LCD is not a magic wand that will make all the best decisions at all times. If it was, everyone would be using it all the time and there would be no fun setting the lineups. It is imperfect, but it had changed slightly lately and that's all I wanted to point out. The effectiveness of the tactic is a whole other matter.
I ran a couple tests on LCD lately to see what it would do with a reversed lineup (C at PG, PF at SG, etc...) and reserves swapped with starters. Not only did all of the starters start as assigned, but the coach never deviated from the depth chart, and the inferior starters still played more minutes than the superior reserves. I was surprised.
Are you saying that it is still possible for LCD to sometimes deviate from the depth chart?