The one minute gun range survivability test

[Pak40]

Some thoughts and questions come to mind regarding the Jeep vs the HT gunner:

In CM, you can only target the vehicle, not the driver or any other exposed crew member. So how exactly does CM handle targeting when there is an exposed crew member? Is there some sort of decision making behind the scene?

Also, does CM treat soft targets with exposed crew members differently than hard targets with an unbuttoned tank commander/gunner? The results of the test seem to suggest that the jeep driver, although just as exposed as the half track gunner, isn't being specifically targeted but rather the jeep itself.

[Bulletpoint]

52 minutes ago, Pak40 said:

The results of the test seem to suggest that the jeep driver, although just as exposed as the half track gunner, isn't being specifically targeted but rather the jeep itself.

Good point. I think the rationale is that they know they can destroy the jeep with small arms fire, so they target the vehicle instead of just the crewmember. The same goes for trucks, I think. That's likely why trucks can often keep driving around even after taking quite a lot of fire, where in real life, a burst of automatic fire to the windshield would stop the vehicle very fast.

[rocketman]

1 hour ago, Pak40 said:

Some thoughts and questions come to mind regarding the Jeep vs the HT gunner:

In CM, you can only target the vehicle, not the driver or any other exposed crew member. So how exactly does CM handle targeting when there is an exposed crew member? Is there some sort of decision making behind the scene?

Also, does CM treat soft targets with exposed crew members differently than hard targets with an unbuttoned tank commander/gunner? The results of the test seem to suggest that the jeep driver, although just as exposed as the half track gunner, isn't being specifically targeted but rather the jeep itself.

I was about to put these ideas forward as well. In my test, does the TacAI see two targets: HT and gunner and picks one? Would it be different if I set a manual Target on the HT? Might have to check for that as well.

Also, is there some reason as to where the TacAI aims at a target - always at its center?

[A Canadian Cat]

21 hours ago, rocketman said:

I was about to put these ideas forward as well. In my test, does the TacAI see two targets: HT and gunner and picks one?

Yes, I believe it does at least in one sense.  In the sense that the gunner manning the gun is seen as a greater threat than the HT itself.  The TacAI knows that the gunner is the danger and so targets him.

 

21 hours ago, rocketman said:

Would it be different if I set a manual Target on the HT? Might have to check for that as well.

No it would not change anything because in this sense there is only one target.  If you tell the TacAI "target that HT" it will fire at the HT vehicle if the gunner is buttoned up but if the gunner is up it will see the gunner as the threat and target him.  The decision is all about what you ordered the TacAI to do and what the threat of the target is.  So, when you tell the TacAI to target the HT it will but it will also decide that an active gunner is the biggest threat and specifically target him.  In this situation the TacAI does not even have to counter man your orders because the gunner is part of the HT.  If there were two HTs, one buttoned and one with an active gunner and you told the TacAI to target the buttoned HT it would have a decision to make: obey your order or break from your order to address the greater threat.  That decision could vary soldier by soldier, the willingness to break your order will depend on the motivation of the troops (fanatic troops will likely not break from your orders), it will also depend on the experience of the troops (crack troops will want to address the greater threat more than keep to your orders). 

As you can see there is a conflict there and you can sometimes end up with rather perverse situations where your crack, fanatic panther crew will continue to fire HE at the now empty building that I ordered them to shoot at and ignore the Sherman tank coming at them from the side.  That is why using direct target orders as little as possible is a good thing (and the main reason I love target briefly).  My crack / fanatic Panther should have been like no other team on the battlefield at foiling silly Sherman tanks trying to sneak around but because I gave them a direct target orders they lost out.

Story exaggerated to make my point (it was only a veteran highly motivated Panther).

 

21 hours ago, rocketman said:

Also, is there some reason as to where the TacAI aims at a target - always at its center?

Yes the centre but with caveats - they vary their aim slightly with each shot, if shots hit obstacles (e.g trees) they re aim away from the obstacle and troops can switch between targeting the vehicle and exposed crew of said vehicle as necessary (which probably means that the poor crewman becomes the target - at least for small arms - tank main gunners do not do this with the main gun at least).

