That game plan is partially developed by algorithm. But let me tell you a story ...
This past spring, Andrew Knizner was part of a trial run by MLB and a tech company to call pitches using an Apple watch. He wore one. The pitcher wore one. He would tap the type of pitch he wanted, and then the watch would offer a grid for the strike zone, and there were areas for out of the zone and in the dirt. Like you'd see with any strike zone. He'd tapped the area he wanted. That would be relayed to the pitcher. He could accept or shake, and if he shook the process would repeat. Knizner used this tech in bullpen sessions, not in game. The idea from MLB is to create a closed circuit that would be harder to crack and would limit team's ability to steal signs -- or eliminate it. Baseball is consumed by this, by the way. Anyway, they gave it a run. And now back to your point about an algorithm. This Apple watch thing strikes me as untenable for a few reasons.
-- One, it was laborious. Not efficient at all. Way slower than dropping a finger, getting a shake, dropping another finger, moving on.
-- Two, it opens up a whole new way for teams to evaluate catchers and catcher efficiencies. They could grade a catcher on the game he calls with data from the watch. How often he is shaken off. How often he drifts from the plan. We would see the same thing that has gripped umps with the strike zone, grip catchers, and maybe they'd be less confident in their calls.
-- Three, or just have the computer run the calls and the Apple Watch would tell the catcher where the next pitch is coming. The catcher and pitcher would learn at the same time from the A.I., and away w go.