Small arms casualties - tank crews

[Rokko]

26 minutes ago, Battlefront.com said:

Yes, accuracy decreases, but chances of hitting increases overall because more lead flying downrange is more lead to hit with.  Not proportional per shot, but more than not shooting at all.  As for the RoF issue, see previous comment about the TacAI controlling RoF by determining how many guys are shooting, now how much each guy shoots in a given period of time.  Now that I think about it I'm pretty sure that's correct. Which is why having 1 or 2 guys is definitely better than having a dozen for this sort of test.

I am sorry but how is that not totally controlled for by only counting the shots it took? Doesn't matter how long it took, doesn't matter how many people shot, etc. Especially with the new test where it's only a single guy.

26 minutes ago, Battlefront.com said:

Since you can't control the TacAI behavior you have to adapt what you are testing to the circumstances you do not have control over.  This gets to "what are you trying to test for".  If you are trying to test for "how many shots does it take to get a state change of the target" then it works fine.  If you are trying to figure out how many shots it takes to get a KIA/WIA result then it doesn't work because of the reasons I stated.  And your subsequent test, more-or-less along the lines I suggested, confirms that.

Does confirm what? Do you mean it is equally flawed in your estimation?

26 minutes ago, Battlefront.com said:

Have you shot at things at even 100m distance?  I can assure you even a moderately competent shooter can get a tighter hit grouping on a shield than a Human silhouette.  It has to do with how well your brain is sure of what it is shooting at.  This is why soldiers do not wear bright colored clothing and instead wear camouflage or dark single colors.  The easier it is to see a target the easier it is to hit a target. And a large piece of vertical metal is a lot easier to see than a dark colored Human figure at 200m.

No I haven't btw, not that it matters. The real question is, how does the game handle aiming? In my understanding it is mainly based on center of visible mass. Camouflage and whatnot is not at all taken into account IIRC. And please take a look at the pictures I've posted, I don't see how anyone could get to the conclusion the prone infantry man would be easier to hit.

26 minutes ago, Battlefront.com said:

Yup, I understood that is what you did.  However, Experience/Motivation/Leadership does make a difference when being shot at because it determines how likely the soldier is to duck and how quickly it will expose itself to more danger.  Which may have an impact on perception when you're trying to compare it to actual in game results.  Which is generally why I use Regular because Regular is the most common situation in an actual game setting.

But thats exactly the (suicidal) behaviour one has to replicate at the moment, because...

26 minutes ago, Battlefront.com said:

Yup, but unfortunately it also makes the crew more likely to stay upright.  Either way, keeping the Experience/Motivation/Leadership stuff the same for all Targets is generally a good idea.

...no it plainly doesn't. A halftrack gunner will NEVER EVER refrain from staying upright. Only by being told so or by dying will he leave his post, doesn't matter what experience or motivation he has. The only exception I know if is ratteling while the vehicle is immobilised, in that case I've seen crews abandon their vehicle entirely (and getting hit while doing so).

[A Canadian Cat]

39 minutes ago, sttp said:

A guy's 100 meters away from you in this flat and empty field of short grass.

He yells over and says "hey, I'm gonna start firing my 30 calibre rifle at you. So you'd better either get into that halftrack, behind the gun, or you can just stay out there in the open in whatever position you choose. Your choice. But either way, I'm gonna fire once every two seconds."

Which option do you take?

I'd just want to go home

[JoMac]

Now, Rokko, Can We All Learn to Get along...

I do know that a HT Gunner in a standard game will button down, then come back up from time to time in a turn, depending on situation, etc.

But, your right thou, in your tests the HT Gunner is at Max Moral, Leadership & Motivation, and so that does mean the HT Gunner will die in place rather then duck down more often like he should. 

Which, I hope Steve will look into...He did mention something about it (possible ducking up and down more quickly depending on threat levels, etc).

On top of that, I think that even if more shots are targeted at a Vehicle (because its a bigger target to see, etc) the Gun Shield should provide much better protection then it does now ( Steve, also mentioned he will look into ).