[Bulletpoint]

52 minutes ago, IanL said:

My crack / fanatic Panther should have been like no other team on the battlefield at foiling silly Sherman tanks trying to sneak around but because I gave them a direct target orders they lost out.

Story exaggerated to make my point (it was only a veteran highly motivated Panther).

I think this is a bit of a problem with the game actually, because being a highly motivated soldier doesn't mean you're also stupid. Yes, it was a direct order to fire at the house, but that should not mean you just let another tank drive up to you and destroy you. And this even happens with "regular" motivation sometimes. I don't even think history's fanatic troops fought like that.

In my view, motivation should determine how long troops run before they take cover, and how much fire they can tolerate before falling back, but they should still use common sense against targets of opportunity - especially against those that are mortal threats to them.

In the end, we're never going to see a TacAI that always reacts perfectly believably to all game situations, but I think it oculd be improved about the area fire.

(Speaking of TacAI, yesterday I had a panther turn around to face a bazooka contact marker 160 metres away, while at the same time turning its back to a Sherman that it had direct visual contact with - the enemy tank should have been seen as the bigger threat)

Edited September 2, 2016 by Bulletpoint

[rocketman]

Thanks Ian for an insight into the inner workings of the game. So, if I understand correctly, a unit will start to aim at the center and then gradually adjust aim which seems perfectly well and realistic. But does that mean as a unit or per soldier? That is, a single sniper will much sooner adjust his aim and hit its target (let's say afte 10 shots) rather than a five man team firing two shots each - everything else equal?

[A Canadian Cat]

1 hour ago, rocketman said:

Thanks Ian for an insight into the inner workings of the game. So, if I understand correctly, a unit will start to aim at the center and then gradually adjust aim which seems perfectly well and realistic.

Actually I didn't mean correct their aim over time - but they do do that too if they miss.  I meant that each subsequent shot that *hits* will not hit the exact same place every time.  The game simulates scatter between shots that hit.

1 hour ago, rocketman said:

But does that mean as a unit or per soldier? That is, a single sniper will much sooner adjust his aim and hit its target (let's say afte 10 shots) rather than a five man team firing two shots each - everything else equal?

This kind of stuff is per solider.  And they are aiming at something; could be a vehicle, could be a particular enemy soldier. 

[A Canadian Cat]

1 hour ago, Bulletpoint said:

I think this is a bit of a problem with the game actually, because being a highly motivated soldier doesn't mean you're also stupid. Yes, it was a direct order to fire at the house, but that should not mean you just let another tank drive up to you and destroy you. And this even happens with "regular" motivation sometimes. I don't even think history's fanatic troops fought like that.

Yeah you could make a successful argument that this is not optimal - I will not disagree.  In my situation it is pretty clear but if you switch it up and now its a elite fanatic scout team with a tight cover arc: is it right for them to ignore your order and expose their position the deal with a threat outside their cover arc?  Probably not.  The TacAI is trying to simulate soldiers but it is a computer so has to use some rules.  For us humans to be able to deal with it we have to know what those rules are and how it behaves.  So, BFC cannot program the TacAI to "do the right thing as I see it".  It is just impossible to do: so we have what we have.  I personally believe on balance the behaviour is good.  What I meant by that is, averaged out, the behaviour of the TacAI under combat is very reasonable. 

Even the Panther doing the wrong thing is OK really - their Platoon CO had them lay some HE fire down for the infantry they were supporting and because they were doing the job they asked they missed spotting the Sherman until it was too late.  It *is* possible to sneak up on a Panther and KO it on the battlefield.  So, it *is* OK to see it happen in game.  Just because I might think that they should have been smart enough to do the right thing doesn't mean that's what a real crew would have done.  A real crew could have been surprised and not been fast enough too.  And the real failure might have been the platoon leader who should have had another tank in the platoon watching that flank.  Just like the game commander (i.e. me) should have probably done the same or at least used target briefly to fire at the house so the crew would have been more alert to other threats at least most of the turn.
 

1 hour ago, Bulletpoint said:

In the end, we're never going to see a TacAI that always reacts perfectly believably to all game situations, but I think it oculd be improved about the area fire.

Actually, I think the quote should be "In the end, we're never going to see a TacAI that always reacts perfectly to all situations and that is a good thing."

