there may have been a time – that we've had a back like (Darren) McFadden come running up there on a belly play, or on a lead draw and just be stopped cold. He was just stopped like he was running into a wall. Well, he was running into Fili and Mookie in a lot of cases. I don't know if they got credit for a tackle, but the bottom line was the defense did a terrific, terrific job. I don't know that they're capable of playing a whole lot better than they played Sunday. And I'm not sure that any defense in the NFL played any better than we played Sunday.

    152
    by Michael K (not verified) :: Thu, 12/30/2010 - 1:31am

    Ok, that's considerably more context, but I still think he's got the wrong of it. Moala had two big tackles, which should rightfully be recorded, but what else did he do? The problem here is that Polian is essentially using anecdotal evidence and indulging in the human brain's propensity to magnify some things in memory while downplaying others, and then he's using that "evidence" to disparage actual statistical evidence that looks at every play of the game and potentially says, "Moala made a few excellent plays, but was also frequently defeated," or "Moala made a few excellent plays and they are symptomatic of how he played all game."

    As for Freeney's contributions being undervalued...that's ridiculous on Polian's part. Hurries are among the most common statistics you'll see; they're all over TV broadcasts. Actually, it could be argued that the evolution of DL measurements was an early stat advancement; recognizing that merely measuring sacks wasn't indicative of how effective a defensive lineman was, statisticians created new categories to measure them.

    199
    by Purds :: Thu, 12/30/2010 - 2:32pm

    Wait, in a game where one play can make all the difference (think tuck rule, think Eli Manning ridiculous escape and then helmet catch, think Harper's pick-6 on Peyton in last year's SB, think Big Ben's end zone pass), why aren't players who can make big plays less valuable that stat machine players?

    The use of "big play" events is magnified when discussing the Colts defense (you know, the group Polian creates with personnel choices), because as I have said many times on FO and people seem to forget -- the Colts don't look to be traditional stoppers on defense. They play a bend-don't-break deep safety game, and they rely on a big play or two to get the offense in trouble. In good times, that means getting a sack or better a fumble-sack. Against a team that can and will be methodical (see Patriots and Brady, see NO and Brees) without making a mistake, the Colts D is in huge trouble. But, against an offense that must strike quickly because they aren't consistent, or are just plain not very good (see Bal of year's past, NYJ), the Colts defense works precisely as Polian sees the world.

    Thus, Polian tries to build a team of anecdotes. Guys who can make one great play even if they make 5 poor plays as long as the sum total is giving up 20 yards then causing a turn over or punt.

    Now, I am not saying this is the BEST way to create a defense, but it is clearly Polian's way, and he can get a lot of undersized players for cheap because most NFL teams can't use those guys in their standard defenses. Remember how ridiculed the Colts and Polian were at first for taking Freeney in the first round? "He's too small!" Yet, Freeney does usually command a double team. Look at last year's SB -- the Colts defense was fine in the first half while Freeney was playing, but when the second half rolled around and Dwight's pain killers wore off, they were awful.

    Finally, with your second paragraph: hurries are a nice stat, but is there no value to causing an offense to need to use two players to block you on a high percentage of downs? Where is that statistic on FO or any other site? To really eveluate De's, Polian is suggesting, you'd need to create stats that accumulate all sorts of things FO can't do -- something like .2 points per double team taken up, then 1 for each sack, .5 for a hurry, and then multiply the sack or hurry by some number (1.5?) for each time tha lineman can do that while double teamed. Where are the stat packages that can do that? And if they can't, how can the stats accurately measure a player's value?

    And then, you have to set up stats systems for each type of defense in the league, because chewing up double teams every play might not mean squat to a different type of defense.

    Whew! I'm tired.

    201
    by tuluse :: Thu, 12/30/2010 - 2:44pm

    Thus, Polian tries to build a team of anecdotes. Guys who can make one great play even if they make 5 poor plays as long as the sum total is giving up 20 yards then causing a turn over or punt.

