To make a meaningful recommendation for special teams strategy, you would need to access the value of each outcome across the curve. You touched on adjacent, negative side effects of returns but don't provide a compelling metric of value other than gross yards. In football analytics theory, not all yards are created equal.
I'm not saying your final conclusion is incorrect, but it needs to take into account the disproportional impact of the sparse right-tail outcomes to the more likely outcome out losing yards compared to a touchback on the return.
The analysis seems incomplete.
To make a meaningful recommendation for special teams strategy, you would need to access the value of each outcome across the curve. You touched on adjacent, negative side effects of returns but don't provide a compelling metric of value other than gross yards. In football analytics theory, not all yards are created equal.
I'm not saying your final conclusion is incorrect, but it needs to take into account the disproportional impact of the sparse right-tail outcomes to the more likely outcome out losing yards compared to a touchback on the return.