Preferred WRA's?

[ARCNA442]

Setting WRA's is one of the most important tactical choices in CMANO. So I was wondering what the go-to settings are for the members of the community?

Personally, I tend to reset all torpedoes and defensive missiles to 1 weapon per salvo and limit the automatic firing range for surface to air missiles to 30nm.

What are your tricks to make optimal use of your weapons?

[SeaQueen]

It depends. I'm not sure there's an "optimal" choice. On the defense, the issue is balancing the probability of annihilating an incoming raid versus munitions expenditures. In scenarios where I'm worried about munitions, I shoot 1. In scenarios where I've got more than enough, I shoot 2. Sometimes if there's an incoming leaker I'll pause the game, and tell it to fire all remaining munitions at it, just to be sure to knock it down. Lasers are great because they can keep shooting where other weapons can't.

On the offense, I typically feel like I have more than enough weapons so I'm more likely to shoot 2, particularly against things like bombers or AWACS because I REALLY want those things dead.

[kevinkins]

There are so many variables. But what interests me is settings for the AI when designing a scenario. Once set, you would have to use lua to change them and the coding might be tricky and time consuming to write. The human player will have their own rules of thumb but they might not be best for the AI over the length of a scenario.

Kevin

[SeaQueen]

Too much is said about making "the best" decision or the "optimum" decision as if there's a mathematically dictated coarse of action. In some cases that might be true, but typically if it does exist, it's subject to very tight constraints and assumptions. If those assumptions turn out not to be true then they don't matter.

The best thing to do is make an decision (any decision!), explain to yourself why you think that's the best thing to do, and then try it. Fundamentally, the real measure of success in warfare is whether or not something works. If it works, it might not be the "mathematical optimum" but that doesn't matter. It's the "good enough" solution, which is what you really need anyhow. Then experiment with something different and see how you do. Ask yourself what the trade-offs might be? Is there something else you could change to alter what those tradeoffs might be? For example, having SAM sites shoot 2 improves their effectiveness at the expense of longevity. If you add in some ammunition trucks, and have them store a bunch of reloads on them, then maybe it's a better trade off?

[ARCNA442]

While I agree that tactics can and should change depending on the tactical situation, I believe that there are indeed a set of optimum tactics that cover the majority of common situations - particularly when it comes to something like WRA's, which can basically be reduced entirely to mathematics.

For instance, if you have a long range SAM, firing 2 at subsonic high altitude targets may be "good enough", but mathematically it offers an insignificant increase in effectiveness in exchange for a massive reduction in magazine depth. Now, you can come up with some specific scenario where that insignificant increase is warranted or where magazine depth isn't an issue, but that really doesn't change the fact that there is a mathematically optimum course of action to achieve maximum effect with minimum weapons and that the vast majority of the time you wont be in those specific scenarios that demand something else.

[RIPperSVK]

Some typical settings on my side:
* reduce range of air-to-air missiles to 75% - to increase PK against modern enemies
* I allow kinematic range for torpedoes for manual launches
* for anti-missile defense (e.g. ship defensive missiles), I often set 1 missile per target (if hit % is high, ie for modern missiles) - this handles swarm attacks better as I want to attack as many missiles in one salvo as possible. For leakers (engaged by followup salvos) I can manually increase missile count. I.e. quickly destroy most of incoming missiles, and for the few leakers be damn sure to kill them.

I'm handling attacks against surface or land targets mostly manually, so I use "mental" WRAs there:
* missiles against ships - depends on the ship's anti-missile defense capabilities. If it's strong, then usually "launch everything!!!!"
* runways - 8
* taxiways - 8
* tarmac spaces & similar (when I want to hit parked planes) - 1
* for the above - if the airfield is heavily defended, if I have lots of ammo, or I want to be really sure I hit them - add a half

[SeaQueen]

Well... okay... in the case of WRA, there is an optimum, and it actually does come down to mathematics.

What's the probability of annihilating a raid consisting of N aircraft or weapons? When the probability of killing a single weapon is P. That's easy!

Pa = P^N

Since P<1, Pa is a decreasing function of the raid size. So now, the question is, what's the probability of killing that weapon, P as a function of salvo size, M and single shot probability of kill Pk < 1. Assuming that it takes one or more hits to knock down the weapon, then

P = 1 - (1 - Pk)^M

All that says is that the probability of scoring at least one hit for a given salvo is one minus the probability of all the missiles missing. Putting both equations together means:

Pa = (1 - (1 - Pk)^M)^N

That's the threshold value for what's is considered to be an acceptable risk of not completely annihilating the raid (in which case there might be leakers), of course we're assuming a single shot opportunity. With very long ranged SAMs that might not necessarily be a valid assumption, but lets say you would prefer not to have to take it even if you have the second opportunity, so you're going to optimize for probability of annihilating on the first shot opportunity. Therefore:

M = (ln(1 - Pa^(1/N)) / ln(1- Pk)

There's your salvo size necessary to get a 0.9 probability of raid annihilation on the first shot opportunity. This is an increasing function of salvo size. So for fixed Pk, and Pa, bigger raids demand more and more missiles to ensure a fixed risk. The problem with this is that there's still not a single "optimum" choice. The solution needs to be constrained.

Warships have fixed magazine sizes, so obviously you can't shoot more and more missiles at larger and larger raids indefinitely. Additionally, they don't want to have to withdraw immediately following a battle because their magazines are empty. Warships are there for a reason, they protect important things and attack important things. Protecting themselves is important, but ultimately if they have to withdraw it constitutes only a tactical victory because their mission can't continue.

So let's assume a warship has W weapons in its magazines. If you're shooting salvos of M weapons at raids of size N, then you're expending M*N weapons. To continue the fight you need J weapons remaining. Let's suppose that you want W - N*M >= J. All that says is that there's at least J weapons left in your magazine to continue the fight with. There's the constraint.
Therefore

(W-J)/N >= M

If you plot both curves and look at the intersection, you can find that mathematical optimum. But look at all the assumptions you have to make! I need to assume a single shot Pk, an acceptable level of risk, and a desired number of weapons remaining to continue my mission. While it's probably reasonable to make some engineering based assumption about Pks versus a given threat. What's necessary to continue my mission or what an acceptable level of risk might be are not necessarily clear. Personally, if it's me taking the hit, I'd say make the risk as small as possible, which is to say, I want J=0 and that gets me back to more is better, basically, with me just exhausting my magazine on every raid that comes in.

If you start putting in reasonable values, though, what you end up with is for fairly substantial raid sizes using an Arleigh Burke sized destroyer, somewhere between 2 and 3 shots to be pretty good. For very large raids sometimes it gets up to 4 missiles per salvo, but that's big raids (7+ incomming). For smaller raids between 1 or 2 shots leaves you with plenty left in the magazine.

[morphin]

I think it could be possible to dynamic adjust WRA for the AI based on mathematical principels. I will sometimes not be 100% right (tooo many varriables), but it would be interesting to developp and programm the code...

Something if the develoopper has nothing else to do [:D]

[SeaQueen]

I doubt it would be any more realistic if you did it that way. Maybe with certain systems like Aegis which can operate nearly autonomously but that's fairly unique, and pre-Aegis systems relied on just operator training (doctrine).