    Really? Because that's not my view of his teams at all. If that was what he wanted, he would have Mike Vick and Devin Hester, and Reggie Bush on his teams. Instead he shies away from flashing players who can boom from time to time and goes after consistently good players like Reggie Wayne.

    As for defense they play Tampa-2, which is the most boring consistent system I know of in the NFL. It's not a big blitz scheme like Rex Ryan's which is designed to create big plays.

    As for your last point, FO doesn't really try to evaluate DEs on an individual basis based on stats alone, so I guess they agree with Polian to some degree. Also, forcing a double team, but not being able to defeat it, is only useful if the team has players who can take advantage. See: Harris Tommie 2009.

    206
    by Purds :: Thu, 12/30/2010 - 3:00pm

    Really? Because that's not my view of his teams at all. If that was what he wanted, he would have Mike Vick and Devin Hester, and Reggie Bush on his teams. Instead he shies away from flashing players who can boom from time to time and goes after consistently good players like Reggie Wayne.

    Perhaps I wasn't clear enough. Polian looks for anecdote guys on defense. And yes, a guy who demands a double team works really well when you have one at each DE position (Mathis anyone?) There is a reason these two guys are near the NFL career lead in forced fumbles.

    You noted all offensive players as anecdote guys. The Colts offense is a much different story.

    210
    by Sander :: Thu, 12/30/2010 - 3:11pm

    To be fair though, the Tampa 2 does emphasize turnovers over yardage. The idea is that if an opponent has to take 4-yard checkdowns to get down the field, he'll make mistakes somewhere in the drive. And if the opponent isn't satisfied with 4-yard checkdowns, they'll take bigger risks and the Tampa 2 is designed to punish that.

    131
    by bigpurpleguy (not verified) :: Wed, 12/29/2010 - 6:21pm

    Did anyone click on the actual article?

    Polian is quoting Phil Simms!

    Oh that is so classic. Polian is messing with FO for sure on that one.

    133
    by Staubach12 :: Wed, 12/29/2010 - 7:26pm

    Only the first sentence is a quotation of Phil Simms, and Polian endorses the statement wholesale? The vast majority of the quotation was Polian himself (he also quotes one of his coaches).

    134
    by bigpurpleguy (not verified) :: Wed, 12/29/2010 - 7:31pm

    The essence of the insult, such as it is, is in the Simmms quote:

    I think it was (CBS broadcaster and former New York Giants quarterback) Phil Simms on a broadcast I heard who said, 'Statistics are virtually meaningless in football unless you can match them up with videotape and say this is what this statistic specifically means against this defense in this situation against this offense in this situation. Standalone statistics such as exist in baseball – sabermetrics, if you will – have no validity in the NFL. Zero. I was glad to hear Phil say it publicly.

    There's no closing of the quotation, but it sure looks like it should be after "Zero."

    135
    by bigpurpleguy (not verified) :: Wed, 12/29/2010 - 7:43pm

    OK, looking at it now it seems this Dulevy character is trying to quote Polian only, and he would know better than me what were Simms' words and what were Polian's.

    BUT

    To have Polian quote Simms as an inspiration for dissing FO et al. is still pretty funny.

    144
    by Mr Shush :: Wed, 12/29/2010 - 9:24pm

    He's quoting Phil Simms in response to a leading question about Freeney having two tackles and no sacks but a big impact on the game. He's talking about conventional individual player statistics, and specifically those for wide receivers and defensive linemen. Which really are pretty damn useless. Although possibly neither of them are as useless as wide receiver DYAR, so if he had wanted to have a go at FO, as opposed to just say nice things about his own players in response to a softball question, he could have picked worse targets.

    Please note, I'm all in favour of advanced statistics, and think FO have some pretty good ones and are working well towards more. But getting hacked off about this comment from Polian is ludicrous.

