Belichick not a fan of analytics websites.

[Cynjin]

http://www.cbssports.com/nfl/eye-on-football/25423617/bill-belichick-rips-analytics-websites-says-theyre-not-real-accurate

 

 

I have to agree with him on this.  Those of you that worship at the alter of PFF will probably disagree.

[2006Coltsbestever]

3 minutes ago, Cynjin said:

http://www.cbssports.com/nfl/eye-on-football/25423617/bill-belichick-rips-analytics-websites-says-theyre-not-real-accurate

 

 

I have to agree with him on this.  Those of you that worship at the alter of PFF will probably disagree.

I am not a big fan of Analytics either. Either the player you choose is talented or he isn't the way I look at it.

[2006Coltsbestever]

I also think things like QBR are flawed. It doesn't show how Great a QB really is. Andrew Luck was near the bottom in QBR his Rookie season and won 11 games and was Great in the clutch.

[Cynjin]

Just now, 2006Coltsbestever said:

I also think things like QBR are flawed. It doesn't show how Great a QB really is. Andrew Luck was near the bottom in QBR his Rookie season and won 11 games and was Great in the clutch.

 

Completely agree, I have not seen a QB rating system that I believe to be good.  They all have significant flaws.

[amfootball]

I would love to know how Bill determines his analytics on a player. I really hope when he is retired he writes a book on this stuff. I think it would be fascinating.

[Cynjin]

Just now, amfootball said:

I would love to know how Bill determines his analytics on a player. I really hope when he is retired he writes a book on this stuff. I think it would be fascinating.

 

Maybe when he retires, though I doubt it, but until then he is not going to tell anyone.

[2006Coltsbestever]

2 minutes ago, amfootball said:

I would love to know how Bill determines his analytics on a player. I really hope when he is retired he writes a book on this stuff. I think it would be fascinating.

I would too.

[BOTT]

21 minutes ago, 2006Coltsbestever said:

I also think things like QBR are flawed. It doesn't show how Great a QB really is. Andrew Luck was near the bottom in QBR his Rookie season and won 11 games and was Great in the clutch.

He was actually in the top 10....not that it means anything.

[amfootball]

1 minute ago, Cynjin said:

 

Maybe when he retires, though I doubt it, but until then he is not going to tell anyone.

"Do Your Job" is probably the closest we are ever going to get to understanding Bill and I was happy he did and allowed his coaches to speak so openly and frankly. I think he understands the history of the game and wants his accomplishments documented so I do hope he shares more after retirement.

[BOTT]

6 minutes ago, amfootball said:

I would love to know how Bill determines his analytics on a player. I really hope when he is retired he writes a book on this stuff. I think it would be fascinating.

Well, he actually knows what he's looking at/for....unlike some of the websites.

[2006Coltsbestever]

7 minutes ago, BOTT said:

He was actually in the top 10....not that it means anything.

Andrew Luck's QBR was 26th in 2012. Look it up on ESPN.com. It was 76.5. Blaine Gabbert had a better QB Rating.

[BOTT]

1 hour ago, 2006Coltsbestever said:

Andrew Luck's QBR was 26th in 2012. Look it up on ESPN.com. It was 76.5. Blaine Gabbert had a better QB Rating.

http://espn.go.com/nfl/qbr/_/year/2012

[2006Coltsbestever]

1 hour ago, BOTT said:

He was actually in the top 10....not that it means anything.

You meant last year I see that now which he was. I was pointing out how the QBRating is flawed though by Andrew being 26th in 2012 when he was arguably a Top 10 QB then by going 11-5 and breaking the Rookie record for most Passing Yards.

[2006Coltsbestever]

3 minutes ago, BOTT said:

http://espn.go.com/nfl/qbr/_/year/2012

I was talking about Quarterback Rating. I guess I got it mixed up with QBR, It is 2 different things actually. If you go to espn.com and look up QB stats he was 26th regarding rating behind Gabbert.

[BOTT]

3 minutes ago, 2006Coltsbestever said:

You meant last year I see that now which he was. I was pointing out how the QBRating is flawed though by Andrew being 26th in 2012 when he was arguably a Top 10 QB then by going 11-5 and breaking the Rookie record for most Passing Yards.

See link above.  He was 8th in QBR in 2012.  Did you accidentally look at passer rating?

[2006Coltsbestever]

1 minute ago, BOTT said:

See link above.  He was 8th in QBR in 2012.  Did you accidentally look at passer rating?

Yes that is what I meant, Passer Rating. Thanks.  Passer Rating is overrated is what I meant from the beginning being overrated.

