Redwolf

Well, we are now looking at two different issues:
1) outright killing a tank with direct hits (this page)
2) Pelican Pal's findings that near misses never cause subsystem damage other than tracks
I find #2 much more interesting. Maybe there can be an engine 5 fix for it. PP has made an excellent argument and analysis.
 
 

I don't think anybody in this thread has a problem with how they are spotted after they fire...

Quite right, that was the "first clue". Tests since then have been carried out with the TC vs individual soldier, and have shown the same weird jump in accuracy.

Again, both circumstances are the same weapon, firing directly at the soldier - so both are the AI aiming and firing at a man-sized target.

The end result is that the shots that are targeting the tank commander/gunner are significantly more accurate that those on a target on foot.

CMCW module?
Phillipines 1898.

My understanding is that there is a software component to the tournament system (which is not in CMCW yet).

Sorry, I forgot to answer this question.
10,000 trials for each spotting test - that means loading the test save 10,000 times, calculate the action phase, let run for up to 7 seconds (until the first spotting action) and collect data of which enemy tanks (in which position) were spotted.
1000 trials for the duels.

They are trying to target the individual soldier in both cases (and granted they're not exactly the same, but they're as close as you can get), but that the dispersion from the first is significantly more.

This is clearer with automatic fire, since that will inevitably spread more for all of the obvious reasons. This really looks like a bug in the underlying calculation, rather than something intentional or working as designed.

Where that bug actually is obviously not known (it'd require a look at the algorithm, for a start), I wouldn't take my suggestion of the cause as anything other than baseless speculation, but it's where I'd start looking personally.

Well, yes, but the point here is that SdKfZ 251/1 and M113 gunners have shields - but seem to die as easily as tank commanders who do not.

Well, Pak40 is 120 points, cheap Sherman is 190. So a gun is 2/3rds of a cheap, full tank
More importantly, even if you get one tank out of every AT gun that engages your purchased AT guns do not all engage. Some are out of way of the chosen attack path, some are discovered by infantry before they can shoot at a tank.
As long as we can't boost the gun with better cover and concealment in fortifications I think pricing them more at a factor of 0.4 or maybe even 1/3rd would be in order. It is all about actually getting the historical benefit out of the guns.

I don't mind how easy they are to kill in open ground.
What nags me is how useless foxholes and trenches are in CMx2 for protection, and even have a negative effect on being spotted.

Black Sea badly needs scenarios on maps that are of the scale of Cold War. And Cold War badly needs quick battle maps that are of the size and quality you get into the campaigns.
The QB maps are by far the biggest letdown of both titles.

I plan to test individual infantry at some point, which should give us an answer.
Just don't hold your breath for it - each test takes 3 days and I have a long list to go through!

It's ironic, I've had a Steam account from the beginning and I've played CM for decades, but the idea of combining the two still seems wrong somehow. 

I wondered, it seems to be like this
British - L1A1 to L85A1 -> in service 1985
Germans - G3 to G36 -> in service 1997
Dutch - FAL to Diemaco C7 -> selected for service 1994
 

The gun shields do help if you use them correctly, but the main issue is that the accuracy has been shown to increase dramatically when targeting gunners/tank commanders.

This has all been tested, logged, etc. The data is all there. Shots versus tank commanders are more accurate than against a single individual in the open - the cone of fire is significantly tighter.

This isn't so much about crews not buttoning up, it's about tests indicating that pixeltruppen are far more accurate when shooting at unbuttoned crewmembers,

I think there is. You put a single trooper in the open, and see how long it takes a squad to kill them from a given range.
It may not be a perfect test, but it is indicative of the underlying problem, especially when you consider that 1 man in the open is roughly twice as big as a head and torso showing over a vehicle hatch, especially if there is a turret shield involved.

So there have been tests. The first clue was some forum posts from here years ago that showed SMG tracers firing at the unbuttoned tank commander, or at the vehicle. The vehicle had tracer fire going all over the action spot, and the unbuttoned one had a focused, tight cone going straight for the TC. The same weapon firing at the same range, but suddenly much more accurate?

The main methodology to generate the data was to set up firing ranges with unbuttoned tcs/halftracks and individual HQ soldiers, at the same distance and on sand. I don't have the test results to hand (can dig this up), but it showed a marked increase of accuracy towards the commanders, over the soldiers on foot. Same weapon, same distance, statistically significantly more accurate fire.

Yeah. If the real world was like CM's fortifications then nobody would have bothered developing direct fire larger than 75mm.
A good exercise would be to improve heavy cover to the point where (inside CM) you scream for 150mm HE.

It is also a matter of fitting the platoon into 3, not 4, BMPs.

It is also a matter of fitting the platoon into 3, not 4, BMPs.

the testing range...

Hopefully we get a 2S1 or somesuch in the next module.