Edited April 12, 2016 by JoMc67

[Rokko]

3 minutes ago, JoMc67 said:

Doesn't the HT Gunner in a standard game, button down, then come back up from time to time in a turn ?...

Thou, in your tests the HT Gunner is at Max Moral, Leadership & Motivation, and so that does mean the HT Gunner will die in place rather then duck down more often like he should.

In my experience HT gunners never button up due to being suppressed whatever their experience/morale settings are, just like StuG MG gunners don't. The reasoning for this has explained earlier in this thread when it wasn't at all about halftracks, I just now realized I've kinda hijacked this thread 

But as I have explained, my hypothesis is that those issues are connected.

3 minutes ago, JoMc67 said:

Now, Rokko, Can We All Learn to Get along...

Have I come at any point during this discussion across as unkind or agitated? If so, I wasn't aware and apologize.

[Battlefront.com]

52 minutes ago, sttp said:

A guy's 100 meters away from you in this flat and empty field of short grass.

He yells over and says "hey, I'm gonna start firing my 30 calibre rifle at you. For 60 seconds. So you'd better either get into that halftrack, behind the gun, or you can just stay out there in the open in whatever position you choose. Your choice. Either way, I'm gonna fire once every two seconds."

Which option do you take?

Unsure.  It's all about the statistical chance of being hit and, as you are now finding out, that might not be as obvious as you think.  Especially in real world settings.  There's a reason why Soviet infantry learned to ride ON TOP of their BTR/BMP instead of inside them

Now, if I were able to use the MG in the Halftrack the decision is a no-brainer.  Which is a reminder that by forcing the gunner to be exposed without giving him the ability to fire you're weighting the test in a way that does not equate well to real game situations.

Steve

[JoMac]

26 minutes ago, Rokko said:

Have I come at any point during this discussion across as unkind or agitated? If so, I wasn't aware and apologize.

Actually, you are doing just fine, LOL, and being very constructive 

And, don't worry about hijacking a 'Thread', because a possible issue might sometimes lead to another bigger issue that needs addressing. 

Edited April 12, 2016 by JoMc67

[Battlefront.com]

54 minutes ago, Rokko said:

I am sorry but how is that not totally controlled for by only counting the shots it took? Doesn't matter how long it took, doesn't matter how many people shot, etc. Especially with the new test where it's only a single guy.

I was responding to the previous conversations.  Your current test is fine in this regard.

Quote

Does confirm what? Do you mean it is equally flawed in your estimation?

Quite the opposite.  You did your first test and I said the ducking thing was not just a footnote, but a fundamental flaw in the logic of your test.  Your second test eliminated (as best one can) the flaw and the results were not surprisingly radically different.  In a good way

Also, I meant to give you kudos for putting the shooters behind a wall in order to give them a consistent kneeling stance.  Yes, firing prone does change many significant variables regarding accuracy, so it's definitely a good idea to keep the shooting stance consistent.

Quote

No I haven't btw, not that it matters. The real question is, how does the game handle aiming? In my understanding it is mainly based on center of visible mass. Camouflage and whatnot is not at all taken into account IIRC. And please take a look at the pictures I've posted, I don't see how anyone could get to the conclusion the prone infantry man would be easier to hit.

(you mean harder to hit)  Here's where you need to pay more attention to reality than to how the game portrays it.  Everything I said about targeting is how the real world works.  Aiming is a complicated amalgamation of weapon, eyes, body, and brain.  Use a rifle instead of a pistol and the weapon has more room for error.  Add optics to a rifle and the eyes have more room for error.  Add a bipod to the rifle and the body has more room for error.  Make the target massively obvious and you give the brain more room for error.  I don't want to have an argument about that stuff any more than I want to argue if the Earth if flat or round.  It just is what is.

Ask yourself this.  Why did you use maximum zoom when you took the screenshots?  I know the answer already but I would rather you phrase it in your own words.

 

Quote

But thats exactly the (suicidal) behaviour one has to replicate at the moment, because...

...no it plainly doesn't. A halftrack gunner will NEVER EVER refrain from staying upright. Only by being told so or by dying will he leave his post, doesn't matter what experience or motivation he has. The only exception I know if is ratteling while the vehicle is immobilised, in that case I've seen crews abandon their vehicle entirely (and getting hit while doing so).

