The numbers look very promising, but one thing that concerns me about the pitch consistency set to zero is that, if I understand correctly, it makes a pitcher very erratic in a "random" way, isn't it? I've read the whole discussions last year how people experimented with pitch control/consistency and a sort of unexpected way these sliders affect how CPU pitches, but I still don't know how exactly control/consistency work.
My understanding (from the online manual) is that pitch control affects the size of circle around the desired pitch location, and the actual pitch is more likely to land inside that circle. So a pitcher with better control has a smaller circle whereas a poor control pitcher has a larger circle. If this is all there is, a pitcher would almost never throw a very erratic pitch (like behind a hitter).
Now, the pitch consistency changes the size of that circle by a huge margin occasionally (the frequency controlled by the pitch consistency slider) to sort of imitate how a pitcher just totally misses the target once in a while.
If this is the way it works, is it desirable to make pitchers very erratic like that?
Being a nerd I tried to do a simple experiment this evening to actually test this hypothesis, but I can only see 25 pitches at a time with pitcher analysis, so the small sample size kinda killed it for me. Data can still be accumulated, but it's quite boring to keep pitching fastballs down the middle with classic pitching (the best way to intend to pitch to the same location) and collect data in the end.
Does anyone know exactly the relation between control and consistency?