Jump to content


How would moneyball work in Football?

Bill Polian Analytics

  • Please log in to reply
75 replies to this topic

#41 C.Biscuit97

C.Biscuit97

    Ladies' Man

  • Members
  • PipPipPipPipPipPip
  • 10,721 posts

Posted 03 January 2013 - 10:05 AM

It's something nerds invented so they can hang around jocks to get their leftover groupies.

#42 26CornerBlitz

26CornerBlitz

    Read The FN Manuel

  • Members
  • PipPipPipPipPipPip
  • 8,738 posts

Posted 03 January 2013 - 10:06 AM

View Postbowery4, on 03 January 2013 - 10:03 AM, said:

I mean there is this link, so I get the jist of it... but my question is how new and effeftive are they?
http://www.wgrz.com/...tball-Analytics
I haven't seen any discussion about it on here, just wondering what you all think.

Sorry just saw the money ball thread :doh: this can be deleted mods

Check out this site:

http://www.footballoutsiders.com/

#43 since79

since79

    Probation

  • Members
  • Pip
  • 5 posts

Posted 03 January 2013 - 12:23 PM

First post sorry for the length but I wanted to jump in.

Buddy Nix’s response is exactly what Billy Bean faced when he first introduced money ball.  Scouts would go by feel or he looks like a ball player.  What his approach did was identify players who don’t look the typical part who had potential because they would be overlooked.  How he applied it was to be economical because he would not have to pay the same price for a player overlook and often times was statistically better than going by feel.  In baseball the metrics are easy to apply because ever pitch, hit ,fielding play, at bat are charted routinely.  In football it would be harder because there are more intricacies in each play but trends most certainly could be developed and understood.

I agree you cannot pick a player based on these analytics alone.  If you take the Patriots as you example, they have been using these metrics to help for several years.  Does it seem year after year they produce better than other teams?  Granted on the offensive side of the ball Brady can make anyone look good.   San Franciso has adopted analytics to assist and they have turned their franchise around as well.  Certainly Harbaugh helps so we cannot give the analytic or him all the credit.

What is interesting to me, the most is how we can use this in ferreting out our next quarterback, both by the draft and FA.  What is best about Brady, Maning etc is their ability to perform under pressure.  We need to find a quarterback who gets yards when you need them most with minimal mistakes.  Not in games trailing by large margins, or blowouts.  It’s the close games that separate the best.

Before the Brandon promotion I was thinking we need to analyze the quarterbacks in some way to determine who has these qualities. I was perusing the net and came across this article and it seems interesting to me.  
Predicting the Unpredictable…Projecting a College QB to the NFL with a Mathematical Formula   Vers 3.0

NFL Draft "Moneyball" for College Quarterbacks


http://www.fantasyfo...0_2-17-2011.htm

We all want change,  we want forward thinking and are we going to rely on Buddy, Whaley’s or the new coaches gut to pick our next Quarterback.  We cannot afford to miss again.  

After Ferguson we had names like Dufek, Ferragamo, Marangi before Kelly came along.
We have had a similar unimpressive list since Flute and Bledsoe.  

Think outside the box and realize this may be a good thing to be analytical.

#44 RealityCheck

RealityCheck

    Veteran

  • Members
  • PipPipPipPipPip
  • 2,004 posts

Posted 03 January 2013 - 12:53 PM

Here is a metric that I would like to know. A QB ranking based on performance stats taken on their last 3 possessions of every game when the score range of losing by 13 through leading by 2. That might be revealing about the QB while putting in perspective the team's overall trajectory in adverse conditions vs the QB's ability to overcome said adversity. Before Haley I would bet that Rothlesberger is one of those QBs that would be at the top of those listings, Brady too. My idea of a franchise QB is a guy that if he has a 2 minute drill to work with while trailing by no more than 6 points that you can pretty much count on him to march down the field and score the TD. JMO. The final 2 minute drill is where we learn to love or hate our respective QBs.

#45 BillsVet

BillsVet

    Veteran

  • Members
  • PipPipPipPipPip
  • 6,255 posts

Posted 03 January 2013 - 01:01 PM

Bears GM Phil Emery apparently knows:

http://blogs.suntime...tremendous.html

A wordy but insightful answer to an issue that significantly limited Chicago this season. For all the heat he's taking about Lovie Smith, he's got a lot of evidence to demonstrate why he made the move.  I have a feeling Whaley is well-versed in stuff like this and it won't matter if Nix knows or not.