[chad72]

Bill's analytics strategy is "get lucky with a 6th round QB" and hope to ride the lightning in the bottle as long as you can. And of course, check the rule book for every rule that you can use to your advantage and hire a World War memory expert to decode opponent's taped defensive signals, and paper boys to steal scripted plays.  /kidding 

[BloodyChamp]

Bill's not only a good coach based on obvious results, but on the fact that he does that job of a coach by definition better than anybody else. He makes the most out of every resource, gets the most out of every player and gets the most out every team at his mercy.

 

QBR is garbage as we've discussed here many times. 

[oldunclemark]

who cares?

[Patagonia Chris]

I've always been skeptical of advanced analytics in football, simply for the fact that there is so much grey area. Unless you have a really experienced former coach breaking down the tape and assigning grades it's a flawed system. It's not like baseball where every offensive stat can be broken down into some hardcore stat. There was a Colts writer who used to be a slave to these new "advanced analytics" and it'd drive me crazy. Used it as an end all, be all when it really just came off as he was trying to hide his lack of knowledge about the game behind these numbers.

[Malakai432]

6 hours ago, Cynjin said:

http://www.cbssports.com/nfl/eye-on-football/25423617/bill-belichick-rips-analytics-websites-says-theyre-not-real-accurate

 

 

I have to agree with him on this.  Those of you that worship at the alter of PFF will probably disagree.

But......but......but.......PFF....... 

 

[Dustin]

Nobody is going to read the article, but Belichick says literally nothing about analytics in his comments just FYI. 

[Dustin]

1 hour ago, Pagano's Realtor said:

I've always been skeptical of advanced analytics in football, simply for the fact that there is so much grey area. Unless you have a really experienced former coach breaking down the tape and assigning grades it's a flawed system. It's not like baseball where every offensive stat can be broken down into some hardcore stat. There was a Colts writer who used to be a slave to these new "advanced analytics" and it'd drive me crazy. Used it as an end all, be all when it really just came off as he was trying to hide his lack of knowledge about the game behind these numbers.

 

I'll be the first person to tell you that PFF "Grades" (which are not the same as analytics btw) are worthless, but part o their grading process literally includes verification from former coaches at both the NFL and NCAA level. 

[SteelCityColt]

9 hours ago, amfootball said:

I would love to know how Bill determines his analytics on a player. I really hope when he is retired he writes a book on this stuff. I think it would be fascinating.

 

The book already exists

[amfootball]

I think analytics work better in baseball for obvious reasons given that sport is more individualistic. Football has so many moving parts and schemes and coaching philosophies. It really depends on what a team needs vs a player's skill set/production. Still, I do enjoy analytics as a fan.

[Superman]

8 hours ago, amfootball said:

I think analytics work better in baseball for obvious reasons given that sport is more individualistic. Football has so many moving parts and schemes and coaching philosophies. It really depends on what a team needs vs a player's skill set/production. Still, I do enjoy analytics as a fan.

 

Analytics is just as valuable in football as it is in any other sport. Good coaches figure out what information is useful and they use it.

[amfootball]

3 minutes ago, Superman said:

 

Analytics is just as valuable in football as it is in any other sport. Good coaches figure out what information is useful and they use it.

I am not denying its value but a sport like baseball lends itself better or perhaps what I mean to say it is easier to find the useful info than football.

[DougDew]

Statistics are used when you want/have to make a decision but have poor information to do so.  Real knowledge lies with knowing the circumstances behind the creation of the datapoints.  You need a lot of manpower close enough to the situation to observe what is going on...then you know the facts that underpin the data.

 

You need manpower close to the source.  The opposite uses stats.  Stats are used when you don't have manpower, and you are far away from the source, which is why computer techno guys like those sites. 

 

They can do their work and never leave their bedroom or even change out of their pajamas. 

 

They sell it like its a superior product, when it is actually far inferior.  Its just done that way because they can create "knowledge" using only a few employees.

[Superman]

1 hour ago, DougDew said:

Statistics are used when you want/have to make a decision but have poor information to do so.  Real knowledge lies with knowing the circumstances behind the creation of the datapoints.  You need a lot of manpower close enough to the situation to observe what is going on...then you know the facts that underpin the data.

 

You need manpower close to the source.  The opposite uses stats.  Stats are used when you don't have manpower, and you are far away from the source, which is why computer techno guys like those sites. 

 

They can do their work and never leave their bedroom or even change out of their pajamas. 

 

They sell it like its a superior product, when it is actually far inferior.  Its just done that way because they can create "knowledge" using only a few employees.

 

Faaaaaaaalse.

 

Analytics tells you that your QB is 20% more efficient on play action, and the defense you're playing gives up 20% more yardage against play action passing. Good coaching says "let's use more play action this week." 