True, but if you're trying to figure out how things are working in an actual game setting then you shouldn't skew tests conditions away from "normal".  My approach is to, at least initially, keep the "soft factors" as neutral as possible, then if there seems to be a reason to change them I change them.  That said, in this case if you are testing for "chance of hitting" then you should keep the truck driver cranked up so he is less likely to duck down.  If you're trying to figure out something about protection levels, then maybe not.  It doesn't really matter as long as you keep in mind that tests generally don't lend themselves to evaluating more than one thing at a time.

Steve

 

[Andrew H.]

1 hour ago, Battlefront.com said:

  For example, I don't know how well the armor deflects a 30.06 round at 200m.  For all I know there's a chance rounds penetrate.

 

Steve

 This is worth looking into.

The Garand's AP ammo had a steel penetrator and could penetrate about 10 mm of vertical armor at 100 yards.  The MG gunshield on the StuG was a 10 mm armor plate (I'm not sure if it's the same gunshield on HTs, although they look similar).  The gunshield is sloped, though, so at 200 meters the shooter is probably safe from penetration.  Most of the time. 

[Rokko]

I have now run 30 tests each. Averages so far:

vs HT: 10.63

vs Inf: 30.43

If you discount the 4 outliers (?) that were in the range of 100 (98, 97, 105, 100) the average is still higher at 19.63

Some of the hits against the HT gunner were clearly shield penetrations or at least hits on the shield. Still, to me this result seems wrong. I'd like to get at least 100 runs each before pushing the test to the 200m mark were shield penetrations would possibly not be an issue anymore.

[Bulletpoint]

24 minutes ago, Andrew H. said:

 This is worth looking into.

The Garand's AP ammo had a steel penetrator and could penetrate about 10 mm of vertical armor at 100 yards.  The MG gunshield on the StuG was a 10 mm armor plate (I'm not sure if it's the same gunshield on HTs, although they look similar).  The gunshield is sloped, though, so at 200 meters the shooter is probably safe from penetration.  Most of the time. 

If the bullet can penetrate a maximum of 10mm vertical plate at 100 metres, wouldn't that mean it cannot penetrate the plate if it is angled just a little bit? Which means at 200 metres there should be no chance at all of penetrating?

[Battlefront.com]

10 minutes ago, Rokko said:

I have now run 30 tests each. Averages so far:

vs HT: 10.63

vs Inf: 30.43

If you discount the 4 outliers (?) that were in the range of 100 (98, 97, 105, 100) the average is still higher at 19.63

Some of the hits against the HT gunner were clearly shield penetrations or at least hits on the shield. Still, to me this result seems wrong. I'd like to get at least 100 runs each before pushing the test to the 200m mark were shield penetrations would possibly not be an issue anymore.

Excellent.  Now I think we're getting somewhere   For this few tests (30 is not many) the best thing to do is discard the major outliers.  The four you picked are clearly way out there, so taking them out of the mix is the right thing to do.

OK, so now what do we have at this point?  That it is takes roughly 50% as many shots to affect a state change on a soldier in the open vs. a soldier in a Halftrack.  Next thing to look at are the state change types.  Which means noting the number of each of these for both HT gunner and driver (separately of course).

1.  Ducking

2.  WIA

3.  KIA

Moving the test out further should be interesting.  I'd do both 200m and 300m as previous tests indicate there is a significant difference between 200m and 300m.  I think 30 tests each is OK for preliminary testing.

With the tests moved further out we can better assess if there might be a protection problem with the 251's gunshield modeling, though I'd test that theory a different way.

Steve

[Battlefront.com]

11 minutes ago, Bulletpoint said:

If the bullet can penetrate a maximum of 10mm vertical plate at 100 metres, wouldn't that mean it cannot penetrate the plate if it is angled just a little bit? Which means at 200 metres there should be no chance at all of penetrating?

Yup, but the test noted above was at 100m.  I don't know what we have the shield rated for, but a quick check shows the armor thickness for the vehicle as a whole varied between 6mm and 14.5mm.  I'll see if I can get a figure.

