IMO----One of the most frustrating things with CMO is the AI-handled attacks on subs. It generally goes something like this:

A sub fires off a missile a one of your ships. The missile is detected as it breaks the water.

You dispatch some kind of air unit (helicopter or Poseidon) to prosecute the sub.

The air unit flies there, drops one, maybe two sonobouys, and then flies around in aimless clockwise circles without locating it, in what seems like forever.

You ask yourself why isn't the air unit dropping more sonobuoys?

After a while, you get tired of watching him, so you manually drop one here (active), one there (passive), one here (deep) and one there shallow, and within a minute you have located the sub. Now you let the AI-controlled a/c back into the action.

But, the A/C isn't ready just yet. He continuously flies in circles. Sometimes he gets close to the target, and sometimes he gets further away.

Meanwhile, the sub moves further away from where you knew him to be.

Again, you get tired so you manually tell the a/c to attack. It does, and it gets a hit, but the sub doesn't die just yet.

You sit back and let the a/c finish off the sub, but nope, he's back to flying around in circles, and the sub is still moving further away from the area.

And, again you get tired, so you manually direct the a/c to re-attack the sub, and it does so and sinks it.

I then replay the events once again. This time I manually send the plane to a location, and handle all of the sonobuoy drops myself. In 5 drops or less, I have the sub located. In two, or no more than three, attack passes, the sub has been destroyed.

So, it's not really a question of sensors not finding the sub, or the torpedoes not being able to destroy it. Those can happen easy enough. It's a question of the AI's control over what it should be doing to locate and attack subs.

Do most people just handle everything manually, or am I being too critical? I suppose that my a/c would have eventually found the sub, if I would have left it alone, and then destroyed it. But, sometimes these a/c are flying so far away that they barely have enough fuel left when they get there, so they can't really afford to fly in circles forever.

I believe that the AI-ASW function could use a tweak.

Thoughts?

Agree 100%

I'd bet that additional sonobuoy drops don't happen because the game is programmed to drop sonobuoys 1.5 (I think) the sonobuoy sensor range away. In this case, the detection uncertainty zone probably isn't big enough to be outside of that, so no auto drop.

I think the best approach is not use the sub-strike or launch and manual attack for long-range sub-intercepts but set your own ASW area mission in the area, which would allow for it. you can also move and reshape based on the info you know, and the AI isn't programmed to yet.

Mike

BDukes wrote: ↑Fri Aug 18, 2023 12:24 pm I'd bet that additional sonobuoy drops don't happen because the game is programmed to drop sonobuoys 1.5 (I think) the sonobuoy sensor range away. In this case, the detection uncertainty zone probably isn't big enough to be outside of that, so no auto drop.

I think the best approach is not use the sub-strike or launch and manual attack for long-range sub-intercepts but set your own ASW area mission in the area, which would allow for it. you can also move and reshape based on the info you know, and the AI isn't programmed to yet.

Mike

My general style is to use buoy equipped aircraft on ASW patrol missions that way they lay a larger buoy net and usually get better results on auto attack. I try to use, when possible, aircraft equipped with dipping sonars, or mad, or maybe on older scenarios just an MD-500 with a torp on ASW strike as a backup since I usually let search planes stay up after wespons used as long as they have fuel to monitor the buoy and will manually launch a replacement patrol plane before the first RTBs on fuel so there is overlap.

tylerblakebrandon wrote: ↑Fri Aug 18, 2023 12:58 pm My general style is to use buoy equipped aircraft on ASW patrol missions that way they lay a larger buoy net and usually get better results on auto attack. I try to use, when possible, aircraft equipped with dipping sonars, or mad, or maybe on older scenarios just an MD-500 with a torp on ASW strike as a backup since I usually let search planes stay up after wespons used as long as they have fuel to monitor the buoy and will manually launch a replacement patrol plane before the first RTBs on fuel so there is overlap.

Yeah that's a good point! Dippers definitely help.

M

[/quote]

Yeah that's a good point! Dippers definitely help.

M
[/quote]

I like your idea of using smaller patrol areas and adjusting them based on available data. Would be useful especially if I just have buoy equipped choppers.

I'm definitely an area-search guy, I basically only use strike missions for prosecuting large, fixed land targets where I want to only hit some of the facilities (airbases, for example). I have seen the circling behaviour though, I think it's to do with aircraft not extending enough before turning to attack, so they're still in the turn when they reach the correct launch point and they miss the opportunity.

I spent part of the afternoon comparing the Search/Patrol method to the Strike method.

There is really no comparison, in my opinion.

The Search/Patrol method works great. The Strike method does not.

While the Search/Patrol method methodically processes the target, the Strike method usually drops a couple of sonobuoys in the general area of where the sub WAS, and then flies around in circles until it runs out of fuel and has to RTB.

I agree with you, the Search/Patrol method is much better.

DWReese wrote: ↑Sat Aug 19, 2023 9:32 am I spent part of the afternoon comparing the Search/Patrol method to the Strike method.

There is really no comparison, in my opinion.

The Search/Patrol method works great. The Strike method does not.

While the Search/Patrol method methodically processes the target, the Strike method usually drops a couple of sonobuoys in the general area of where the sub WAS, and then flies around in circles until it runs out of fuel and has to RTB.

I agree with you, the Search/Patrol method is much better.

The trick is programming the AI side to do this. I have some code now that will generate reference points x miles around something now (it could be the contact or another reference point). Just need to modify the code to generate around the contact and add to a mission. Might just build out a sub-scenario next (after Baldurs Gate muahahahahaha!)

M

Any scenario that you make will be much appreciated and liked.

Yeah AI lack of over extension and maneuvers probably my only complaint in CMO. Especially dogfights. For instance I had a Newer F16 in a dogfight with a Mig23. The F16 can easily out outmaneuver or just go low or high and get up over the Mig. Instead both just sit there on the same altitude circling each other for 5 minutes trying to get a missile or gun shot. I had to manually tell the F16 to extend out and come back into the Mig23. I do hope the WVR coding for dogfights gets better in future ive lost way too many higher Gen aircraft to lower gen because of simple lack of tactics especially in large scenarios where I cant manually keep up with 20+ aircraft engaging at once

DWReese wrote: ↑Sat Aug 19, 2023 8:05 pm Any scenario that you make will be much appreciated and liked.

'

Thanks

Mike