Depending on the priority set by the sim engine for all the scripts and doctrines running for a single AI controlled platform (and there can be MANY running at once for each platform, a tribute to the programmers) the platform may or may not respond to certain scripts or even doctrine level commands.
There is a variable entered by the sim in the doctrines called SetPriority, and this is largely what is used to resolve conflicting doctrine/script commands, and although I believe it can be defined, for every doctrine/script cycle for the platform it gets shifted on the fly by the sim.
Amizaur, knows much more about this than me, but we have had the same problem figuring out why certain doctrine level commands didn't work in certain situations and it is very much the same problem that mission designers seem to have with certain scripts not working correctly or at all, and we have narrowed down some to the way multiple command paths are run and handled by the sim, which is ultimately a strength of the game because of the flexibility it allows.
For example, the straight running torpedo bug (which has been fixed in the doctrine update for LWAMI 3.00b) was a result of multiple copies of the Torpedo doctrine opening for each contact and throwing off a variable counter that reset the torpedo to search again each time the sim shifted priority between them based on the contact environment.
Fortunately, this counter was unnecessary, most likely a holdover from Sub Command, so removing it solved the problem and now the torpedoes seem to work great.