What I am saying is that looking at the specific details of one situation and "fixing" the TacAI and repeatedly doing that will lead to two possible outcomes: uber soldiers (and we will still not be happy because the other guys uber soldiers killed my uber soldiers) or a mess (and we still will not be happy because we will have chaos but be unable to figure out the rules).

There are improvements I would suggest for the TacAI but this detailed level "make them behave perfectly" is not it.

[Bulletpoint]

6 minutes ago, IanL said:

There are improvements I would suggest for the TacAI but this detailed level "make them behave perfectly" is not it.

I never said anything about making the troops behave perfectly. I said I would like them to behave "believably". There's a world of difference.

Even your elite scout team would try to fire in self defense when they started taking fire from a close by flanking team. Maybe the craziest hitlerites in history would have let themselves get shot, but even that is a big "maybe". I think their fanatiscism had more do do with refusing to give up than to sticking to orders to the letter.

About your panther doing area fire, of course it would be possible to flank it without the crew noticing it. I was talking about whether the crew would react IF they noticed that there was a Sherman creeping up on their flank quite close by. Again, I believe they would have excused themselves from their suppression duty long enough to engage the Sherman, then returned to bombardment job once the Sherman was taken care of.

 

[A Canadian Cat]

1 hour ago, Bulletpoint said:

I never said anything about making the troops behave perfectly. I said I would like them to behave "believably". There's a world of difference.

LOL Yeah I know but that's all I hear when people say they want this or that tweaked for fix some specific situation .  We should probably stop since @rocketman is investigating a real issue and our philosophical discussion is just distracting from that.

Quote

Even your elite scout team would try to fire in self defense when they started taking fire from a close by flanking team.

No I set the circular covered arc to indicate what "close by" meant.  They should respect it.  Otherwise what's the point of having that command?  I'm being serious now.  If orders are not going to be obeyed the majority of the time we give them then we have a different problem.

 

Quote

About your panther doing area fire, of course it would be possible to flank it without the crew noticing it. I was talking about whether the crew would react IF they noticed that there was a Sherman creeping up on their flank quite close by. Again, I believe they would have excused themselves from their suppression duty long enough to engage the Sherman, then returned to bombardment job once the Sherman was taken care of.

OK so let say the platoon LT did not screw up and there is a tank protecting their flank.  What then?  Now there is someone who's job it is to deal with that Sherman.  So the crew should continue their task as ordered now right?  Careful how you answer that because in my example I said the building was empty but how does the tank crew really know that?  They were ordered to fire at that building.  It could be there is an enemy MG in there that is holding up the advance and the entire attack rests on them doing this one job.  After all they were ordered to do it.  And that's what orders are for  success.  My point is just that a, normal, possible, outcome that could have happened in real life did happen in the game.  So, we should probably not be fussing about it because the TacAI is never going to have information about how important any particular orders are, what other units are up to etc.  What we have now is pretty clear and already leads to believable results.  In my opinion. Its all good-ish.

Edited September 2, 2016 by IanL

[Bulletpoint]

21 minutes ago, IanL said:

@rocketman is investigating a real issue and our philosophical discussion is just distracting from that.

Yes, you're right. Sorry for derailing the thread.

Edited September 2, 2016 by Bulletpoint

[rocketman]

Ok back on topic then

I decided to change the test somewhat to try what I above called the "HT hit bonus" idea. Ian's comment above that the HT and gunner are one and the same target, but when the enemy spots the gunner they direct their fire at him - made me think that the HT being a large target makes incoming fire adjust almost too well, compared to a man alone in the open. So I let the HT be stationary at 280 m and a sniper team reduced to 50% (one man) walk (so not to lie down if stationary) and eventually he starts running (can't be avoided). So he gets a bit closer than the HT but for now that has to do. This time I didn't count number of bullets to kill or time to kill, but rather how close the shots were to each target. It is not easy to capture screenshots of tracers but it got good enough. I edited in Photoshop (not perfectly good) and put tracers into one picture. The camera was behind the unit all the time but the sniper sometimes strayed off the path.

Here are the pics:

To me it seems like incoming fire towards the HT gunner is considerably more concentrated and "on target" than the ones aimed at the sniper. So there seems to be some kind of "HT hit bonus" in effect vs the gunner. And yes, I know it is stationary, but I saw much the same pattern in earlier tests when it moved.

