AFM Missile PK Test Thead

[Pyroflash]

Okay, so I am setting up a series of standardized tests using particular setups which are 1v1 only, usually with only one missile launch per test. The idea is to gather data on the Pk of these missiles using a controlled environment with particular parameters to avoid any bias as to the behavior of missiles. Track files/ACMIs are appreciated, but results of the test(s) that are performed will be generally satisfactory for this. The tests will be divided into two sets, offensive, and defensive. These should be self explanatory, and relatively easy to set up and perform.

 

General setup is one OPFOR aircraft, either F-15C is play testing the Su-27/MiG-29S, or Su-27 when play testing the F-15C. The single bandit aircraft will be set at 400 knts, at 60 nm (using standard nautical mile of 6080'), and 25% fuel for all tests. Defensive bandits will NEVER have any guns, tanks, or missiles. The player aircraft will ONLY ever carry 25% fuel. Each test however, will have different parameters in parentheses, so pay close attention to these.

 

When posting results, post the results of individual tests, and if the missile does not hit, list whether it was because it failed to continue tracking and went dumb, or because it lacked the energy to intercept the target.

 

[AIM-120B]

 

AIM-120B, Offensive:

 

Start at 40,000' in an F-15C, climb to 50,000', level off, and accelerate to Mach 1.2. Fox 3 at 40 miles using STT. Maintain valid lock through missile flight.

 

AIM-120B, Offensive:

 

Start at 40,000' in an F-15C, climb to 50,000', level off, and accelerate to Mach 1.2. Fox 3 at 30 miles using STT. Maintain valid lock through missile flight.

 

AIM-120B, Offensive:

 

Start at 40,000' in an F-15C, level off, and accelerate to Mach 1.0. Fox 3 at 20 miles using STT. Maintain valid lock through missile flight.

 

AIM-120B, Offensive:

 

Start at 20,000' in an F-15C, level off, and accelerate to Mach 1.0. Fox 3 at 20 miles using STT. release lock at 10 miles

 

AIM-120B, Offensive:

 

Start at 20,000' in an F-15C, level off, and accelerate to Mach 1.0. Fox 3 at 20 miles maddog.

 

[AIM-120C]

 

AIM-120C, Offensive (bandit 40,000 @ 180 aspect):

 

Start at 40,000' in an F-15C, climb to 50,000', level off, and accelerate to Mach 1.2. Fox 3 at 40 miles using STT. Maintain valid lock through missile flight.

 

AIM-120C, Offensive:

 

Start at 40,000' in an F-15C, climb to 50,000', level off, and accelerate to Mach 1.2. Fox 3 at 30 miles using STT. Maintain valid lock through missile flight.

 

AIM-120C, Offensive (red air at 20,000 traveling directly towards you, default options):

 

Start at 40,000' in an F-15C, level off, and accelerate to Mach 1.0. Fox 3 at 20 miles using STT. Maintain valid lock through missile flight.

 

AIM-120C, Offensive (red air at 20,000 traveling directly towards you, default options):

 

Start at 20,000' in an F-15C, level off, and accelerate to Mach 1.0. Fox 3 at 20 miles using STT. release lock at 10 miles

 

AIM-120C, Offensive (red air at 20,000 traveling directly towards you, default options):

 

Start at 20,000' in an F-15C, level off, and accelerate to Mach 1.0. Fox 3 at 20 miles maddog.

 

[R-77]

 

R-77, Offensive (red air at 20,000 traveling directly towards you, default options):

 

Start at 40,000' in a MiG-29S, climb to 50,000', level off, and accelerate to Mach 1.2. Fox 3 at 40 miles using STT. Maintain valid lock through missile flight.

 

R-77, Offensive (red air at 20,000 traveling directly towards you, default options):

 

Start at 40,000' in a MiG-29S, climb to 50,000', level off, and accelerate to Mach 1.2. Fox 3 at 30 miles using STT. Maintain valid lock through missile flight.

 

R-77, Offensive (red air at 20,000 traveling directly towards you, default options):

 

Start at 40,000' in a MiG-29S, level off, and accelerate to Mach 1.0. Fox 3 at 20 miles using STT. Maintain valid lock through missile flight.

 

R-77, Offensive (red air at 20,000 traveling directly towards you, default options):

 

Start at 20,000' in a MiG-29S, level off, and accelerate to Mach 1.0. Fox 3 at 20 miles using STT. release lock at 10 miles

 

R-77, Offensive (red air at 20,000 traveling directly towards you, default options):

 

Start at 20,000' in a MiG-29S, level off, and accelerate to Mach 1.0. Fox 3 at 20 miles maddog.

 

defensive tests and semi-actives will come later, probably late Sunday, or manday (yes, manday, you know, for MEN) when I get back and have time. The semi-active tests will be different by nature, and won't be the exact same like the active tests are. I will also be adding tests where the bandit is running away at missile launch, to check PK on an aft aspect launch.

Edited January 21, 2013 by Pyroflash

[falcon_120]

Although maybe a good idea, right now ,with the game in beta phase and in constant

evolution all this talk about missiles is getting a bit crazy. We should wait till some more

updates are done and we are closer to the actual realease version.

[theChris]

i think the idea of this is to show information maybe the beta testers havn't found like every detail about these missiles to help the updates come quicker so we can plaay

[Kuky]

Not good to let each person create such scenarios, they will vary no matter how accurate they try to make it.

