The problem with analytics is not in the numbers themselves but in the interpretation of those numbers. This can be quite complex but a simple example would be if the NFL 2-point conversion rate was 60% then it would make sense for teams to go for the 2-point conversion after every touchdown. However, if you distill it down further and your team only converts 30% then it would make sense for your team to kick the extra points. But how far do you go with the numbers? What is your teams 2-point conversion rate in a dome on the road? What about at home on wet field? What about at home on a wet field heading toward the dawgpound vs towards the open end?
There are so many variables, which are important? Which are not? How do you weight the value of each variable?
Then there is the role of momentum and differences in performance under pressure situations? Up 10 going into the half vs down 1 late in the 4th.
And what about sample size? How many times do you have to try something in a certain situation to get a large enough sample size that your numbers are significant?
I like analytics and think there is certainly value to using it. But you cannot make all your decisions based purely on analytics. JMO
| 
