Effectivness of decoys

[Hawk66]

Don't know if this was discussed before, but haven't found a corresponding thread.

I'm currently trying to improve the behaviour of AI-Subs and asking myself if it's realistic, that torpedos always gets distracted by decoys.

Is that 'real life' behaviour or should there be a modifier like that a torpedo gets only distracted with a probability of 50% ?

[SeaQueen]

Originally Posted by Hawk66

I'm currently trying to improve the behaviour of AI-Subs and asking myself if it's realistic, that torpedos always gets distracted by decoys.

Is that 'real life' behaviour or should there be a modifier like that a torpedo gets only distracted with a probability of 50% ?

It's not realistic for a variety of reasons. It would be better if there was a simple probability of the countermeasure being effective. You have to be careful when you do this, though, because with the LWAMI and torpedoes that can potentially re-attack their targets, the statistics of this might get kinda funny. Then you'll find yourself asking what does that probability of 0.5 really mean and is it too high or low? I can see people here geeking about that for years to come.

Also... something to consider, is that difference countermeasures might have different effectiveness values against different torpedoes. Countermeasure A, might work really well against the old Torpedo Mk.1 but be not-so-good against snazzy new Torpedo Mk.5 Spiral 15.

[Hawk66]

Thanks for the input SeaQueen. I've added a 50% probability, that the torpedo ignores CM...I 'looks' more realistic now...


BTW: Since AI subs cannot 'wireguide' their torpedoes do you think it would be a workaround to increase the probability, that the torpedo ignores the CM (let's say 75%), if started from a AI sub?

I don't know the situational awareness capability of the weapon officer/team in a real sub.
Do they guide the torpedo away from the decoys (if they are identified by sonar or even via the feedback-signal of the torpedo's sonar)?

[Dr.Sid]

I wonder how much torpedoes are fired against countermeasures for training.
The bigger chance for wireguided torpedoes is good idea. Isn't it part of LWAMI already ?

[Molon Labe]

Originally Posted by Dr.Sid

I wonder how much torpedoes are fired against countermeasures for training.
The bigger chance for wireguided torpedoes is good idea. Isn't it part of LWAMI already ?

LWAMI 4 had "variable decoy logic" but this was never put into the LWAMI 3 series for torpedoes. I think the logic looks a lot like what was done for "advanced" seaskimming missiles on the doctrine level.

[Hawk66]

I did a couple of changes regarding the sub/torpedo doctrines (based on current LWAMI):

• Torpedos:
  • probability that an AI sub fired torpedo uses CircleSearchPattern for a limited time after enabling or if it's lost its target / spoofed by a decoy. This is surely not a realistic behavior but it compensates a little the missing wire guidaince and against AI subs the kill rate seems to be improved
  • probability that a decoy is ignored
• AI subs
  • more variable/randomized attack logic with the possibility to fire a third weapon
  • clears the datum after the first weapon is fired (but might continue engagement)
  • in regulary, but randomized intervalls the sub 'listens' ("stealth mode") and also might do the 'Madman Ivan' manoever (uh, is this the right term for it? ) if no other platform is currently detected
  • if the sub manages to escape it might bypass the area
  • ToDo: possibility of active sonar search (limited time) if a torpedo is detected but not the launcher platform
  • ToDo: Snapshot on incoming torpedo. Might work with a special Attack doctrine command but if it works there would be the need of several doctrine files

When I'm finished and there's some interest I can upload the files.

[nemo67]

Hawk the term is not Madman Ivan but Crazy IVAN :rotfl:

[Dr.Sid]

The active is bad idea. In many scenarios the torpedo you first detect is not aimed at you. The active would help only if it is one sub on one sub. Even then you can use the torpedo to scare the target away, not knowing his position or at least not knowing it exactly, expecting him to increase speed and show up.

The more common term for the maneuver is 'crazy ivan'.

[Hawk66]

Originally Posted by Dr.Sid

The active is bad idea. In many scenarios the torpedo you first detect is not aimed at you. The active would help only if it is one sub on one sub. Even then you can use the torpedo to scare the target away, not knowing his position or at least not knowing it exactly, expecting him to increase speed and show up.

I see your point. Nevertheless I've seen this tactic in "RedStormRising" (sub simulator by Microprose) and sometimes it was successful.
The question is: Does it harm? So, if the chased sub prepares to accelerate anyway does it give its hunter any additional infos (apart of the frequency of the active sonar?) Perhaps it gets a datum and can fire a better snapshot...

Originally Posted by Dr.Sid

The more common term for the maneuver is 'crazy ivan'.

