Roy Halladay, ace pitcher of the Jays makes around $103 000 a pitch. When I mentioned this to Michael this week he responded with something along the lines of ‘I wonder what the ROI of each pitch could be measured?’. In theory, it likely could be, but it isn’t that simple.

How for instance, do you measure the return of a pitch that misses on the inside? It was a ball, when you want to be getting strikes, but inside pitches also tend to make batters move off the plate a bit more which means the opposite side of the strike zone is a little further out.

Or what about pitches that hit someone? The batter gets the base. But what if it was intentional?

There is just too many variables involved in the decision tree, I think, to be able to do that calculation.

So could you calculator the ROI of a test? I don’t think so, for much the same reasons.

What if by doing test A, I am setting the system up (quite accidently) in order to discover a problem identified by executing test B? And even more challenging is calculating the ROI of not running a test (either independently or in deference to another one), which is a decision we make all the time.

Just a quick thought that has been rattling around my head since Thursday. Further analysis is left up to you. 🙂