[mango_prom]

I've done a few tests with coach profiles and at least from my perspective, there's no impact on AI playcalling whatsoever. Also, it seems like the effect of shot locations is quite a bit different from the individual values we can globally edit with game sliders.
For example, 3pt maxed out in coaching profiles leads to more long range attempts, but not in an extreme way. Compared to maxing out 3pt in game sliders, or giving a players some high values manually, the effect appears to be more subtle.
So maxing out every shot slider in coaching profiles might be an idea to make teams play more fast paced without disturbing the overall distribution of shot attempts.

About playcalling in general, I have tested different touches scales in my association, using teams with mostly generated players.
I've been using a touches scale somewhat based on usage% lately, so most players are between 10 and 30, which should not make much of a difference if the distribution is correct.

Point guards do not get any touches if they're within this scale. So if my 2 best players have a touches value of 30, one of them being the starting PG, he gets no plays called for him (the same is true for fantasy draft rosters btw).
What's strange is that if I scale PGs differently, starting at 30 up to about 50, the AI actually calls plays for them, but overall play distribution gets kind of messed up by that.

Now even without any set plays called, PGs will still get involved since after plays breaking down, the CPU will often pass to them to take bailout shots or force a drive. But if I truly want to run the offense to a PG in a scenario different from default rosters, due to generated rookies or simply trades at the PG position, the only way is to work around a touches tendency that is not implemented correctly.

So right now, excluding PGs from the overall scale kinda works, but I'm not satisfied at all. The next step is to find out if the same differences exist for other positions, but maybe less pronounced.

It's kind of the same thing looking at Slimm's results using the Heat. Both test runs, no plays being called for Battier/Chalmers. Of course, most plays should be for Lebron/Wade, but looking at touches values, it makes no sense that the AI ignores the 2 completely.

In theory, shouldn't it all even out in the long run? So for example, giving Wade/Lebron both 90 touches and Bosh/Chalmers/Battier 60, wouldn't that mean in general, that plays should be distributed in a way that the sum of plays called for Lebron and Wade equals the sum of plays called for the other 3?
Seems logical to me looking at how the values are scaled (Wade+Lebron=180, Bosh/Chalmers/Battier=180). Will test this if I have the time.