Missiles, base pK, agility and degree of control

[JWNoctis]

First, I wish to say that I'm glad to see this comes to fruitation...Been having a blast!

Though there are still something confusing, particularly about missiles. Here are my observations so far:

First, the current, or shall I say, Harpoon-like system of anti-air pK calculation by (1-CM1)*(1-CM2)*...*(1-CMn)*(BasepK*DistancePenalty-SpeedPenalty-TargetAgility*0.1) felt a bit biased against low-tech missiles firing against low maneuverability targets, and newer missiles, such as AIM-9X or R-77 against highly agile targets, making them only marginally more effective when compared against previous generation ones like AIM-9M and R-27, and less noticeably making low-tech ones barely fit for service since most of them are unlikely to hit their targets, even if they are literally coming in dumb without any ECM, and even though the systems themselves were -presumably- being continuously improved over the years of their service....Though this is just a gut feeling, without any hard evidence.

A global increse in base pK to only take account of the missile's intrinsic inaccuracy and unreliability and an additional multiplier for the target's agility to reflect the missile's own maneuverability during endgame maneuvers, as well as a greater representation of older weapon system's susceptibility to modern ECMs, which -might- have had a much more pronounced and diversified effect in real world might help, or at least further differentiate things a bit. Of course, your mileage may vary.

Also, all aircrafts seemed to be evading every single missile coming at them with the same agility, regardless of any energy bleed-offs during previous evasions. I think this is mentioned in some other threads, and maybe could use some improvement later on.

Then there is an unusual feature and a potential bug, where missiles like R-77 could be manually steered by F3-waypoints inflight just like a wire-guided torpedo, and the UI controls will be locked to waypoint designation mode if the missile self-destructs with designation mode on, which could only be resolved by a save and a reload.

Any thoughts? Can't wait to see more post-v1 improvements!

[ComDev]

Hi JWNoctis, thanks for your feedback [8D]

So... where to start.... Jakob, any chance you could explain the new PoK/PoH model that is being designed?

The F3 waypoint bug is on the to-do list and will be looked into [8D]

[Jakob Wedman]

Thanks JWNoctis! These questions, along with some great ideas, were raised during beta test and we have been working with a new PoK/PoH model for some time (but not top priority [:)]).

The idea is to use a non-linear logistic transformation of the variables that effects the PoK/PoH, like TargetAgility (and SpeedPenalty?) today, but also new like aspect, energy and loadout. The effects would then be added and the sum (X) would determine the final PoK/PoH with 1/(1+EXP(-X)).

If we had enough data of anti-air missile engagements when could easily estimate each effect with logistic regression but as we add more variables we will be forced to make assumptions and test until we have a model that works (better).

What do you think?

[JWNoctis]

Sounded like a good idea...And indeed, didn't realize about those parameters at first. Also that speed penalty is that -50% pK for high supersonic targets -- which sounds like something that will be modeled in another way in the new model, I guess!

This is awesome -- Thanks!

[smudge56]

Ok what's pK?

[Maesphil74]

ORIGINAL: Blighty56

Ok what's pK?

pK = probability of kill

[smudge56]

Cheers that makes sense [:)]

[Tomcat84]

ORIGINAL: Jakob Wedman

Thanks JWNoctis! These questions, along with some great ideas, were raised during beta test and we have been working with a new PoK/PoH model for some time (but not top priority [:)]).

The idea is to use a non-linear logistic transformation of the variables that effects the PoK/PoH, like TargetAgility (and SpeedPenalty?) today, but also new like aspect, energy and loadout. The effects would then be added and the sum (X) would determine the final PoK/PoH with 1/(1+EXP(-X)).

If we had enough data of anti-air missile engagements when could easily estimate each effect with logistic regression but as we add more variables we will be forced to make assumptions and test until we have a model that works (better).

What do you think?

Sounds very interesting. You may remember I brought up that in my opninio agility affects missile PH too much, so this sounds good.

Looking forward to testing this out when it arrives. I'm also interested to discuss all variables you are planning and their effects.