Question on Analytics

If you're using statistics/analytics to determine what you should do when, aren't you sometimes getting in your own way? For instance, if the league average for running the ball on first down is something like 70%, the coaches are going to guard against it, so passing on first down may be easier. That means that in an analytical game plan, you would throw on first down a lot.

So, if suddenly everyone is throwing on first down because analytics tells them to, won't defenses adjust and make it more difficult to throw on first down? How is this captured in Analytics? Do teams just look at the week before, or maybe the last few games? Or do they look at a body of work over several seasons?

Kyle Allen got sacked 4-5 times on first down Sunday. That doesn't happen unless the defense knows a pass is coming and is out to stop it.

 

In a similar vein, who chooses what numbers to plug in to their formulas? I think I can make almost any decision seem like the right one based on the numbers, as long as I get to choose the numbers.

Except maybe that bizarre punt formation Indy tried a couple years ago...

 

It's a tool, not a religion. I also don't know enough about it's overall efficacy yet. The Eagles are apparently data driven, but they're also 5-7.

 

I started writing about baseball twenty years ago and got very into the analytic aspect, even coming up with a couple of new metrics of my own that never caught on. I rely on certain analytical metrics very heavily because they're useful and reliable, but that's because it's easier to isolate variables in baseball, especially when it's just pitcher versus hitter. I've been writing about football nearly as long but I rarely use any metrics for that due to how many variables are affecting every single play. It's just harder to boil down performance in football to this number or that number because of how many different things are going on. When they bring up "analytics" on NFL broadcasts, they're generally just talking about percentages for how often teams succeed when going for it on fourth down from a certain distance and other similar situations. Analytics won't tell you when to pass or run, as that's always dependent upon your scheme and personnel as well as the scheme and personnel of the opposition.

 

Based on Tepper's interview at the beginning of the year, he was talking about Analytics when he discussed how great first down passes were.

 

I don't listen to interviews of sports franchise owners for the same reason that I don't listen to talk radio. Why waste time on people who don't know what the fuck they're talking about?