 

Has nothing to do with manpower close to the situation or number crunching in your bedroom. It requires tons of film work and charting to come up with the raw numbers to begin with.

 

This silly resistance to having more information is, well silly. 

[DougDew]

31 minutes ago, Superman said:

 

Faaaaaaaalse.

 

Analytics tells you that your QB is 20% more efficient on play action, and the defense you're playing gives up 20% more yardage against play action passing. Good coaching says "let's use more play action this week." 

 

Has nothing to do with manpower close to the situation or number crunching in your bedroom. It requires tons of film work and charting to come up with the raw numbers to begin with.

 

This silly resistance to having more information is, well silly. 

Oh sure, stats can be helpful.  But I would want to know WHY my QB is 20% more efficient on PA and WHY the defense gives up 20% more yardage against PA.  Each situation in which each play is run is different from the previous.  Stats needs homogeneous situations in order for the data points to have meaning....situations that are the same each play, so the performances can be better assessed.  A football game, let alone a season, is anything but homogeneous. 

 

What kind of play was run just before the PA, was I successful because the opponent's CB was out a few plays with cramps and will it work this time against Josh Norman?.  Did my rookie ILB play the first time we met the team and he now has 7 games under his belt the next time?  That's real knowledge.  And every event is dissimilar from the next.

 

Experience and a good memory beats stats.

 

Not just in football, but, yes, paying three people to run stats on a computer while in their pajamas is much less costly than paying a labor force to observe, record, and analyze nonhomogeneous situations.  Its simply an inferior way to form an opinion because the people in the pajamas are assuming that every data point is created from a situation that is similar to the others.

 

Again, its not that stats are not useful, but they are primarily used when people want to try to understand what happened during a game or a season when they didn't actually watch it. (or were too drunk or distracted to notice)

[Superman]

10 minutes ago, DougDew said:

Oh sure, stats can be helpful.  But I would want to know WHY my QB is 20% more efficient on PA and WHY the defense gives up 20% more yardage against PA. 

 

Once you figure out why, a good coach still plans to exploit that presumed advantage. The "why" doesn't change anything.

 

Quote

Each situation in which each play is run is different from the previous.  Stats needs homogeneous situations in order for the data points to have meaning....situations that are the same each play, so the performances can be better assessed.  A football game, let alone a season, is anything but homogeneous. 

 

What kind of play was run just before the PA, was I successful because the opponent's CB was out a few plays with cramps and will it work this time against Josh Norman?.  Did my rookie ILB play the first time we met the team and he now has 7 games under his belt the next time?  That's real knowledge.  And every event is dissimilar from the next.

 

I used a very simplistic example, but all of what you're talking about can be broken down into statistical analysis. Formations, personnel, down and distance, etc., all can be charted and distilled so that you can analyze everything you're talking about.

 

Quote

Experience and a good memory beats stats.

 

A good memory of what? The exact statistical analysis we're talking about. Putting it in a spreadsheet doesn't change anything, unless you think a human memory is more reliable than good record keeping.

 

Quote

Not just in football, but, yes, paying three people to run stats on a computer while in their pajamas is much less costly than paying a labor force to observe, record, and analyze nonhomogeneous situations.  Its simply an inferior way to form an opinion because the people in the pajamas are assuming that every data point is created from a situation that is similar to the others.

 

Show me this argument that suggests football teams (or any sports organization) should stop watching film and just rely on the stats compiled by a group of people sitting in their pajamas. #strawman

 

The benefit of compiled stats and analytics to fans is that none of us has the time to sit down and watch every team, every game, and chart these things ourselves. That's before I even mention that the average fan doesn't notice changes in personnel, formations, etc. A group of pajama watchers charting and sharing this information is better than a group of casual fans arguing with each other about whether a team should use more play action or not. 

 

Quote

Again, its not that stats are not useful, but they are primarily used when people want to try to understand what happened during a game or a season when they didn't actually watch it. (or were too drunk or distracted to notice)

 

Statistics is basically just organized data. Whether you watched, understood, remembered, the difference between what you're talking about and what I'm talking about is basically just writing it down so that you can reference it later.

[jvan1973]

Then why does he have a guy like Erie Adams on the staff?  He is the ultimate analytics guy

[DougDew]

40 minutes ago, Superman said:

 

Once you figure out why, a good coach still plans to exploit that presumed advantage. The "why" doesn't change anything.

 

 

I used a very simplistic example, but all of what you're talking about can be broken down into statistical analysis. Formations, personnel, down and distance, etc., all can be charted and distilled so that you can analyze everything you're talking about.

 

 

A good memory of what? The exact statistical analysis we're talking about. Putting it in a spreadsheet doesn't change anything, unless you think a human memory is more reliable than good record keeping.