#46 since79

since79

    Probation

  • Members
  • Pip
  • 5 posts

Posted 03 January 2013 - 02:15 PM

In my first post, this is the type of thing I was talking about.  I really like your thoughts.

If you don't want to only rely on end of game stats you could add any drive when score is within 7.
Completion percentage when blitzed,  with zone or man coverage or both.  Including, yardage distance short or beyond gaining fist down on play.  Taking sacks or turnovers would also have to be included.

I think, metrics like this would help define poise under pressure,  abitly to make reads when pressured that provide positive results.  Keeping drives alive is the name of the game.

#47 San Jose Bills Fan

San Jose Bills Fan

    San Jose Bills Fan

  • Members
  • PipPipPipPipPipPipPip
  • 20,764 posts

Posted 03 January 2013 - 09:26 PM

http://profootballta...nt-work-in-nfl/

“As a practical tool, Moneyball does not work in the NFL because there are very few undervalued players and no middle class because of our salary cap,” Polian said.  “There is no middle class in football because the minimum salaries are so high, and because of the salary cap, a player will reach a point where you can’t keep him.  They go.  They’re going to get big money elsewhere.”

Mods, feel free to merge. I thought because of Polian it might deserve its own topic.

#48 jboyst62

jboyst62

    Hall of Farmer

  • Members
  • PipPipPipPipPipPipPip
  • 16,878 posts

Posted 03 January 2013 - 09:29 PM

I do not think we took the whole Analytics thing like it is meant.  It sounds fun to think of it like MoneyBall but we know it would not work.  The backing behind it I think is going to be a sort of Pro Football Focus kind of technique, and mostly to those already on the roster.  It is 3rd and 6, CJ gets X, FJ gets Y, Brad Smith gets Z = do Z.  It is not going to be something like trade Spiller to get Houstons #3 RB because he gained 5.5 yards on 3rd and shorts.  It's going to be what does this player offer us, ok good, use that to our advantage.

#49 BuffaloBillsForever

BuffaloBillsForever

    Veteran

  • Members
  • PipPipPipPipPip
  • 1,982 posts

Posted 03 January 2013 - 09:30 PM

I'm also VERY SKEPTICAL  of Polians scouting of player personnel in the modern NFL. He is even a dinosaur in his own game (scouting) these days. I doubt he has any knowledge about statistics and player evaluation.

#50 K-9

K-9

    Long Timer

  • Members
  • PipPipPipPipPipPip
  • 9,632 posts

Posted 03 January 2013 - 09:34 PM

View PostBuffaloBillsForever, on 03 January 2013 - 09:30 PM, said:

I'm also VERY SKEPTICAL  of Polians scouting of player personnel in the modern NFL. He is even a dinosaur in his own game (scouting) these days. I doubt he has any knowledge about statistics and player evaluation.

You mean current stats and evaluation, right?

GO BILLS!!!

#51 Punch

Punch

    Veteran

  • Members
  • PipPipPipPipPip
  • 2,085 posts

Posted 03 January 2013 - 09:38 PM

It's worth listening to Football Outsiders' Aaron Schatz on both the strengths and limitations of football analytics:

http://audio.wgr550....tm?pageid=28473

Additionally, WGR's Matthew Collier outlined what the Bills Dept. of Analytics should look like--- for the record, he's the station's sort of go-to guy in regards to advanced statistics:

http://wgr550.com/Wh...ook-Li/15183323

#52 BuffaloBillsForever

BuffaloBillsForever

    Veteran

  • Members
  • PipPipPipPipPip
  • 1,982 posts

Posted 03 January 2013 - 09:39 PM

View PostK-9, on 03 January 2013 - 09:34 PM, said:

You mean current stats and evaluation, right?

GO BILLS!!!

No, "moneyball" in football would be delving into more advanced stats. Beyond that of sacks, QBRating (which is flawed to begin with), total passing/rushing yards etc....

#53 K-9

K-9

    Long Timer

  • Members
  • PipPipPipPipPipPip
  • 9,632 posts

Posted 03 January 2013 - 09:48 PM

View PostBuffaloBillsForever, on 03 January 2013 - 09:39 PM, said:

No, "moneyball" in football would be delving into more advanced stats. Beyond that of sacks, QBRating (which is flawed to begin with), total passing/rushing yards etc....

You really think after nearly 40 years in the pro game that Polian doesn't have knowledge of player evaluations? Let's take OLmen for instance. I would say he has intimate knowledge of the various grading systems used throughout his tenure. And these are metrics that are unique to every team and wouldn't be available to outside organizations.