Steve

[Rokko]

6 minutes ago, Battlefront.com said:

Excellent.  Now I think we're getting somewhere   For this few tests (30 is not many) the best thing to do is discard the major outliers.  The four you picked are clearly way out there, so taking them out of the mix is the right thing to do.

OK, so now what do we have at this point?  That it is takes roughly 50% as many shots to affect a state change on a soldier in the open vs. a soldier in a Halftrack.  Next thing to look at are the state change types.  Which means noting the number of each of these for both HT gunner and driver (separately of course).

1.  Ducking

2.  WIA

3.  KIA

Moving the test out further should be interesting.  I'd do both 200m and 300m as previous tests indicate there is a significant difference between 200m and 300m.  I think 30 tests each is OK for preliminary testing.

With the tests moved further out we can better assess if there might be a protection problem with the 251's gunshield modeling, though I'd test that theory a different way.

Steve

Oh, no I didn't count state changes, only hits. HT gunners don't have any state changes, their suppression increases but their state is always "spotting" I think. Since HT gunners don't have state changes I didn't count them for the prone truck driver either, they often went to "cowering" but only for very short amounts of time. I did differentiate between KIA (including "heavy" wounds) and WIA though. For the HT test there were 29 K's and 1 W.

For the other test there were 3 W's and 27 K's.

[Battlefront.com]

Hmm.  OK, I thought I was clear in previous suggestions that the state changes are what you should key on and therefore should be recorded.  Oh well, let me restate again why this is important.

Because there is no way to keep the truck driver kneeling and no way to make the HT gunner duck down, we have ourselves a bit of a problem when doing a direct comparison.  And that is in one case the TacAI is allowed to pursue a behavior that decreases the chance of being hit at the expense of being able to shoot back.  In the other case the TacAI is NOT allowed to pursue this behavior because it wishes to remain able to shoot.  That isn't "fair" and when trying to do a direct comparison things have to be as "fair" as possible.

Here's my recommendations:

1.  Put the wall back in front of the truck driver

2.  Run the test again and any time the truck driver ducks behind the wall count how many shots it took and note the result (duck, WIA, KIA)

3.  Since the HT gunner can not change states just note how many shots it took to be WIA or KIA

4.  Record 30 state changes, no matter what the change is, for each

You should get something like this:

HT Gunner
# shots WIA (# instances)
# shots KIA (# instances)

Truck Driver
# shots Duck (# instances)
# shots WIA (# instance)
# shots KIA (# instances)

Now that you have this data we can see what we can make from it.  We can say "let's remove the Duck instances from the equation because the HT Gunner can't Duck".  We could say "let's presume the HT gunner would Duck proportionally as much as the truck driver".  That sort of thing.  That allows us to compare WIA/KIA apples to apples for each.

If the results show that there are almost no WIA/KIA for the Truck Driver because the first 30 instances recorded were pretty much all Ducks then we have ourselves more evidence that the problem isn't so much targeting or protection, but rather the TacAI for the HT gunner being flawed when artificially prevented from working as designed.  And from there we can form some opinions on that as well.

So on and so forth.

Steve

 

[Rokko]

14 minutes ago, Battlefront.com said:

Hmm.  OK, I thought I was clear in previous suggestions that the state changes are what you should key on and therefore should be recorded.  Oh well, let me restate again why this is important.

Because there is no way to keep the truck driver kneeling and no way to make the HT gunner duck down, we have ourselves a bit of a problem when doing a direct comparison.  And that is in one case the TacAI is allowed to pursue a behavior that decreases the chance of being hit at the expense of being able to shoot back.  In the other case the TacAI is NOT allowed to pursue this behavior because it wishes to remain able to shoot.  That isn't "fair" and when trying to do a direct comparison things have to be as "fair" as possible.