    150
    by bigpurpleguy (not verified) :: Thu, 12/30/2010 - 1:22am

    Yes, our friend should not get so bent out of shape about this.

    But it is far worse for someone in Polian's position to talk out of his ass like this. Where is FO on individual D lineman? The stat doesn't exist, and the reason probably is because they have no decent metric for it, so they leave it out.

    Polian probably has more of an issue with Pro Football Focus, which does grade individual lineman. Funny thing about that is I bet what they do is far closer to a conventional coaching grade than what FO does.

    138
    by Jagwire (not verified) :: Wed, 12/29/2010 - 8:12pm

    How often does a fan get to be on an NFL rule committee, that changes rules to favor the team that the GM is employed by?

    142
    by Michael K (not verified) :: Wed, 12/29/2010 - 9:12pm

    I find this kind of funny, actually. Look at this:

    We're now talking about receivers being “targeted” as opposed to how many balls they catch. All that tells you is: a, that the ball was thrown his way – it doesn't tell you why. It doesn't tell you what kind of a ball it was – whether it was good, bad or indifferent. It tells you whether he caught it.

    He fundamentally doesn't understand the targets stat. It doesn't tell you whether he caught the ball, it tells you how often he was thrown the ball. He's being quite honest when he says he has no idea what that's about. Honestly, I think the guy probably has some stat wonks he pays to condense stat info into something he can understand ("Peyton's throwing to Wayne way more than anyone else, and when he throws to Collie and Garcon, they don't catch most of the passes.") and uses those conclusions to aid his gut evaluations of players that have (by and large) worked out for him over the years.

    Bill Polian's a great exec, even if he's an asshole, but in regards to advanced statistical analysis, he's just another old man who feels vaguely threatened by younger people talking about things he doesn't understand and doesn't care to understand. There's no point in getting angry at him for it, anymore than there's a point in getting angry at your grandparents for using IE6 and still trying to surf to the AOLs. He'll rage in his older years and then eventually he'll be put out to pasture and dollars to donuts the next guy in line will end up being a stat guy. Football's about winning within the constraints and advanced stats can help find the unused space within those constraints. The progress will march on and the angry old men will get left (hopefully respectfully and kindly, but still left) behind.

    151
    by bigpurpleguy (not verified) :: Thu, 12/30/2010 - 1:25am

    Exactly.

    Bottom line is we all use stats, whether they be "hunches" based on our personal experience or "advanced metrics."

    And if we were as successful in our field as Polian is, we'd be pretty biased toward the kind of stats we use too.

    153
    by Michael K (not verified) :: Thu, 12/30/2010 - 1:33am

    And if we were as successful in our field as Polian is, we'd be pretty biased toward the kind of stats we use too.

    Oh, absolutely. I'd kill to have Polian GMing the Bears, even if hearing stuff like this would make me grind my teeth.

    154
    by bigpurpleguy (not verified) :: Thu, 12/30/2010 - 1:45am

    Well, the guys with the phalanx were pretty good at their game until the arrow or whatever came along.

    155
    by Mr Shush :: Thu, 12/30/2010 - 4:05am

    Actually I think they were more just undermined by infighting and a series of bad hires after their Hall of Fame head coach died young. The scheme at least arguably still had merit, but the execution was shot, and they came up against a hungrier, younger team which still had the discipline to do the little things right.

    Kinda like the Aussie cricket team.

    Now the armoured knight, on the other hand, ran into the longbow and had a really bad time, but the yeoman archer system never took off around the league because it was too hard to install.

    167
    by Will Allen :: Thu, 12/30/2010 - 11:45am

    Then the league got completely ruined by the Mountain West offense, otherwise known the Alamogordo mushroom spread. Put real limits on the game.

    223
    by Bowl Game Anomaly :: Fri, 12/31/2010 - 12:27am

    The archer system is really overrated. There were a few high-profile examples of it trumping the knights (and everyone forgets to take the weather into account), but if you look past those few you would see that the knights were fundamentally sound and dominated offenses in the European league. They didn't get outschemed until the rise of the pikemen centuries later.

