Jump to content

good article on analytics in the NFL


Recommended Posts

Bills nugget: Lorenzo Alexander was resigned at a low price because he had, in baseball terms, a suspiciously high BABIP (batting average on balls in play). He got 12.5 sacks, but didn't get to the qb much beyond those 13 times. The view is that he can't reproduce those numbers given his low number of hurries.

 

http://mmqb.si.com/mmqb/2017/06/27/nfl-analytics-what-nfl-teams-use-pff-stats-llc-tendencies-player-tracking-injuries-chip-kelly

 

Link to comment
Share on other sites

What did you think of it? You didn't opine.

 

Thought the history of number crunching in the NFL was interesting even if it didn't go by the modern name of analytics. That's why I posted the piece to begin with. The analysis on LoRax is something to look at closely for the '17 season.

Link to comment
Share on other sites

The lorax comment is one that has been discussed here. Honestly, I think any reasonable person would expect him to go back towards his averages but still hopes he stays high producing. Even if he trends down though he can be a good contributor and role player

Link to comment
Share on other sites

The lorax comment is one that has been discussed here. Honestly, I think any reasonable person would expect him to go back towards his averages but still hopes he stays high producing. Even if he trends down though he can be a good contributor and role player

Agreed. And at least the Bills didn't overspend.

 

I also thought the passage on Julio Jones was interesting given the Sammy situation.

Link to comment
Share on other sites

I think the point was that it was already being discussed

No it wasn't -- the reason i hadn't seen it was because it was buried on p. 2. It has no commentary either -- just one quick and non-substantive response. Thread nazis are the worst (not that 26cb is one).

Link to comment
Share on other sites

They are putting a lot of faith in Lorax to be able to switch positions and replicate his unicorn of a season in a different scheme and as a 34 year old. He's a good leader and I like his attitude but I don't think his play on the field is sustainable. He will do ok but that will be a position of need they will need to address next offseason

Link to comment
Share on other sites

The lorax comment is one that has been discussed here. Honestly, I think any reasonable person would expect him to go back towards his averages but still hopes he stays high producing. Even if he trends down though he can be a good contributor and role player

 

This. He's making a shade under $3M/year, which is hardly bank-breaking.

 

And is he even pencilled-in as a starter? I'd figure the starting LB's would be Hodges-Ragland-Brown.

Link to comment
Share on other sites

They are putting a lot of faith in Lorax to be able to switch positions and replicate his unicorn of a season in a different scheme and as a 34 year old. He's a good leader and I like his attitude but I don't think his play on the field is sustainable. He will do ok but that will be a position of need they will need to address next offseason

I think it depends on what you are penciling him in for expectation wise. As a vet leader that fills a narrowly defined role he's fine.

 

That said, if investing a little extra in a first off the bench guy this offseason I would've done MG over lorax (who also likely doesn't reproduce last season but is at a position we lean on heavily and I think is more likely to have a larger impact).

 

That last bit just an opinion, not one I'm presenting as strongly factual

Link to comment
Share on other sites

 

This. He's making a shade under $3M/year, which is hardly bank-breaking.

 

And is he even pencilled-in as a starter? I'd figure the starting LB's would be Hodges-Ragland-Brown.

And really if we are in nickel 2+ out of 3 snaps then the starter we need to discuss is 3rd corner/safety not 3rd LB which has become a sub package player in the NFL these days

Link to comment
Share on other sites

And really if we are in nickel 2+ out of 3 snaps then the starter we need to discuss is 3rd corner/safety not 3rd LB which has become a sub package player in the NFL these days

Yup.

Link to comment
Share on other sites

The problem with using "analytics" in sports is there are too many variables and the sample size is not nearly enough to support a statistical analysis. It sounds nice and I know Mike Dope is all about it, but you will get the same results using analytics as not. It's like a coin toss. Flip a coin 4 times and tell me if you get 2 heads and 2 tails. I'll bet not.

Link to comment
Share on other sites

The problem with using "analytics" in sports is there are too many variables and the sample size is not nearly enough to support a statistical analysis. It sounds nice and I know Mike Dope is all about it, but you will get the same results using analytics as not. It's like a coin toss. Flip a coin 4 times and tell me if you get 2 heads and 2 tails. I'll bet not.

As with any statistics- you need a wise person figuring out the meaningful metrics, how to measure them and when they become significant.

 

To pretend this is worthless work is as silly as to pretend it's infallible

Link to comment
Share on other sites

As with any statistics- you need a wise person figuring out the meaningful metrics, how to measure them and when they become significant.

 

To pretend this is worthless work is as silly as to pretend it's infallible

I think that's the toughest part about analytics with football vs the other sports too. 16 games per season makes for a much smaller sample size, so you really need to try to nail down what stats you feel transcend the surrounding cast & schemes since they change so much more frequently within the data set.

Link to comment
Share on other sites

As with any statistics- you need a wise person figuring out the meaningful metrics, how to measure them and when they become significant.

 

To pretend this is worthless work is as silly as to pretend it's infallible

I didn't say statistics was worthless but I was implying the idea of trying to use it to predict out comes was foolish for the reasons I posted. Which wise person could accurately determine the number of heads if you toss the coin 5 times. Now if you toss it a thousand times I bet he is pretty close, and that is with only one variable.

Link to comment
Share on other sites

But is that assumption correct? All the data in the world doesn't help if you make the wrong decisions once you have that data

I guess you didn't read the article. The necessity of wisely using the growing quantity of data available was a central point.

Link to comment
Share on other sites

×
×
  • Create New...