Here's my recommendations:

1.  Put the wall back in front of the truck driver

2.  Run the test again and any time the truck driver ducks behind the wall count how many shots it took and note the result (duck, WIA, KIA)

3.  Since the HT gunner can not change states just note how many shots it took to be WIA or KIA

4.  Record 30 state changes, no matter what the change is, for each

You should get something like this:

HT Gunner
# shots WIA (# instances)
# shots KIA (# instances)

Truck Driver
# shots Duck (# instances)
# shots WIA (# instance)
# shots KIA (# instances)

Now that you have this data we can see what we can make from it.  We can say "let's remove the Duck instances from the equation because the HT Gunner can't Duck".  We could say "let's presume the HT gunner would Duck proportionally as much as the truck driver".  That sort of thing.  That allows us to compare WIA/KIA apples to apples for each.

If the results show that there are almost no WIA/KIA for the Truck Driver because the first 30 instances recorded were pretty much all Ducks then we have ourselves more evidence that the problem isn't so much targeting or protection, but rather the TacAI for the HT gunner being flawed when artificially prevented from working as designed.  And from there we can form some opinions on that as well.

So on and so forth.

Steve

 

The removal of the wall in front of the truck driver also caused him to NEVER kneel. He is constantely prone, just like in the picture I've posted. In fact there was only one case where the truck guy started the test run in kneeling position and I aborted that one.

So the only states were "normal" prone and "cowering" prone, besides being dead or wounded.

So since he was prone either way and the terrain he was on was pavement I figured the benefits from cowering/ducking were minimal, i.e. almost non-existant, so it would hardly make a difference. If you tell me there is some abstraction going on that makes being in cowering state so much more beneficial in terms of protection even without any cover and on totally even ground, then I will have to reconsider the setup indeed.

[Battlefront.com]

Ah.  I didn't understand the bit about the wall having the guy go prone right from the start.  Smart TacAI

Given this then we can go with what you have already.   Under ideal shooter conditions you discovered

1 in 10 shots hit the MG gunner (10% hit chance)

1 in 20 shots hit the prone soldier (5% hit chance)

Given these are perfect conditions for the shooter (including short range) and the MG guy isn't allowed to duck or to fire back, I don't see any problems with the results.  Re running the test at 200m and 300m should be interesting because that's kinda the range where complaints start.  Re running the test where the MG guy can shoot back would be a fun one, but I think we all know what the results will look like

Steve

[JoMac]

KHAAAAAANNN !

[Battlefront.com]

Confirmed the shield is 10mm thick in the game.

Steve

[Andrew H.]

2 hours ago, Bulletpoint said:

If the bullet can penetrate a maximum of 10mm vertical plate at 100 metres, wouldn't that mean it cannot penetrate the plate if it is angled just a little bit? Which means at 200 metres there should be no chance at all of penetrating?

Sources are inconsistent in what they mean by "penetration".  Usually when penetration is stated as a number, it means that the projectile has a 50% chance of penetrating that number; presumably it has a smaller chance of penetrating a few millimeters more.  But the gun shield on the HT does seem to be sloped pretty well - eyeballing it, it looks like it's angled about 30 degrees from the vertical, and then goes back about the same on each side as well.

 

 

[Lethaface]

Great to see the discussion follow up in intelligent and civilized fashion 

I don't (yet?) fully understand why it should be more difficult to hit the smaller chunk of 'meat' behind the shield versus the larger chunk of meat prone on the pavement. From a computing viewpoint the 'visible surface' of the prone soldier is larger and thus easier to hit when accuracy is constant. One thing I'm interested to know more about is the effect Steve mentioned regarding the tighter groupings versus a shield then a human silhouette. In RL I do have 'experience' that it is easier to spot/aim for an object A that is partially sticking out from a larger object B versus the same object A being somewhere in 'the open'. I don't know the science behind that (anyone does?) and it's an extra challenge to formulate this stuff in English at 1:30 AM while having a beer, but even logical thinking can explain why it's easier for the brain to recognize an abnormal pattern on a large object (something sticking out) versus the same object that was sticking out laying around on the all encompassing floor (even if it is on a surface like pavement). 

Is that (part) of the reason why it should be harder to hit the prone infantry on the pavement in-game, versus the HT gunner? If so and the game actually simulates such effects, that's pretty slick!  
A follow up question would be whether and how much this 'effect' would be affected by the range from spotter/aimer to the target.

