re: Former Auburn WR Marcus Davis to be named WR coach…

quote:

---

I just really want to know the formula. Are these guys going back and watching receivers drop rates before and after being coached by Davis? How quickly they get in and out of breaks? How well they use their hands to beat press coverage.

I would love to see their laughable criteria. Probably the same clowns running PFF.

---

This is why I want to see the methodologies. These types of metrics also need to be refined over years and years.

- Construct valid definitions of that you're measuring (statistics), for each coach, OC, DC, HC (some of this would be done already)

- Deciding which individual statistic has a causal relationship to "success"

- What are the weights/factor loadings for each of the causal statistics chosen?

- Are the weights/factor loadings static, or do they change? (e.g., If a QB has great TD/INT ratio, but his negative rushing yards are absurdly negative versus only slightly negative. Would this matter?)

- If the factor loadings change, then you have to develop what's called a nomological network, where you map all the interrelationships between all of the statistics used.

IOW, this type stuff can spin up and get really complex, really quick. Anyone know if these analytics guys have PhD's?

FYI, this is an example of a simple nomological network :