Comments on Advanced NFL Stats: The Passing Paradox Part 3

Wow. I stumbled upon this site looking for some s...

2008-07-17T14:09:00.000-04:00

Wow. I stumbled upon this site looking for some stats helping in my fantasy football rankings and I come up with this potentially game changing research. Wonderful work. Im not a math whiz so the formulas are a touch geeky to me but you do a brilliant job simplifying it.

One thing I wanted to add however.

1- Field position should factor as a team that is 3rd and 5 say inside the 40 yd line may run the ball on a draw thinking they would pound it on 4th and 1 if the element of surprise didnt work. Dallas did this a lot the past couple of years with Barber running shotgun draws.

2- A teams defensive capabilities may affect the offensive thought process as a confident coach who is defensive minded would want to score explosively (with more risk) via the pass to put his defense at an advantage forcing the opponent to pass.

Just my 2 cents.

True, but the Football Outsiders' approach is simi...

2008-02-21T09:57:00.000-05:00

True, but the Football Outsiders' approach is similar only in that they grade each team on a play-by-play basis. But DVOA is an ad hoc retrospective scoring system for ranking teams, not a risk/reward analysis tool.

What I'm proposing is using the risk/reward math to identify future optimum strategies for play selection and other decisions. I'm not proposing this as a prediction model or as a system for some sort of power ranking.

Isn't the Football Outsider's approach similar (ie...

2008-02-19T19:59:00.000-05:00

Isn't the Football Outsider's approach similar (ie, rate each play as success/fail depending on the down and distance (well the percent distance to go)?

Yes, the distributions would be heavily skewed rig...

2008-02-16T00:21:00.000-05:00

Yes, the distributions would be heavily skewed right. I'm not sure how it changes things either (yet).

Good point about "getting out of a hole." A good passing team might have more flexibility on 2nd down. I wonder though.

What if my team is a poor passing team? I run on 1st down. It's now 2nd and 8. Would I go for a run to make 3rd down more manageable, say 3rd and 4. Or would I want two pass attempts at making the 8 yards?

Great stuff. It's funny, and unsurprising, to see...

2008-02-13T18:16:00.000-05:00

Great stuff. It's funny, and unsurprising, to see the Denny Green Cardinals as the most risk-averse losing team. Of course we all remember the infamous Bears game, where the Cardinals built a big lead, and then decided the best way to maintain it was to go double tight end I formation, run Edgerrin James into the line three times, and punt to Devin Hester.

I disagree with one thing you say:

A coach can call plays with pure yardage optimization balance in mind until third down, commonly considered the do-or-die, make-or-break down. Then, he has to consider the risk of being forced to punt.

In reality, breaking down the analysis by down and distance opens up further complexities. What you should do on earlier downs is dictated not merely by yardage optimization, but by how good your team is at getting out of a hole. Your risk-reward analysis will still lead you to yardage optimization on 2nd-and-10 if you have a great passing attack that can rescue you on 3rd-and-10. But if you don't, then the fear of 3rd-and-10 leads you to more conservative play calls.

Brian,

Wow, fascinating stuff. I'm not muc...

2008-02-13T13:27:00.000-05:00

Brian,

Wow, fascinating stuff. I'm not much of a stats whiz. Are yards gained distributed with a standard Gaussian distribution? It seems that there would be a fat tail on the right (positive yardage) side. Not sure if that changes the analysis at all.