Metrics are for.... | Page 2 | FinHeaven - Miami Dolphins Forums

Metrics are for....

I wonder how much analytics Bill Belichick or Ozzie Newsome use?

The Pats do use it:

http://www.fredberinger.com/2010/06/musings-on-the-future-of-data-and-predictive-analytics-3/

"The new England Patriots, Oakland A’s and the Boston Red Sox all use software to make decision before and during the game"

---------- Post added at 04:16 PM ---------- Previous post was at 04:14 PM ----------

The thing is none of it makes sense

This is also true. There are many statistical relationships that simply cannot be explained by rational thought. That doesn't mean they don't exist though :)
 
"Stats are for losers, the final score is for winners"
- Bill Belichick
 
There's a balance. It's needed just as much as film work, game planning, etc. You're outcry probably comes from Shouright's threads, which are more applicable to fantasy football, betting and the such. Or if I could say so myself; his spew is more like a college-keg-party-funneling-drink-til-you-puke-on-yourself version of what should be derived from statistical measurement and it's application to scouting, player performance, etc as it pertains to the actual game of football that is played on the field.
 
Here is the only "stat" that matters. And it has been proven. If you score more points than the other team, 100% of the time you win, everytime
 
The thing is none of it makes sense. There are to many variables that can't be determined by some statistics professor, who doesn't know the first thing about football. I don't care what anyone says numbers are deceiving and don't tell the whole picture. Film is the closest thing you can get. "Analytics" and "Metrics" are just the flavor of the month, everyone is caught up on the Billy Beane book/movie. This isn't baseball. Football is to damn complex for some formula or statistic. I wonder how much analytics Bill Belichick or Ozzie Newsome use?

Futurescout - you would probably be surprised by the data that Belichick looks at. He is functioning at an extremely high level and has a lot of metrics that he probably looks at. He probably sets the key metrics that he wants to see.
The biggest improvement this season was in the Chip Kelly fast pace Offense. Every team wants to better understand how Kelly is speeding up the game and we are delighted to see his QB Coach, Bill Lazor now join us as OC. Anyway the team that has devoted more resources to sophisticating the metrics for every position is the Philadelphia Eagles. They have employed the best brains available in a group reporting to GM and approach the draft in a highly analytical way. This is the future.
 
Futurescout - you would probably be surprised by the data that Belichick looks at. He is functioning at an extremely high level and has a lot of metrics that he probably looks at. He probably sets the key metrics that he wants to see.
The biggest improvement this season was in the Chip Kelly fast pace Offense. Every team wants to better understand how Kelly is speeding up the game and we are delighted to see his QB Coach, Bill Lazor now join us as OC. Anyway the team that has devoted more resources to sophisticating the metrics for every position is the Philadelphia Eagles. They have employed the best brains available in a group reporting to GM and approach the draft in a highly analytical way. This is the future.

Until people rely on it too much and stop watching tape. Then their team sucks and they get fired.
 
Until people rely on it too much and stop watching tape. Then their team sucks and they get fired.
They have to watch the tape 3 - 4 times to extract all of the data for future analysis. It isn't either watch the tape or analyze the metrics, it is watch the tape AND analyze the metrics.
 
Analytic tools are useful as a supplement for traditional scouting.

It's that way in any business. When I examined the manufacturing process my firm uses while looking for ways to integrate some best practices from Six Sigma, I did two things: I personally inspected the facilities and watched an entire lot go through production from start to finish, and I compiled statistical data for every lot we had produced since 2006.

I would have been missing some pretty key information if I'd neglected to examine trends, just like I'd have been missing some pretty important stuff if I hadn't actually watched what was going on.
 
It's that way in any business. When I examined the manufacturing process my firm uses while looking for ways to integrate some best practices from Six Sigma, I did two things: I personally inspected the facilities and watched an entire lot go through production from start to finish, and I compiled statistical data for every lot we had produced since 2006.

Ahh, DMAIC. Not to get off on a tangent, but...

I worked in mortgage loan servicing prior to the Subprime mess (I saw the writing on the wall & bailed), the company was big on Six Sigma which IMO did not mesh well when you are playing with other peoples $.

That said, JDW hit the nail on the head mentioning trends. Patterns too are crucial to football as well.
 
Ahh, DMAIC. Not to get off on a tangent, but...

I worked in mortgage loan servicing prior to the Subprime mess (I saw the writing on the wall & bailed), the company was big on Six Sigma which IMO did not mesh well when you are playing with other peoples $.

That said, JDW hit the nail on the head mentioning trends. Patterns too are crucial to football as well.

OFF TOPIC:

Six Sigma principles are definitely over-used. They were useful for us because we were in kind of a unique situation:

1 We have a product where failures are costly. It's not a simple matter of replacing the unit. If we don't pass lot inspection with import authorities because they pulled two bad units, we're out hundreds of thousands of dollars. Even if a single bad one slips in and gets caught, that means a fine and expensive laboratory testing. Not to mention we've just paid freight on useless merchandise that we THEN have to pay to have destroyed so that it can't be sold on the grey market.

2. We have limited manufacturing space (red tape), a labor intensive process that can't be fully automated, and we're at full production capacity year round. We couldn't overproduce and simply reject.

So it made sense to focus on process quality control.

END OF OFF TOPIC

-

At any rate, the whole idea behind applying metrics to football is not to come up with ridiculous pie in the sky conclusions, but to have a better understanding of how individual attributes and skills translate into production on the field. Statistical analysis can also be very useful for comparing and even predicting the production of a lot of different players.

Just as an example, a scout who does not use analytical tools looks at Andy Dalton and says "that is a player with a low ceiling." Someone who uses statistics to analyze players but does not actually evaluate them by watching the game film looks at Andy Dalton's career stat line and says "This guy is great!" Someone who uses both probably sees him for what he is: a player who has athletic limitations but who can still run an effective offense if put in the right situation.