Edited April 12, 2016 by Lethaface

[Battlefront.com]

43 minutes ago, Lethaface said:

Great to see the discussion follow up in intelligent and civilized fashion 

I don't (yet?) fully understand why it should be more difficult to hit the smaller chunk of 'meat' behind the shield versus the larger chunk of meat prone on the pavement. From a computing viewpoint the 'visible surface' of the prone soldier is larger and thus easier to hit when accuracy is constant. One thing I'm interested to know more about is the effect Steve mentioned regarding the tighter groupings versus a shield then a human silhouette. In RL I do have 'experience' that it is easier to spot/aim for an object A that is partially sticking out from a larger object B versus the same object A being somewhere in 'the open'. I don't know the science behind that (anyone does?) and it's an extra challenge to formulate this stuff in English at 1:30 AM while having a beer, but even logical thinking can explain why it's easier for the brain to recognize an abnormal pattern on a large object (something sticking out) versus the same object that was sticking out laying around on the all encompassing floor (even if it is on a surface like pavement). 

We're talking about the same effect, but you're forgetting that the shield is attached to a halftrack.  What is easier to spot and aim for at 100m, a halftrack of a guy lying prone?  Halftrack, right?  Right   Aiming works in stages.  First stage (in this case) is aiming in the general direction of the halftrack, then tightening the aim on the halftrack, then tightening the aim at the shield, then at the middle of the shield.  This might sound like a lot of work, but anybody who has shot a gun of any sort, kinda sorta even in an FPS, knows that it is not.  It's totally intuitive.  Contrast this with trying to get a bead on what amounts to a helmet and shoulders on the ground.  Much more difficult.

Now, in these artificial conditions it might be argued that it's about equal or there's maybe even a slight advantage to shooting at the soldier.  And here's where we run into some limitations of artificial tests.  The game mechanics were not programmed to simulating shooting ranges, rather they were designed for combat environments.  This means there are assumptions built into the various calculations and sub routines that might not work as you'd expect in a sanitized, artificial environment such as this.  Often times it's subtle, but sometimes not.  Aiming and suppression are some of the things that are more subtle.

Quote

Is that (part) of the reason why it should be harder to hit the prone infantry on the pavement in-game, versus the HT gunner? If so and the game actually simulates such effects, that's pretty slick!  

Yes it does, but it's not as slick as you might think   It's pretty standard programming practice.  Every single thing in the game belongs to one or more (usually more) "class" of similar objects.  Each class is designed to represent generalized attributes common to all members of that class.  These define lighter/heavier, easier/harder, faster/slower, nosier/quieter, etc. vs members of other classes.  These attributes are then used as shortcuts for the CPU and memory when performing functions, such as spotting and targeting.  We have some degree of ability to tailor the specifics of any member of a class depending on the flexibility of that specific code.  Weight classes, for example, determine what unit can go over which type of bridge.  It doesn't matter if the tank is 48 tons or 52.69 tons.... if they are in the same class they behave identically in this particular circumstance even though they might behave totally different in others.  Other things are designed to be more flexible and the class is used only as a "rule of thumb" with further refinement for members within.  Infantry units have a lot in common, but a 1 man Team is a lot more difficult to spot than a 7 man Team.  A 1 man Flamethrower is a higher priority target than a 1 man Ammo Bearer.  Etc.

In this case the US ammo bearer is targeting a Halftrack in one case and Infantry in another.  This is where targeting starts.  It progresses from there with the TacAI deciding exactly what part of the Halftrack to target (obviously the meat sack behind the shield is a great place to start!) and which Soldier within that Vehicle Crew to go after (in this case the choice is rather obvious).

Quote

A follow up question would be whether and how much this 'effect' would be affected by the range from spotter/aimer to the target.

Yes, weather screws around with all kinds of stuff.  Put the soldier in snow and he'd be even harder to hit while the dude in the halftrack would be just as easy.  Add falling snow or heavy rain and the truck driver would likely be proportionally much harder to hit than the dude in the halftrack.  Why?  Again, because it's easier to establish and maintain an aimpoint on a vehicle than it is a dude lying down on the ground.