 

 

Show me this argument that suggests football teams (or any sports organization) should stop watching film and just rely on the stats compiled by a group of people sitting in their pajamas. #strawman

 

The benefit of compiled stats and analytics to fans is that none of us has the time to sit down and watch every team, every game, and chart these things ourselves. That's before I even mention that the average fan doesn't notice changes in personnel, formations, etc. A group of pajama watchers charting and sharing this information is better than a group of casual fans arguing with each other about whether a team should use more play action or not. 

 

 

Statistics is basically just organized data. Whether you watched, understood, remembered, the difference between what you're talking about and what I'm talking about is basically just writing it down so that you can reference it later.

Yes, it takes a lot of data to get it right.  By the time you feed all of the nonhomogenous data for all of the different situations into the computer, you will have gathered as much data as the old-school method.  From that point, its just a matter of choosing to pay three guys in pajamas and the capital investment and "ongoing upgrade maintenance" of a computer system, or paying the guys who gathered the information in the first place to analyze it. 

 

I think a lot of folks think that using modern methods simplifies the data gathering process.  It actually makes it more complicated because you have to gather just as much data as before, then take the next step in translating it into a language that can be fed into a program.  (assuming nothing gets lost in translation...i.e. not all data is fed into the system correctly...which often happens...by accident). 

 

Gathering a lot of data is an expensive time consuming PITA, so some statisticians short cut it and try to look at every situation as homogenously as possible. But if you don't gather enough data...the expensive part.....the created data point becomes a less accurate explanation of what happened. and statistics becomes a less useful method of predicting the future.

 

There are more than a few people on this forum who will say that player X is better than player Y and rattle off a stat to justify it.  They have no idea what actually went into creating the stat...which is the useful part when you're trying to explain the past or predict future performance.

 

Stats are useful.  They are a good supplement.  Incorporating them into your analysis adds value...and its better than guessing.  I would say that a RB that averages 4.5 YPC is better than a RB who averages 3.2 YPC, but if I didn't know A LOT about what all went into those numbers, there is a better chance of overpaying for a FA or reaching for a draft pick, IMO.

[shakedownstreet]

In other news, analytic websites are not a fan of Belichick

[Synthetic]

 

We live in an information age where everything is over analyzed to the extremes. 

 

I have to agree with him, cause all it takes is one bad game from a quarterback, and the internet goes wild with over-analyzing them. Andrew Luck has fallen victim of this over this past season with his rough games. 

[NewColtsFan]

On December 23, 2015 at 4:23 PM, 2006Coltsbestever said:

I am not a big fan of Analytics either. Either the player you choose is talented or he isn't the way I look at it.

 

The story doesn't say that.

 

The story says Belicheck is a BIG fan of analytics.    He's just NOT a fan of the current analytic websites that are our there and we all know.    PFF, and PFO and others.

 

[NewColtsFan]

On December 23, 2015 at 3:06 PM, jvan1973 said:

Then why does he have a guy like Erie Adams on the staff?  He is the ultimate analytics guy

 

Belicheck LOVES analystics -- ESPECIALLY Adams!   (Good find by you!)     But he doesn't like the analytic websites that are out there.    PFF, PFO and others.

 

All that said,  I'd take everything he says with a huge grain of salt.    If Belichick likes these sites,  I can't imagine him publicly saying good things about them.    That doesn't see very Belichick to me.

 

[jvan1973]

21 minutes ago, NewColtsFan said:

 

Belicheck LOVES analystics -- ESPECIALLY Adams!   (Good find by you!)     But he doesn't like the analytic websites that are out there.    PFF, PFO and others.

 

All that said,  I'd take everything he says with a huge grain of salt.    If Belichick likes these sites,  I can't imagine him publicly saying good things about them.    That doesn't see very Belichick to me.

 

Oh I have no doubt he has no use for those websites.  But a gut like Bill certainly knows the usefulness of analytics 

[2006Coltsbestever]

4 hours ago, NewColtsFan said:

 

The story doesn't say that.

 

The story says Belicheck is a BIG fan of analytics.    He's just NOT a fan of the current analytic websites that are our there and we all know.    PFF, and PFO and others.

 

I know I went back and read it. I am still not a fan of it though. It can be a useful tool but some GM's swear by it and mainly just use Analytics when analyzing players. I think it's a better tool for Baseball.

[jvan1973]

1 hour ago, 2006Coltsbestever said:

I know I went back and read it. I am still not a fan of it though. It can be a useful tool but some GM's swear by it and mainly just use Analytics when analyzing players. I think it's a better tool for Baseball.

I doubt any gm analyses players on analytics alone