GO BILLS!!!

#54 UConn James

UConn James

    I'm a little behind on my paperwork

  • Members
  • PipPipPipPipPipPip
  • 8,231 posts

Posted 03 January 2013 - 09:49 PM

I find it a worthless comparison b/c Beane/Epstein et Al's mantra was to "buy statistics" rather than take the vague words of scouts.

I'm unconvinced that the largely individualistic stats of baseball can be transposed to football, and therefore, be some kind of predictor of production/wins. In baseball, it's cut and dried --- this guy hits this % against lefthanders/righthanders at 60/70/80/90/100mph curves/sliders/fastballs high and inside.... The variables for football are so widespread, I don't get how you can take the number of catches by a TE at Notre Dame, divide by however you arrive at a metric for how good the ND QB is, multiply how good your QB is, factor for the differences in quality of defenses, and arrive at some number that you think quantifies a football player. A play can go wrong for SO many reasons that would be totally unrelated to one individual.

Baseball and other sports like tennis, etc. are mano-a-mano, largely with no one else responsible for the important parts other than the dueling pair (besides fielding errors and such). And while I'm sure there probably are sophisticated rundowns of everything and everything, football is so much more a team sport that applying a numbers game loses a lot of its wind.

#55 Nitro

Nitro

    "Lead me, follow me, or get out of my way. " Patton

  • Members
  • PipPipPipPipPip
  • 1,602 posts

Posted 03 January 2013 - 09:58 PM

New school versus old school football thinking.  Trying something new is not a bad idea. Baltimore and Jacksonville are implementing this approach.  San Francisco is leading the way in the NFL.  Polian's touch is scouting talent wanned while in Indy leading to some poor drafts and was the partial reason he was fired.  Time to explore ways to improve the organization.

#56 djp14150

djp14150

    RFA

  • Members
  • PipPipPipPip
  • 705 posts

Posted 03 January 2013 - 10:08 PM

The big difference with football nd baseball....is baseball is more an individual driven sport because a batter is up aginst a pitcher....the only comparison to this natives would be CBs on Wzr in man free overage.

In football it's about everyone doing their responsibility for a play to work.  

There are numerous examples of players who excell with one team but fail hen changing teams.. Some of a players success can be attributed to surrounding players.    

When looking at game tape you can't analyze mistakes by players unless you knew whT the play call was.   You see a CB let a receiver go past him thinking he has safety help but there is none..thus a long pass ply....who is at fault?   You can't measure that.


A guard fails to pull to make a blok...why....is it because of his slowness or was there a mistake by another OZl that preventing him pulling by blocking his path.  

A RB doesn't pick up a blitz thinking the TE on that side is supposed to stay in and block before he releases thus picking up any blitzes.  Who is a fault?


I could go on....play mistakes are judgement calls...there is too much subjective gray area for good analytics to be done

Also you have more situations occur so results are more statistically significant in analysis.

For example you can look at a batter facing righties or lefties...or hitter tendencies with 2 strikes.  Over a career you can look at hitter vs  pitcher numbers and see who has had unusual success or failure.   This type of data isn't as robust given player shorter careers and some players only face each other once ever in there careers if st that.



View PostBuffaloBillsForever, on 03 January 2013 - 09:39 PM, said:



No, "moneyball" in football would be delving into more advanced stats. Beyond that of sacks, QBRating (which is flawed to begin with), total passing/rushing yards etc....

It's also system and style driven....under Shanahans system he can find good RB cheap and they excel because of his system.  The RB style needs to match the team running style.  

Cedric benson is an example of a RZb who was a bust then turned very good after changing teams.  

An example is Herschel walker trade....he was a bust in part because his running style didn't match the running style on his new team.

Edited by djp14150, 03 January 2013 - 10:11 PM.


#57 Dragonborn10

Dragonborn10

    Veteran

  • Members
  • PipPipPipPipPip
  • 2,106 posts

Posted 03 January 2013 - 10:20 PM

Analytics and moneyball do not mean the same thing.  I think what Polian is saying is that there are no small market teams that can't keep up with the Yankees.  Every NFL team can and will spend up to the cap.  There is very little difference between highest and lowest payroll in NFL and in fact with the new CBA there is a very high minimum amount that must be spent on salary.  In baseball the entire budget of the Royals or Athletics may be less than what the Yankess pay their infield.
Analytics however is using new version of stats while breaking down film to evaluate players or groups of players - ie. sub-packages or units like nickel secondary and 4 WR sets.  The days of the short shuttle and 40 yard dash in shorts should be long gone.  Why they still put players through that non-sense is mind boggling. The greatest WR of all time ran a 4.6 in shorts.  Was Jerry Rice every caught from behind before the age of 38???
Analytics tries to seperate the value of the RB from his offensive line and the defense the RB is facing.  It is just a tool that aides what guys see on film.  It is not cap-management and salaries though I'm sure it helps when it comes time to give a guy a big contract.