    233
    by Mr Shush :: Fri, 12/31/2010 - 9:02am

    Meh. The European League was kind of a backwater at that point anyway. Those all power no speed knight offenses took a pasting every time they came up against Eastern League sides with a properly implemented missile cavalry system. Look at their results against Coach Subutei's classic Golden Horde teams.

    235
    by Bowl Game Anomaly :: Fri, 12/31/2010 - 12:19pm

    I'm not going to deny that the Euro league was a backwater. We all know that any performances in that era have to be taken with a grain of salt because the teams were only drawing from a small segment of the population, not to mention the lack of modern nutrition and medical techniques. But I think the research will show that those Golden Horde teams were never tested against the top Euro league teams and that they feasted on easy schedules.

    237
    by Will Allen :: Fri, 12/31/2010 - 5:41pm

    I just love the marketing from that era; what says "We are going to play smash-mouth" more than a pyramid of decapitated heads!

    240
    by Mr Shush :: Sun, 01/02/2011 - 10:09am

    Yeah, but the Horde didn't just beat Poland and Hungary, they absolutely dominated them, and once you factor in the two-step connections via Turkey, it's pretty hard to argue they weren't the pre-eminent team of the era.

    And don't start with the whole Plaguegate thing. They were caught, they were docked a first round pick, half the teams in the league were doing it anyway, and it's really not clear that catapaulting corpses into cities was technically against the rules in the first place.

    174
    by Noah of Arkadia :: Thu, 12/30/2010 - 12:47pm

    Nice

    160
    by Jay Jesse (not verified) :: Thu, 12/30/2010 - 10:38am

    The problem is that for as much as you pay lip-service to interpreting the statistics in context, you don't do it yourself. You say "ok remember DVOA doesn't explain everything" and 2 sentences later you're saying "the best team in the NFL is x with a DVOA of y." If you want to equate "highest DVOA" with "best team" then don't get upset if somebody calls you blind number crunchers.

    I think the analysis you do is great but you put too much weight in it, and that's what Polian is saying.

    164
    by Colts1919 (not verified) :: Thu, 12/30/2010 - 11:35am

    Lies - Damn lies - and statistics.

    Fine for baseball...a gradual rounding out of averages over a long season....hitting/pitching is a one on one matchup.

    How do you incorporate blocking into a running backs success or failure....when you would need EVERY block (or lack of) to be accounted for.

    Stats in football are essentially mental masturbation.
    Also - did BP really say exactly what you QUOTE in your "article"?

    Either way - he is correct.

    165
    by Colts1919 (not verified) :: Thu, 12/30/2010 - 11:37am

    PS - get over it. There is good reason why BP is the greatest GM of all time....and you have never touched a football field.

    175
    by Noah of Arkadia :: Thu, 12/30/2010 - 12:50pm

    You wanna know why this is funny? Polian never played the game! I guess Emmitt Smith should use your argument to get the Colts GM job

    178
    by tuluse :: Thu, 12/30/2010 - 12:58pm

    Hey now, that's not Jim Finks.

    209
    by Mikey Benny :: Thu, 12/30/2010 - 3:08pm

    FO Staff, sorry to be a broken record: please force people to register to post to the web site.

    It would make it way easier to send certain people back to the bridges they came from.

    218
    by Nathan :: Thu, 12/30/2010 - 4:21pm

    i kind of like the (not verified)... it's like a red badge of pay no attention

    169
    by Steve F. (not verified) :: Thu, 12/30/2010 - 12:05pm

    "We consult for NFL teams. And a number of other teams use our publicly available data. The Colts just aren't one of them."

    I believe this comment just proves the point Bill Polian is making. Look at the insane success his Colts organization is having over the last decade and then tell us which teams are using your data. Let me guess- the Bills, Lions, Texans, 49ers, Browns and Bengals just can't get enough of your data.