Range also has a huge factor in this.  Move both to within 25m of the shooters and I think you'd see that the delta between HT gunner and truck driver would narrow considerably.  If the average dropped to 2 shots for each then you'd have the truck driver losing far more margin of safety compared to the HT driver (20 down to 2 vs. 10 down to 2).  Likewise, push both targets out to 1000m and you'd probably have a very low chance of picking off the HT gunner and probably would never even spot the truck driver.

Which is why a generalized "feeling" about this sort of thing is only a possible starting point for further investigation, not a "that's all you need to know, now fix it" sort of thing.  Contrast this with the second crew member becoming a casualty complaint.  That one doesn't need any more testing than the quick one done many pages ago.  Why?  Because there's a really straight forward cause/effect and the TacAI is not sensitive to much of any variables in this case.  Which is why we are looking into making a change

Steve

[Lethaface]

Thanks for the detailed explanation! While object orientated abstractions are part of my job (fortunately programming them isn't anymore ;-)), I'll need some time to ponder about this information.
Feelings are indeed just feelings, however in my life the 'abnormal pattern effect' has often caused a subliminal cue (feeling) that a particular issue is worthy of further investigation. I think that creativity is allowing subliminal solutions/explanations for particular issue to form. The difficult (and interesting, imo) part is then to dissect feeling from fact. How else did Einstein come up with his theory?

Anyway I thought that it's funny that a similar (simulated) effect is causing both the 'accuracy delta' between the HT-gunner/truck driver and the feeling that there's something wrong with that delta 

 

Edited April 13, 2016 by Lethaface

[JoMac]

1 hour ago, Battlefront.com said:

We're talking about the same effect, but you're forgetting that the shield is attached to a halftrack. What is easier to spot and aim for at 100m, a halftrack or a guy lying prone ?...Halftrack, right?  Right   Aiming works in stages. First stage (in this case) is aiming in the general direction of the halftrack, then tightening the aim on the halftrack, then tightening the aim at the shield, then at the middle of the shield.  This might sound like a lot of work, but anybody who has shot a gun of any sort, kinda sorta even in an FPS, knows that it is not.  It's totally intuitive.  Contrast this with trying to get a bead on what amounts to a helmet and shoulders on the ground.  Much more difficult.

Steve

Yes, and that should be right and same thing I was thinking...It's easier to spot something that generally doesn't belong or sticks out in the immediate surroundings and so generally easier or faster to spot...Then, from there ( in case of HT w/Mg or Unbuttoned Crew) the Progression of Area-Fire, then becomes a point on the Vehicle, etc...It's good that this behavior is modeled. 

Now, are you also looking into having the Unbutton Crew of any Vehicle ( Tanks...including Stugs, HT's, etc ) have a TAC AI behavior change to ducking more often as it co-insides with Small Arms Progression ( Hey, those Bullets are getting awfully close, and think I should duck or button up ) . I know you also mentioned no Unbutton if loss of Crew member. 

Anyways, at this point, Steve, recap once more at what your going to consider looking into...unless you think nothing else has changed since your older post. 

 

Edited April 13, 2016 by JoMc67

[John Kettler]

What a dizzying discussion! I have absolutely nothing to contribute on the testing, how the TAC AI treats matters and such, but I would like to ask a few things about how the game treats the typical armored MG-34 mount. We have now established it is indeed 10 mm thick and is angled, but it is also raked, something like the way the gunshield is on the Pak 40.

No idea who did the art!  In any event, does the CMx2 engine treat the shield for the MG the same way it does for a separate ATG on the ground?  Also, there was discussion of the effectiveness of .30 '06 AP fire vs the shield. My understanding, from reading a number of period accounts and books about such matters, is that the riflemen fired ball, but that the BAR fired AP, specifically because of the cover busting it conferred over ball. Does the game model this? Seems to me that if AP penetration of the shield is a factor at 200 meter range, as has been speculated, then drastically reducing the number of AP shooters ought to considerably improve gunner survivability. Similarly, though not to as great a degree, the MG shield on the StuG III is also raked, as shown on this Finnish one at Parola and taken from Wiki.

