Amizaur corrected the SubrocAttack doctrine sometime in the summer of 2005... so if there are any problems, I'd be very surprised if they hadn't been caught already. :)

Molon is correct about the AI TMA solutions creating their range errors, it's not the doctrine itself, as the AI uses the same doctrine as the players, entering the same range parameter.

If the range parameter is correct, the weapons will drop their torpedoes on target. To be precise, the weapon is always on target, but the solution can be quite bad for AI at first, just as for humans.

Cheers,
David

The problem in that the doctrine does not give the period of time for the good decision (Delay ASW missile Launch) - as soon as the contact is received, the trigger of doctrine work.

While the good decision will be produced, ASW of a missile already for a long time will be launched at pick up contact.
I think in 99 % cases it always will fly past, - if you do not ping for AI boat active sonar - I here! I here!!!. :p

I know the AI solution is only has good as how well they do their TMA but it seems they launch the ASROCS too early, sub anyway. I think it is something to do with the attackbest so if they pick you up early at long range they will use their longest range weapon.

Anyway to get them to firm up there TMA before they start lobbing things?

It seems quality of AI solution is only based on count of LOBS it has .. in other words of the time it holds the contact. I guess it could be modded to fire let's say after 6 minutes after detection or something like that.
Anyway .. AI will shoot again if the first missile misses, won't it ? It should have better solution at that time.

Regrettably, attack behavior for surface vessels is not controlled in the doctrine system.

Therefore, SUBROC firing for surface ships cannot be changed by modding.

Cheers,
David

Is it ? Anyway .. ships mostly picks you by active sonar, which is precise enough. While subs will mostly pick you by TA ..

The underlying problem is that the changes made to the aTMA to make it less accurate for human players also carried over to the process the AI uses to gain solutions.

This is not a problem by itself, in fact it's actually much better to have the AI come up with bad solutions first (as opposed to near perfect solutions on contact), however, SCS did not adjust the firing logic for AI platforms to compensate for the new uncertainty in solutions.

Thus, the AI tends to fire too early, on bad solutions (because it thinks it has a good solution).

Cheers,
David