Player Specific Tracking Data

[Pape]

The NFL has had it for a few years now.... But according to PFT, it was only available for teams own players. Next year the gates are wide open.... All tracking data is in play

will love to see if Ernie A can make use of such info.....

NFL teams to get more access to data that tracks players on every play

 

[Kontradiction]

IF? Adams will be on that the second it becomes available.

 

[Ivan]

I can’t remember his name but the Patriots have an IT guy that is already doing draft work in this area. I think he’s the guy they would task with this.

 

[1960Pats]

IF? Adams will be on that the second it becomes available.

And yet some Pats fans I talk to don't even know who he is. Incredible.

 

[WaterfallJumper]

I can’t remember his name but the Patriots have an IT guy that is already doing draft work in this area. I think he’s the guy they would task with this.

You're referring to this guy: Patriots won't say what Sean Harrington does; meet 'genius' who rejected Google for New England's scouting department

I think his software analytics had a slightly different focus, but yes, I assume that he would be involved in parsing through the data.

 

[primetime]

Problem with the tracking data, like all Big Data, is that there's so much crap and basically no way to test the validity of any model you could use to query it. Effectively, with that much informatiin and no real context to it, you can make something that will tell you whatever you want to hear. I'm skeptical there's anything useful at all in there besides noise. (Note I'm a social scientist and statistician by profession, so my skepticism comes with an insider's eye).

 

[Ivan]

You're referring to this guy: Patriots won't say what Sean Harrington does; meet 'genius' who rejected Google for New England's scouting department

I think his software analytics had a slightly different focus, but yes, I assume that he would be involved in parsing through the data.

Yep, that’s the guy and the article. Thanks.

 

[WaterfallJumper]

Problem with the tracking data, like all Big Data, is that there's so much crap and basically no way to test the validity of any model you could use to query it. Effectively, with that much informatiin and no real context to it, you can make something that will tell you whatever you want to hear. I'm skeptical there's anything useful at all in there besides noise. (Note I'm a social scientist and statistician by profession, so my skepticism comes with an insider's eye).

Fair points, but there's also a desired outcome here, rather than just looking at trends (not to say that all you do is just look at trends -- but I think you catch my meaning). Coaches might look for certain "success cases" and then use the data to determine who fulfills the criteria the best, rather than just aggregating data and massaging it until the like the way it looks.

 

[maust]

And yet some Pats fans I talk to don't even know who he is. Incredible.

 

[maineman209]

You're referring to this guy: Patriots won't say what Sean Harrington does; meet 'genius' who rejected Google for New England's scouting department

I think his software analytics had a slightly different focus, but yes, I assume that he would be involved in parsing through the data.

Probably working as an assistant to Adams.

 

[maineman209]

Problem with the tracking data, like all Big Data, is that there's so much crap and basically no way to test the validity of any model you could use to query it. Effectively, with that much informatiin and no real context to it, you can make something that will tell you whatever you want to hear. I'm skeptical there's anything useful at all in there besides noise. (Note I'm a social scientist and statistician by profession, so my skepticism comes with an insider's eye).

Ernie Adams has been tracking similar stuff - and adding the context to make it meaningful and useful - for decades, probably using paper and pencil. So his models are well-tested.

 

[primetime]

Fair points, but there's also a desired outcome here, rather than just looking at trends (not to say that all you do is just look at trends -- but I think you catch my meaning). Coaches might look for certain "success cases" and then use the data to determine who fulfills the criteria the best, rather than just aggregating data and massaging it until the like the way it looks.

The problem is there's no way of knowing what inputs lead to said outcomes, and there really isn't much in the way of theory there to test. You're just tossing a bunch of crap into a neural net and seeing what comes out. At best it's much worse than old-timey scouting.

 

[Sciz]

I feel like the biggest losers in all of this are veteran players. If a guy has a top speed of 19 MPH at age 30 and 18 MPH at age 31, now every team is going to know that without looking at any film. That’s bound to come up during contract negotiations, regardless of whatever the player might have improved at over time to compensate for the inevitable loss of athleticism.

 

[primetime]

I feel like the biggest losers in all of this are veteran players. If a guy has a top speed of 19 MPH at age 30 and 18 MPH at age 31, now every team is going to know that without looking at any film. That’s bound to come up during contract negotiations, regardless of whatever the player might have improved at over time to compensate for the inevitable loss of athleticism.

And without any indication that it even matters at all.

 

[jah]

Fair points, but there's also a desired outcome here, rather than just looking at trends (not to say that all you do is just look at trends -- but I think you catch my meaning). Coaches might look for certain "success cases" and then use the data to determine who fulfills the criteria the best, rather than just aggregating data and massaging it until the like the way it looks.

This. If the teams are smart, they'll apply machine learning by feeding an algorithm the data and (subjective) results. For example, label each play as positive or negative and construct an algorithm to maximize positive plays with that data.

 

[sb1]

Any chance of data tracking PSI in footballs?

 

[XLIX]

I do analytics as a career, and this would be fascinating. I am not good at spatial analytics used here though. But I can appreciate the sort of insight it can give.

 

[Steve102]

IF? Adams will be on that the second it becomes available.

What makes you think they don't already have Ernie on this? This is the Pats, right? Out there on the bleeding edge.

 

[Kontradiction]

I do anal as a career, and this would be fascinating.

Agree on fascinating.

 

[XLIX]

Agree on fascinating.

GFY.
Grow up, ***hole