#58 dave mcbride

dave mcbride

    All Pro

  • Members
  • PipPipPipPipPipPip
  • 8,648 posts

Posted 03 January 2013 - 11:54 PM

View Postsince79, on 03 January 2013 - 12:23 PM, said:

First post sorry for the length but I wanted to jump in.

Buddy Nix’s response is exactly what Billy Bean faced when he first introduced money ball.  Scouts would go by feel or he looks like a ball player.  What his approach did was identify players who don’t look the typical part who had potential because they would be overlooked.  How he applied it was to be economical because he would not have to pay the same price for a player overlook and often times was statistically better than going by feel.  In baseball the metrics are easy to apply because ever pitch, hit ,fielding play, at bat are charted routinely.  In football it would be harder because there are more intricacies in each play but trends most certainly could be developed and understood.

I agree you cannot pick a player based on these analytics alone.  If you take the Patriots as you example, they have been using these metrics to help for several years.  Does it seem year after year they produce better than other teams?  Granted on the offensive side of the ball Brady can make anyone look good.   San Franciso has adopted analytics to assist and they have turned their franchise around as well.  Certainly Harbaugh helps so we cannot give the analytic or him all the credit.

What is interesting to me, the most is how we can use this in ferreting out our next quarterback, both by the draft and FA.  What is best about Brady, Maning etc is their ability to perform under pressure.  We need to find a quarterback who gets yards when you need them most with minimal mistakes.  Not in games trailing by large margins, or blowouts.  It’s the close games that separate the best.

Before the Brandon promotion I was thinking we need to analyze the quarterbacks in some way to determine who has these qualities. I was perusing the net and came across this article and it seems interesting to me.  
    Predicting the Unpredictable…Projecting a College QB to the NFL with a Mathematical Formula   Vers 3.0

    NFL Draft "Moneyball" for College Quarterbacks


http://www.fantasyfo...0_2-17-2011.htm

We all want change,  we want forward thinking and are we going to rely on Buddy, Whaley’s or the new coaches gut to pick our next Quarterback.  We cannot afford to miss again.  

After Ferguson we had names like Dufek, Ferragamo, Marangi before Kelly came along.
We have had a similar unimpressive list since Flute and Bledsoe.  

Think outside the box and realize this may be a good thing to be analytical.
SF started it in 2005 or so and it was a failure for a number of years. They didn't turn it around until they hired harbaugh, although their moneyball guy, paraag marathe, is still a higher up in the organization. http://sfist.com/200...yball_49ers.php

#59 bowery4

bowery4

    "‘Chance favors the prepared mind‘’

  • Members
  • PipPipPipPipPip
  • 4,201 posts

Posted 04 January 2013 - 12:38 AM

View Post26CornerBlitz, on 03 January 2013 - 10:06 AM, said:

Check out this site:

http://www.footballoutsiders.com/
Thanks, yes I was aware of the site. But I was actually curious as to stats about how successful these systems are (in win and loss terms, if there is any kind of thing like that?). I have been aware of the idea since a few years ago TBH when the whole going for it on 4th down thing came up.

Edited by bowery4, 04 January 2013 - 12:38 AM.


#60 dave mcbride

dave mcbride

    All Pro

  • Members
  • PipPipPipPipPipPip
  • 8,648 posts

Posted 04 January 2013 - 08:51 AM

View PostSan Jose Bills Fan, on 03 January 2013 - 09:26 PM, said:

http://profootballta...nt-work-in-nfl/

“As a practical tool, Moneyball does not work in the NFL because there are very few undervalued players and no middle class because of our salary cap,” Polian said.  “There is no middle class in football because the minimum salaries are so high, and because of the salary cap, a player will reach a point where you can’t keep him.  They go.  They’re going to get big money elsewhere.”

Mods, feel free to merge. I thought because of Polian it might deserve its own topic.

No offense to Polian, but I think he has a very thin understanding of the concept: http://fifthdown.blo...ball-football/?