[A Canadian Cat]

Oh that is interesting. To me that looks like the HT gunner is being aimed at while the sniper is not but instead the team is more area firing like their NCO thinks there is a team there.

Note I'm not really trying to say the game is doing that just that it looks like that.

Edited September 2, 2016 by IanL

[c3k]

Tracer screenshots are very informative. Nicely done.

[rocketman]

Consider also that the tracer screenie is from four rifles during less than a turn and not all traces are in the pic. Imagine then in a real battle situation, the gunner unbuttons and all of a sudden becomes the priority target for several squads, some at closer range, with automatic weapons, from several angles - no wonder the gunner doesn't last long

Next test will be to put the HT sideways as that would make it an even bigger target. Let's see if that changes anything. Then I'll probably have a look at a german HT with gun shield. But that will be later today/tomorrow.

[Erwin]

It will be v interesting to see what effect/advantage the german gun shield is re frontal fire vs defilade.  Thanks for doing this.

[Bulletpoint]

On 3/9/2016 at 0:35 AM, rocketman said:

Ok back on topic then

I decided to change the test somewhat to try what I above called the "HT hit bonus" idea. Ian's comment above that the HT and gunner are one and the same target, but when the enemy spots the gunner they direct their fire at him - made me think that the HT being a large target makes incoming fire adjust almost too well, compared to a man alone in the open. So I let the HT be stationary at 280 m and a sniper team reduced to 50% (one man) walk (so not to lie down if stationary) and eventually he starts running (can't be avoided). So he gets a bit closer than the HT but for now that has to do. This time I didn't count number of bullets to kill or time to kill, but rather how close the shots were to each target. It is not easy to capture screenshots of tracers but it got good enough. I edited in Photoshop (not perfectly good) and put tracers into one picture. The camera was behind the unit all the time but the sniper sometimes strayed off the path.

Here are the pics:

To me it seems like incoming fire towards the HT gunner is considerably more concentrated and "on target" than the ones aimed at the sniper. So there seems to be some kind of "HT hit bonus" in effect vs the gunner. And yes, I know it is stationary, but I saw much the same pattern in earlier tests when it moved.

This is very interesting. Maybe halftrack gunner vulnerability will finally be solved?

Edited September 4, 2016 by Bulletpoint

[rocketman]

First preliminary test with German HTs show that at 300 m vs rifles the gun shield is a life saver. The same concentrated fire as in screenshots above was noted, lots of them hitting the shield and harmlessly bounced off. Out of 4 HTs, no gunner was killed during 1 min of fire.

[cool breeze]

While the tracer fire does look strange coming at the mg gunner like that, in the context of CM, because we are used to seeing it more like the tracer fire being shot at the walking guys; I propose that its actually pretty reasonable.  Part of why the regular aim of soldiers in CM seems bad is because its partly simulating that the targets are benefiting from a lot of micro terrain and micro movements that aren't ... um shown... for lack of articulation.  The game assumes there is more stuff blocking los, and more work is done by the soldiers to hide behind it than the engine can show/ simulate so it gets fudge factored.  But the mg gunner on the stationary halftrack has none of that stuff going for him.  While the engine never tries to give us shooting range performance, shooting at the mg gunner is the closest to a shooting range target.  And honestly to me the grouping for tracers in the pic seems like reasonable for a shooting range target under combat level stress environment.  If it wasn't for combat stress they should be like half or more hits, right?

[rocketman]

1 hour ago, cool breeze said:

Part of why the regular aim of soldiers in CM seems bad is because its partly simulating that the targets are benefiting from a lot of micro terrain and micro movements that aren't ... um shown... for lack of articulation.  The game assumes there is more stuff blocking los, and more work is done by the soldiers to hide behind it than the engine can show/ simulate so it gets fudge factored.

If that is the case with the game engine, I'd like BFC to confirm that. Otherwise it is an assumption I don't really believe in. The map is completely flat, but the engine simulates undulations of 1 m over long distances so that isn't the fact here. We can place bushes and even brush on the map, none of which exists here so there shouldn't be any "invisible ones". If there was foilage I'd agree to a degree, without a single tree nearby I tend to disagree. That a soldier that walk and then run would pause to take cover every now and then, then surely they wouldn't advance at a constant speed. And also it is my impression that the engine does simulate bullit hits by the pixel, so if what we see is different than what counts, I'm confused...