Who has the most complete picture?
 
A Mike Mayock sentence goes something like this: Tape, tape, tape, kid, kid, tape, kid, tape, tape kid, tape, tape.

And that's all you need to know.

Mayock has been a boon to the type who somehow think it's a revelation to look at tape. As in, look at me, I'm watching the same thing as everyone else.

Perhaps we should throw a parade.

Fixation on tape is laughably simplistic. The high error rate in scouting is largely due to obsession with tape while ignoring the peripheral variables and how to weigh them. But I concede it won't change. Guys in this era are brainwashed to rely on tape, just like college football fans are now paralyzed to evaluate a team minus strength of schedule.

Two horses can be dead even at the top of the stretch. But how they got there and their likelihood to win the race, or subsequent races, can be vastly different. If you know the pedigree and the surrounding details, you understand that. It's not a surprise when there's a 7 length gap at the wire. But the goof with the binoculars and no background in the sport is cheering for the red saddlecloth and little else. If he sees Ryan Tannehill and Johnny Manziel looking roughly the same in a Texas A&M uniform late in their careers, he concludes they are parallel and their pro potential is roughly the same, if not advantage Tannehill. Heck, draft each one of them in the 8 spot. Disregard that one guy has been a special playmaker all his life while the other has been a perennial tease.

Tape lies all the time when conventional wisdom is the principal ammo of the beholder.

Bar stool types are shocked to be challenged. That's verified daily here. They are accustomed to screaming this guy sucks and that guy sucks and awaiting peer approval. That's how a modern day Steve Deberg-like Alex Smith and his lifetime pathetic 6.6 yards per attempt, just good enough to get you beat, is actually preferred above Colin Kaepernick and his 7.9 lifetime number and adjusted 8.1. It's rare for the second number to be superior to the first but the happy adjusters are wonderfully oblivious. Something in their system screamed suck so they stick with stuck. They don't even understand the criteria that places the second number above the first but none of that matters as long as the guy in the adjoining stool is equally bewildered.

futurescout is talented but he makes basic mistakes. Like yesterday when he announced that Cameron Wake has many top years left since he was inactive in his early 20s. That's happy adjuster caliber. Walrus properly scolded. Now the onslaught on numbers. I remember when I was new to Las Vegas and bet strictly on line of scrimmage mismatches. I scoffed at the eccentric nerds who were babbling about situational bias, and some vague concept of value. I threw away my stubbornness, studied what I'd previously dismissed, and prospered. Let it happen to you.

In the meantime, make sure you never speculate. :lol:
 
I think metrics can be useful but when you try to manipulate a stat to your liking and ignore the ones that are not then the metrics are pretty useless. Any person quoting metrics saying Tannehill regressed last year is being selective in what they choose to explore
 
A Mike Mayock sentence goes something like this: Tape, tape, tape, kid, kid, tape, kid, tape, tape kid, tape, tape.

And that's all you need to know.

Mayock has been a boon to the type who somehow think it's a revelation to look at tape. As in, look at me, I'm watching the same thing as everyone else.

Perhaps we should throw a parade.

Fixation on tape is laughably simplistic. The high error rate in scouting is largely due to obsession with tape while ignoring the peripheral variables and how to weigh them. But I concede it won't change. Guys in this era are brainwashed to rely on tape, just like college football fans are now paralyzed to evaluate a team minus strength of schedule.

Two horses can be dead even at the top of the stretch. But how they got there and their likelihood to win the race, or subsequent races, can be vastly different. If you know the pedigree and the surrounding details, you understand that. It's not a surprise when there's a 7 length gap at the wire. But the goof with the binoculars and no background in the sport is cheering for the red saddlecloth and little else. If he sees Ryan Tannehill and Johnny Manziel looking roughly the same in a Texas A&M uniform late in their careers, he concludes they are parallel and their pro potential is roughly the same, if not advantage Tannehill. Heck, draft each one of them in the 8 spot. Disregard that one guy has been a special playmaker all his life while the other has been a perennial tease.

Tape lies all the time when conventional wisdom is the principal ammo of the beholder.

Bar stool types are shocked to be challenged. That's verified daily here. They are accustomed to screaming this guy sucks and that guy sucks and awaiting peer approval. That's how a modern day Steve Deberg-like Alex Smith and his lifetime pathetic 6.6 yards per attempt, just good enough to get you beat, is actually preferred above Colin Kaepernick and his 7.9 lifetime number and adjusted 8.1. It's rare for the second number to be superior to the first but the happy adjusters are wonderfully oblivious. Something in their system screamed suck so they stick with stuck. They don't even understand the criteria that places the second number above the first but none of that matters as long as the guy in the adjoining stool is equally bewildered.

futurescout is talented but he makes basic mistakes. Like yesterday when he announced that Cameron Wake has many top years left since he was inactive in his early 20s. That's happy adjuster caliber. Walrus properly scolded. Now the onslaught on numbers. I remember when I was new to Las Vegas and bet strictly on line of scrimmage mismatches. I scoffed at the eccentric nerds who were babbling about situational bias, and some vague concept of value. I threw away my stubbornness, studied what I'd previously dismissed, and prospered. Let it happen to you.

In the meantime, make sure you never speculate. :lol:

One thing I totally agree with is the falsehood that players that didn't play much early on will play longer than others. I proved that wrong in a thorough analysis of qb's many years ago. Age hits these guys regardless. There are some exceptions but its definitely not the norm.

I remember people saying how John Beck would be able to play for such a long time and hell he is basically already done with his NFL career. Chris Weinke ended his NFl career basically on par with guys that came in at 21
 
Back
Top Bottom