Shaolin Fantastic wrote:
View Post
It's an extreme trend though, unlike net ratings and defensive ratings which tend to differ by like 20%.
Anyway, my point with the 20 games is that we should rely on old data until we have enough new data - meaning we should not make (or advocate for) any changes based purely on current data until we have enough of it. Miles coming off the bench and Powell starting is still based on prior years data, which is more reliable than the 10 games we have. I was just pointing out that even if we really wanted to make the switch based on results so far (an outright bad idea as I've stated, and what people are advocating for based on a perception that Miles is not getting enough clean shots with the bench thus far), the existing data from this year does not support it, making the suggestion completely lack a leg to stand on. I would certainly never do anything so foolish as to suggest, for example, a change to the starting lineup based on samples as small as what is above.
Comment