But I might be completely wrong, but in that case I have the wrong view of the game engine and what goes on under the hood.

[rocketman]

1 hour ago, cool breeze said:

But the mg gunner on the stationary halftrack has none of that stuff going for him.  While the engine never tries to give us shooting range performance, shooting at the mg gunner is the closest to a shooting range target.  And honestly to me the grouping for tracers in the pic seems like reasonable for a shooting range target under combat level stress environment.  If it wasn't for combat stress they should be like half or more hits, right?

The upper body of a HT gunner is about a 0,5 m2 target. At almost 300 m, was a regular soldier in WWII with a rifle without a scope able to hit that consistently? Anyone have data or source to confirm or dismiss?

[JoMac]

Ok, this HT 'Bullet Magnet' & HT Gunner Survivability issue has been around since CMBN...The Laser like Hail of Small Arms Precision along with the increased Firepower patch since CMBN (especially automatics) has simply made the HT too Unsurvivable. 

No Scientific Data here, but reasonably at 250 meters away with Small Arms against even a Vehicle Size Target should be around 50% and say 25% if aiming at top half of vehicle, then try hitting a man size target on the top who is considered Hull Down...I venture to say a 'Lucky Shot' is needed to get a hit.                                                                  

Snipers at the above Range or Close Range Fire of say 100 meters should result in more consistent HT Gunner/Crew Hits.

I don't know what's needed, but me thinks it should take alittle longer to acquire that Gunner/Crew (and not this instant spotting of Crew who are hull down), less overall Firepower (more Pop Shots vs. Aimed Shots), or maybe a better 'Savings Roll' to represent Gunners/Crew ducking in and out of cover when fired upon (not visually simulated), and more quicker Buttoned up of these Vehicles.  

Don't get me wrong, but I think unbuttoned Units are still a priority target and should get hit fairly regularly, but not at the pace CM portrays it now. 

Edited September 6, 2016 by JoMc67

[rocketman]

I just took screenshots from the firing unit's perspective, not that it fully represents the 3d depth perception of a real human but should give an impression of just how large a target they aim at. Will post those tomorrow.

[cool breeze]

2 hours ago, rocketman said:

If that is the case with the game engine, I'd like BFC to confirm that. Otherwise it is an assumption I don't really believe in. The map is completely flat, but the engine simulates undulations of 1 m over long distances so that isn't the fact here. We can place bushes and even brush on the map, none of which exists here so there shouldn't be any "invisible ones". If there was foilage I'd agree to a degree, without a single tree nearby I tend to disagree. That a soldier that walk and then run would pause to take cover every now and then, then surely they wouldn't advance at a constant speed. And also it is my impression that the engine does simulate bullit hits by the pixel, so if what we see is different than what counts, I'm confused...

But I might be completely wrong, but in that case I have the wrong view of the game engine and what goes on under the hood.

 

I think the regular short grass ground cover has built in small undulations.  Playing the game I notice that on paved surfaces my pTruppen die more, and I dont think it just ricochets.  Plus the game isnt designed to simulate the sort of test situations we devise, the fudge factors have all been adjusted with actual game play on regular maps not for test set ups. 

 

 

2 hours ago, rocketman said:

The upper body of a HT gunner is about a 0,5 m2 target. At almost 300 m, was a regular soldier in WWII with a rifle without a scope able to hit that consistently? Anyone have data or source to confirm or dismiss?

  There was no real source for that, I just pulled the number out of my butt.  After I sent it I kinda wished I had used a less questionable made up number.  I'm totally unqualified, having only had one shooting session with a rifle, and it had a scope and bipod, so didn't count  ( but I did nail it, am a natural shooter   ) .  But I'm also vaguely referencing JasonC and his unscoped rifle range condition numbers hes thrown out a few times.  Training was less back then than now but firearms proficiency was more widespread, at least in the US. Marines are supposed to all be able to pretty reliably hit a man size target at 500 yards on the range. 

 

 

[cool breeze]

If the half track is at 700 or more yards the rifles wont even shoot but the 50 cal will work fine.