Originally Posted by DoogMinAmo
To play devil's advocate, first of all, all fly outs are not deep to right...
But when you set out to prove or disprove something you create "test conditions" which aren't necessarily intended to mirror the real world. They are intended to isolate/reduce the variables (in this case the type of outs). Yes, it is unlikely to have a game where all 27 outs would be deep flyouts to right. But by using the hypothetical, you can reduce variables to just those two types of outs and prove that they are not equal.
As a slight digression, this is my issue with some of the stat-based SABRE conclusions I have seen. In some cases, I've seen conclusions which suggest correlations exist or don't exist but there are many statistical variables at work in those studies - no attempt to adhere to the "Scientific Method" of isolating variables with "all other things held equal".