    This is just a case of ego. You don't like that Polian says this stuff may be useful to fantacy players or others but not to his organization. I wonder when people will ever stop second guessing this man? If the Indy media would have had their way Ryan Leif would have been Indy's starting QB. Also, how many other organizations could have Half, thats right, HALF of their opening day starting roster placed on injured reserve and still be in the position Polian's Colts are in today?

    Your stuff is great. . . for my fantacy team. But that's FANTACY, not REAL football. You stick to what you know (which you do well, btw) and let Mr. Polian stick to what he knows.

    251
    by Anonymous1 (not verified) :: Sat, 01/08/2011 - 5:49pm

    Is this a troll? You have both Ryan Leaf and fantasy spelled incorrectly...

    173
    by PatsFan :: Thu, 12/30/2010 - 12:38pm

    Hey Bill -- it's good to see you were willing to talk to some of the younger guys in the Colts' front office to find out how run some internet sock puppetry.

    176
    by Noah of Arkadia :: Thu, 12/30/2010 - 12:51pm

    lol

    189
    by shelterdog :: Thu, 12/30/2010 - 1:49pm

    I don't want to agree with Bill Polian on anything but isn't the point something like "statistical tools don't really work in football because the game is way too complex"? There are so many important variables on a given play that you just can't incorporate the variables and create a useful model-you run into sample size issues very quickly and you're unable to figure out the cause.

    To put it concretely, can anyone tell me whether BenJarvus Green-Ellis is a good running back or is it because the Pats have a good line, a good passing attack, play against sub packages all the time because of the passing threats they have, and call creative plays at the right time for him? Is BJGE actually better than Mike Hart (or Joseph Addai)or is it situational? I bet Bill Polian can anwer that more easily and directly than any statistical model that's been developed.

    192
    by RickD :: Thu, 12/30/2010 - 2:03pm

    BJGE is definitely a better running back than Laurence Maroney. You can see that in the statistics or you can see that on the field. And it's not because the playcalling or blocking suddenly improved after Maroney was traded.
    I have no idea if Maroney is better than Mike Hart. Indeed, I have no idea if Maroney is better than anybody.

    238
    by Daniel :: Sat, 01/01/2011 - 2:21am

    Except you can't see it in the statistics - look at Maroney's 2007 season. Even up to this year, I've seen this trotted out as evidence of his excellence (usually accompanied by some sort of fatuous comment about how his detractors were ignoring the FACTS! ... but I digress.) In reality, as I think any experienced observer could see, he was a mediocre (at best) player being made to look much better by exogenous factors that FO's stats are unable to account for. This was a classic case of needing to look beyond statistics - even FO statistics.

    220
    by Jerry :: Thu, 12/30/2010 - 5:48pm

    "The best is the enemy of the better."

    It's easy to look up which running back has gained the most yards, but we'd all agree that he's not necessarily better than other backs who have fewer attempts or play behind worse lines or .... FO (and others) are trying to come up with metrics that will help us make those determinations. We haven't seen a magic number that definitively describes a player, and we're unlikely to, but that doesn't mean that these attempts aren't useful. These statistics may not be as useful to a GM trying to figure out what players will best fit his team's system, which is why the GM pays coaches and scouts and other football people. They do help some of us understand the game a little better, and I appreciate that.

    239
    by MJK :: Sun, 01/02/2011 - 1:08am

    Scientists try all the time to fit overly simplistic models to a universe that is actually way too complex to be described by them. Some of the most famous are Newton's laws of motion...it turns out, that motion in space-time is way more complex than that, as Special Relativity shows us. But wait...Special Relativity is too simplistic; reality is way more complex; good thing Einstein developed General Relatively. But General Relativity obviously doesn't work...just look at quantum mechanics. Except that even quantum mechanics aren't capturing the complexity of reality...