I knew, I was wrong here :rotfl:

[Dr.Sid]

In all situation it's good to know if the torpedo is aimed at you. Campaign opening missions in DW are good example. 6 to 8 subs fires at each other, and you are there in the middle guessing what to do. Going active would be suicide.

In one on one ti is somewhat less trouble.

Single ping just before launching the torpedo can be useful too, since you are giving your bearing anyway with the shot, and you can get precious range info. But AFAIK ping can be heard at larger distance.

[LoBlo]

Originally Posted by Hawk66

• ToDo: Snapshot on incoming torpedo. Might work with a special Attack doctrine command but if it works there would be the need of several doctrine files

If you get this to work, please share. I've been fiddling around with it off and on and its easy to do IF the AI sub has already detected the firing sub. However, when it gets tricky is when the AI detects the torpedo fired at it, but not the sub that fired it. The AI won't fire at a sub that it still doesn't know exist (in it's logic).

Originally Posted by SeaQueen

Also... something to consider, is that difference countermeasures might have different effectiveness values against different torpedoes. Countermeasure A, might work really well against the old Torpedo Mk.1 but be not-so-good against snazzy new Torpedo Mk.5 Spiral 15.

I always wished their was a modifier on the torp side (not just the CM side) to effect probability to track decoy, but alas, the engine was only made so that a CM has a blanket effectiveness on all torps.

At least one upgrade in complexity could be to start making individualized CM each with their own level of sophistication (read probability) for each country. I heard that the latest a British CM is pretty snazzy (frequency matching and such) and the Germans too.

[Hawk66]

Originally Posted by LoBlo

If you get this to work, please share. I've been fiddling around with it off and on and its easy to do IF the AI sub has already detected the firing sub. However, when it gets tricky is when the AI detects the torpedo fired at it, but not the sub that fired it. The AI won't fire at a sub that it still doesn't know exist (in it's logic).

Yep, it seems to work.
For example, change the SubAvoidWeap.txt doctrine
...
IF Not Snapshot THEN {
DEBUGOUT "Firing Back!"
;FireBest
Attack "53 cm Torpedo"
Snapshot = true
} ENDIF
....

That should work for Platforms, which have the 53 cm Torpedo on board.

Of course that means for a general solution you would need different doctrine files for the subs (since they have different loadouts).

Originally Posted by LoBlo

At least one upgrade in complexity could be to start making individualized CM each with their own level of sophistication (read probability) for each country. I heard that the latest a British CM is pretty snazzy (frequency matching and such) and the Germans too.

Don't think this is an engine, but only the current doctrine implementation limitation. I've added some torp modifier in that way, that in 50% of the cases the torpedo simply ignores the decoy (so if the decoy wouldn't be there at all).
Of course the DB could be enhanced with several kind of decoys and there could be several doctrine files for torpedoes to react differently.
But some tough work.

[Hawk66]

it seems even to be easier...just add an "Attack <Name of Torpedo"> for each available torpedo in the db (but it seems, that torpedoes only capable against surface platforms don't work, but only a minor issue). So, no platform-specific doctrines needed...although I don't know if this solution destabilizes the sim (but it never happened during my tests). Snapshot firing by Ships also works.

If somebody wants to test it I can provide the (isolated) doctrine file...

I'm finished with the changes of the improvements for sub ASW. Next challenge is to improve ASW handling of ship platforms and attack of ships by subs...the doctrine files are either non-existent or over-simplified.

[LoBlo]

Originally Posted by Hawk66

it seems even to be easier...just add an "Attack <Name of Torpedo"> for each available torpedo in the db (but it seems, that torpedoes only capable against surface platforms don't work, but only a minor issue). So, no platform-specific doctrines needed...although I don't know if this solution destabilizes the sim (but it never happened during my tests). Snapshot firing by Ships also works.

If somebody wants to test it I can provide the (isolated) doctrine file...

I'm finished with the changes of the improvements for sub ASW. Next challenge is to improve ASW handling of ship platforms and attack of ships by subs...the doctrine files are either non-existent or over-simplified.

I tried your suggestion and it works surprisingly well. AI subs now snapshot back even if they didn't detect the firing sub. Thanks

Needed a range limit though. Subs were snapshotting at torps 20nm away.

[Hawk66]

Originally Posted by Molon Labe

LWAMI 4 had "variable decoy logic" but this was never put into the LWAMI 3 series for torpedoes. I think the logic looks a lot like what was done for "advanced" seaskimming missiles on the doctrine level.

Pretty late, my question but I've left DW for almost a year due to other stuff:

What happened with LWAMI 4? I guess '4' stands for the 1.04 patch...