I realize these issues are small ones in the vast scope of the overall discussion, but since they would appear to have real leverage over the likelihood of a gunner casualty, I thought I would raise them.  Finally, depending on how the TAC AI threats TCs, the Presented Area (targetable surface area) may or may not be larger than that of the prone infantryman. That worthy is, I would note, not just shoulders and arms, but also a chunk of torso to which are attached splayed legs. Two versions of the position, taken from the newsletter for the DCMP (Director of the Civilian Marksmanship Program), show this very much to be the case.

Either way you slice it, it appears to me there's a lot of GI to shoot at.

Regards,

John Kettler

P.S.

The auto-embed thing is a boon for me!

Edited April 13, 2016 by John Kettler

[Battlefront.com]

3 hours ago, Lethaface said:

The difficult (and interesting, imo) part is then to dissect feeling from fact.

Very difficult sometimes, but generally always interesting.  One of the reasons I know so much about the game's guts is because I've had to explore many issues like this over the years.  Back in the very early days of CMSF we had lots of them because the game engine was still quite young.  I remember one extremely important one that went from feeling to observed effect rather quickly, but it took us quite a long time to figure out exactly what was causing the problems.  For those of you who were around back then it was dubbed the "oblique angle problem".  After we got it narrowed down Charles confirmed it existed but didn't think it would produce such obvious problems, so we had to document that it did.  He then largely fixed the problem, but that then uncovered another problem.  That problem turned into a massive upgraded to the game which we dubbed "Enhanced LOS" (ELOS).  Ah, man those were long days

3 hours ago, Lethaface said:

Anyway I thought that it's funny that a similar (simulated) effect is causing both the 'accuracy delta' between the HT-gunner/truck driver and the feeling that there's something wrong with that delta 

Very, very often we find that "feelings" are rooted in a misunderstanding of how things work in the real world.  This is a good example because it is definitely unintuitive that it would be easier to aim and therefore hit the HT shield than it would be a meat puppet out in the open.  Just like many people don't understand how much harder it is to shoot at a laterally moving target vs one moving directly towards/away.

3 hours ago, JoMc67 said:

Yes, and that should be right and same thing I was thinking...It's easier to spot something that generally doesn't belong or sticks out in the immediate surroundings and so generally easier or faster to spot...Then, from there ( in case of HT w/Mg or Unbuttoned Crew) the Progression of Area-Fire, then becomes a point on the Vehicle, etc...It's good that this behavior is modeled. 

Yup.  Next time any of you are out on a firing range or in a FPS game check it out if you haven't before.  It's so natural most people don't even notice they're doing it.

3 hours ago, JoMc67 said:

Now, are you also looking into having the Unbutton Crew of any Vehicle ( Tanks...including Stugs, HT's, etc ) have a TAC AI behavior change to ducking more often as it co-insides with Small Arms Progression ( Hey, those Bullets are getting awfully close, and think I should duck or button up ) . I know you also mentioned no Unbutton if loss of Crew member. 

Yes, we are looking at this.  Part of the problem is that the current behavior is generally correct, but the player choice to leave vehicles unbuttoned within effective small arms range is making it seem worse than it is.  So we have to be careful we don't get ourselves into a problem where gunners are not doing their jobs and vehicles get whacked as a result.  Because you know we'll see plenty of thread if that happens.  "My stupid f'n gunner is a coward and I lost all my tanks because of your stupid TacAI tweak!!!"

3 hours ago, JoMc67 said:

Anyways, at this point, Steve, recap once more at what your going to consider looking into...unless you think nothing else has changed since your older post. 

Nothing has changed since my last post, though I might describe our thinking a bit differently.  Basically we are looking at:

1.  Buttoned vehicles should be very reluctant to send anybody topside except when absolutely necessary.

2.  When a casualty happens while Unbuttoned the vehicle should go into Buttoned mode with the tweaked behavior of the previous point.

3.  The decision to go topside (reloading, defending, etc.) should be for the shortest time possible and only when absolutely necessary.

4.  Ducking should be allowed and reasonably responsive to enemy fire.  This is the trickiest of all of them, though, since we have to prevent a yo-yo effect.

Steve