[MacD]

Quote: Originally Posted by Da_Czar

Shots for AI are not solely based on tendencies. They are based on ACE and ACE is dynamic. We like to wait until teams play together before assuming how they will play. The default for OKC is that PG and Westbrook share touches but depending on matchups and mismatches and hot cold streaks they could dynamically be altered. Also Westbrook is the pg he will naturally have more opportunities to call his own number. So if Westbrook has a huge mismatch or just a mismatch and he gets it going the game will dynamically give him more touches. And we have no idea how these teams will actually play. Once we have a large enough sample size ACE will be updated. But it will still be dynamic so every single game doesn't play out the same regardless of matchups and all that stuff I mentioned before. We could hard-code it so x always gets x amount of shots but that wouldn't play very smart once you notice that pattern. When you make something dynamic you inherently give up some control. A probable reason for the increased shots for Andre is the defense is most likely to gap or leave him. Especially AI defenders when help or rotations are needed. His spotup 3 point tendency is more important that the default point tendency in determining how many 3's he is shooting. Also as I said how is the defense playing him ? These are all factors. The default 3 point tendency has more to do with spacing than shot attempts because those were made more situational as I mentioned in the blog. A lot of work and testing was put into the tendencies and I'm certain there are some that need to be adjusted but overall they are much better than they have ever been. Also post scorers are difficult to get the ball to because they plays ran for them can branch because of defensive lapses or pressure. Also post plays are longer which means more time for something else to open up. Sent from my iPad using Operation Sports

Thanks for the answer Czar!

When it comes to tendencies I've already mentioned that the sample size is important, I don't want to draw conclusions based on one game. I also love the fact that the system is supposed to be dynamic, based on matchups etc, makes it more realistic and interesting. But in terms of OKC I think there might be a troubling pattern based on a few games I've played and saw online (again, hope I'm wrong here). Of course you guys are right in waiting to see how the teams play. I just assume that Westbrook won't play the same way as last year with PG around, and the few games gave me that impression.

When it comes to Andre it seems to be even more tricky. Defenses will leave him open, yet he's still a reluctant shooter (not sure there's a tendency for that) resulting in him passing up on open looks, but I assume that may be really difficult to implement. I'm okay with Roberson shooting a high number of threes once in a while (his career high in attempts is 9) but if he does it reguraly then it's a problem.