 

What you need to do is create mission for each setup (speed,alt,aspect) and let AI fly both target and attacking aircraft... only problem with this is AI will always launch at R-max. Workaround for this could be let player fly the aircraft BUT each player must set Autopilot straight away after starting mission in order to keep same parameters and only need to turn on radar and lock target and not change his course/speed/alt and only fire the missile at certain range (and note this range)

[arteedecco]

Or.... make "official" test missions with very clear parameters like Pyroflash did ... maybe a youtube vid showing it demo'd with callouts for critical test card parameters.

 

Then... have ppl fly in single player and provide tracks. Then consume all data collected from tracks and average the results.

 

Tangentially... I think the point of Pyroflash's tests is to gain empirical experience / evidence on the way the missiles are working to provide "user feedback" in a sort of standardized format. Simply shooting for AI replications is not a good test either IMO, since nobody wants to watch AI bots fly a sim :) It is the human valuation of the missile performance that matters. Some... with more xp and good arguments should have more input, but throwing out human user testing with the bath water would also be a mistake.

 

"Snipe"

[Kuky]

Nothing to do with "human" evaluation, it either has to be done in scientific approach or it's pointless. People complain 120 is much better than 77, and only way to test this is have exact same conditions for both missiles.

 

Human opinion is useless, numbers and correct approach of testing matters

[Pyroflash]

Nothing to do with "human" evaluation, it either has to be done in scientific approach or it's pointless. People complain 120 is much better than 77, and only way to test this is have exact same conditions for both missiles.

 

Human opinion is useless, numbers and correct approach of testing matters

 

That's what I am trying to accomplish, and you are right that I should probably create a mission set, however I am not simply going to run this with AI as the test would then be next to useless as the AI can't manage the aspects like a player can.

 

To be clear, the purpose of the test is to clear up, or bring to light any problems with the missiles doing unusual things like not managing energy, doing hard u-turns, and generally not flying in-envelope. It is also to get Pk numbers for different missiles under controlled circumstances against targets that are maneuvering at the SAME skill level each time to try and shed some light on the validity of some of the complaints that one missile has some distinct advantage over another.

 

Sadly, the missions are going to have to wait until I get back on Monday. My schedule isn't exactly gaming friendly ATM.

[Grigs]

Wouldn't it be even more accurate to use the same target?

[Kuky]

That's what I am trying to accomplish, and you are right that I should probably create a mission set, however I am not simply going to run this with AI as the test would then be next to useless as the AI can't manage the aspects like a player can.

 

AI doesn't have to manage aspects, that should be set by mission design. You need to first create mission with AI target at certain altitude, speed, have second WP far away to keep going in straight line, and how you want it to respond when being fired upon (do you want it to drop c/m's do you want it to evade).

 

Next add human player aircraft and set it at head on, pick altitue etc so you create scenario 1. Then create another mission based of 1st one and only change initial position of human plane to create new aspect (for example 30° ofset left or right), then repeat again and again each time changing aspect, always leaving AI target stationary... and you will create multiple missions with different aspect scenarios. Once this is done, change player plane (to test different weapon) and reuse all the missions you created already and only change player plane time... this way you create duplicate scenarios.

 

Once you have all the missions, player enters his plane and as mission starts he needs to set to auto pilot (to keep same speed, altitude, heading) and only need to turn radar on and fire missile at specified range... the rest is up to missile... once fired under exact same conditions and under same aspect, then you can compare how each missile performs.

 

PS: every time you test something its important to leave all other variables constant (speed, alt, range etc which ever you are not testing) and only change 1 thing at a time... this is how you test properly

Edited January 12, 2013 by Kuky

[Pyroflash]

AI doesn't have to manage aspects, that should be set by mission design. You need to first create mission with AI target at certain altitude, speed, have second WP far away to keep going in straight line, and how you want it to respond when being fired upon (do you want it to drop c/m's do you want it to evade).

 

Next add human player aircraft and set it at head on, pick altitue etc so you create scenario 1. Then create another mission based of 1st one and only change initial position of human plane to create new aspect (for example 30° ofset left or right), then repeat again and again each time changing aspect, always leaving AI target stationary... and you will create multiple missions with different aspect scenarios. Once this is done, change player plane (to test different weapon) and reuse all the missions you created already and only change player plane time... this way you create duplicate scenarios.

 

Once you have all the missions, player enters his plane and as mission starts he needs to set to auto pilot (to keep same speed, altitude, heading) and only need to turn radar on and fire missile at specified range... the rest is up to missile... once fired under exact same conditions and under same aspect, then you can compare how each missile performs.

 

PS: every time you test something its important to leave all other variables constant (speed, alt, range etc which ever you are not testing) and only change 1 thing at a time... this is how you test properly

 

I think that you misunderstood me somewhere along the line :D

 

When I said aspects, I wasn't referring to aspect angles, but instead the management of certain parameters such as a specific launch distance, speed, angle, etc. that an AI cannot perform, while a player can. The idea was to have the player test against the AI, and that an AI vs. AI test would be next to meaningless because the AI can't adhere to specific fixed parameters that would prove the validity of such a test.

[Kuky]

Yes, I agree AI vs AI works for R-max testing only, I said that earlier in one of the previous posts :)

 

So, if you want to test human vs AI target (but human should switch to auto-pilot), you should create missions with setups like I described, or any other setup depending what you want to test (hi->low, low->high, beaming, head-on, receding, etc)