    I guess the scientific method doesn't really work in understanding nature or physics, because the universe is too complex. There are so many important variables that affect things that you just can't incorporate them all and create a model that answers all the questions. So I guess that means that we shouldn't try.

    I realize I'm oversimplifying your point a little...you're claiming that sample size issues make football analysis impossible. I agree they make it difficult, but that doesn't mean that there isn't anything we can learn from statistical analysis of football.

    195
    by RobC (not verified) :: Thu, 12/30/2010 - 2:23pm

    Polian has a God complex. He is clearly a terrific talent evaluator. He is also a very poor PR type of person. This was clear last season when he ordered the Colts to quit trying against the Jets and give up any chance at an undefeated season in the name of keeping players healthy. That reasoning is logical on it's face until you remember that he tried to justify pulling the starters by saying wins in a decade meant something but an undefeated season did not, then the following week in Buffalo played Manning, Wayne and Dallas Clark for a full quarter in a blizzard just so Wayne and Clark could get 100 catches each. So risking injury in a blizzard to get 100 catches in a season is okay but trying to go undefeated in a season means nothing?

    I'm glad to have Polian as the GM for my team. I just wish he would keep his mouth shut and quit sticking his foot in it like this.

    204
    by Purds :: Thu, 12/30/2010 - 2:49pm

    RobC:

    There is clearly a difference between limiting exposure to injuries, and eliminating injuries altogether. As a fan, I would have loved to have seen the Colts try for 19-0, but then in the SB, I wished that the Colts had taken Freeney out at the end of the NYJ AFCC game so that he wouldn't have gotten hurt and then been ineffective for the second half of the SB.

    200
    by Mikey Benny :: Thu, 12/30/2010 - 2:35pm

    It seems people are looking to FO as if its theories and statistical analysis were purported to be carved in stone. Get a grip and realize:

    1. They are tools. Imperfect ones at that.
    2. They are works continually in progress.
    3. They try to shed light on what is happening on the field. Imperfectly.

    I'll bet you Aaron Schatz wouldn't claim any differently.

    Aaron, I'd recommend you make registration mandatory. The trolls have been coming out in full force lately.

    225
    by neoauspex (not verified) :: Fri, 12/31/2010 - 12:43am

    Polian thinks the mainstream media are idiots, he doesn't know that FBO exists. More specifically, he just thinks the statistics that mainstream media emphasizes should not be emphasized. And that's pretty much why FBO exists: because those stats shouldn't be emphasized. FBO should join Bill Polian in thinking that other people are idiots. And you should change your link to the full article on colts.com so people can read context of the discussion.

    229
    by Justin Zeth :: Fri, 12/31/2010 - 3:30am

    Nobody involved with running a football team should be taken seriously about anything until they quit asserting that punting the ball on 4th-and-4 from inside your opponent's 45 is anything other than suicidally insane.

    231
    by Bright Blue Shorts :: Fri, 12/31/2010 - 6:44am

    The general premise behind Polian's comment seems valid ... in a team sport, you can't breakdown individual stats ... e.g. a QB is dependent on the quality of the offensive line giving him protection, the potential of the running game to take defenders out of the box, and the hands of his receivers ...

    And I guess his point is that if you send him to the combine and he watches two WRs make 40 catches out 50 balls thrown to him; he can see the quality of the pass to judge whether the guy should have been able to catch them, or whether he just has awful hands. He can make a decision between the two guys that allows for these things.

    However what the stats do tell you is quantifiable information. The human mind is easily persuaded by first impressions and "confirmation bias".
    - If Player A drops a high percentage of balls early on, the tendency is to look for reasons why he's not a good receiver. You might draft him as a lower round pick and put him on the development roster.
    - If Player B drops a high percentage of the latter balls you put it down to tiredness or lack of concentration and decide that with a bit of conditioning and good coaching you'll be able to get him catching all 50. You draft him high and make him a third string receiver.

    While quantifiably both guys had the same numbers because of the first impressions you did different things with them.