Well, I havent crunched the numbers yet, but here are my observations from memory...
The basic IR missile modeling in DCS, without accounting for countermeasures, exhibited tangible differences in effectiveness depending upon target aspect and Line of Sight Rate (LOSR). Throttle position and airspeed had no observable effect.
When considering countermeasures, preemptive flares were totally ineffective regardless of number or pattern; the weapon system was able to detect the target, achieve target lock, and engage the target without any degradation in time or range.
Reactive flares were effective, especially when employed at a beam aspect with high LOSR. The most effective flare program actually turned out to be one I had programmed for preemptive use, and used a lower dispense rate than other programs. Thus, I does not appear that "mo' flares are mo' better" is universally true.
As for methodology, I setup a simple mission single client A-10 and a single threat system with Excellent AI. I started outside the known threat ring, and configured the CMSP while driving directly at the threat per the conditions I outlined in the previous post.
The first three runs, I used no IRCM in order to establish baseline performance for a nose-aspect engagement. Using TacView, I noted and recorded all of the datapoints i mentioned before. Specifically, i noted the average range of the first shot, and used that number to determine the next part of the test.
I multiplied the observed first-shot Rmax and multiplied that by 1.5 to determine the range at which i should begin premptive IRCM. Then I performed three runs for each of the preemptive flare programs, noting the data as I went. I observed no susceptability to premptive IRCM whatsoever in the DCS SA-13.
Next I flew three runs for each of the reactive programs, waiting until receiving a MWS warning to begin employing IRCM. I continued IRCM until the missile warhead functioned or it was spoofed/missed. I still need to crunch the numbers for these engagements.
I also flew tests that put the missile and launcher on the beam. For these tests, I created a second mission with a client A-10 at Rmax, flying at a 45deg aspect. Imagine placing a diamond inside the threat ring; I started at the bottom corner of the diamond at the southern-most point of the threat ring, and flew northwest to the western-most corner of the threat ring.
Finally for shits n giggles mostly, I flew a nose-aspect engagement, defeated the first missile with a check turn and IRCM, then defeated follow-on beam shots with just IRCM. I cant really add those encounters to the findings, but its baby steps toward a comprehensive threat reaction.
Rinse and repeat for all the altitudes and various conditions you wish to test for, changing only a single variable each time. Once that is done, determine the set of conditions for which you want to calculate probability data. For instance: a rear-hemisphere engagement using two missiles against a non-maneuvering A-10 employing IRCM program X.
Select the data and calculate the launcher's probability of detection, and each missile's probability of guide, probability of hit/function, and probability of causing enough damage to cause the aircraft to crash within 5 minutes. Multiply all of these together to determine the threat system probability of kill.
From there, you can refine IRCM programs, combine IRCM with defensive maneuvers to create a threat reaction against the DCS SA-13, and train pilots how to mitigate the SA-13 threat.
Edit: I just realised that I didn't completely answer your question regarding susceptibilty to number versus pattern of IRCM. I have to run, but I'll come back and answer that fully later tonight.