Hull-down spotting disadvantage

[John Kettler]

All,

I need to scream, because the confounded BB kicked me out to Login and ate a thoroughly researched post, with many useful vid segments included, so shall summarize.

Driver vision, even direct view through a port from a restored static tank in a museum, is very poor. Naturally, it's much worse while on the move.

U.S. bow MG has no telescopic sight, and when buttoned, is fired using the periscope to aim. No brow pad fitted, so bow gunner can't weld himself to MG and brow pad, as the German bow gunner can, while moving. All German tanks and AFVs fitted with ball mount MGs were fitted with the KZF2 telescopic sight and a brow pad, which came with the sight.

http://www.warrelics.eu/forum/germany-ww1-ww2-armour-artillery-vehicles/kzf-2-kugelzielfernrohr-2-a-19089/

Thus, to some degree, the German bow gunner can see targets at longer ranges while moving and can certainly do so while stopped. Both the German drum type cupola and the later molded one on the Panther G place the TC's eyes higher above the turret roof than is the case for the 76mm Sherman. In the Panther G case, way above.

Given that BadgerDog reports having severe problems finding a tank at 500 meters, as a TC head up with binos, it seems reasonable to conclude that the much lower unmagnified periscopes on a Sherman and other AFVs so equipped will be useless at all but the closest ranges when it comes to target spotting. Therefore, spotting a hulldown static tank, at range, while moving is hopeless for the driver and bowgunner of such Allied AFVs.

Given the above, I believe it's fair to say that the spotting leverage supposedly provided by the driver and bow gunner of the hull up, buttoned Sherman is illusory at all but the closest ranges.

It would seem, therefore, that we're really back to Pd (Probability of detection) as a function of target size. And for that, I've already provided a link to a curve in a visual detection probability study.

Note:

I delved into this set of issues to the level of interior stills, HiRes: through the driver visor footage, footage shot inside Tiger 131 (early cupola, optics about same height above turret roof as 76mm Sherman), exterior and interior footage of two Panzer IV/Gs, Tunisia combat footage inside a Panzer IV/G, to include what could be seen through the vision blocks in the drum cupola.

Sherman research included my own hands on experience, wartime external footage, wartime interior footage, HiRes Stills and videos of the interiors and exteriors of restored Shermans. I paid particular attention to pitching movements of a moving Sherman even while on a runway apron, movements which became more pronounced on a flattish meadow, then went downhill from there. Alternating, variable dwell, views of sky and earth don't make for good spotting!

Regards,

John Kettler

[Vanir Ausf B]

Before hitting Reply: Right click --> Select All --> Copy. Or whatever the Mac equivalent is.

[Vanir Ausf B]

BTW, I will hopefully have some test results by early Friday morning. The Panther testing is (probably) done. I have not entered all the data into the spreadsheet yet, but the second 100 spot times have mean of 139.5 seconds, only .4 seconds different than the first 100 I posted.

[A Canadian Cat]

Here's a fun way of finding out...

...

LOL very visual description.

I stayed at a Holiday Inn last night..

LOL even funnier response.

(Click PIC to Enlarge)

Is that you listening to your hip hop music?

Anyway, I apologize folks (again)..

What apologize for some first hand info and one of the funniest posts on the board? Don't do that.

[A Canadian Cat]

Seriously, the point is very well taken. We're doing various tests to see how much of an impact the driver/radio guy play in the overall scheme of spotting.

Excellent, thanks for looking at this Steve. Glad you guys are so good at trying to improve things when customer's do find odd happenings.

[Jock Tamson]

This has been a very interesting thread, but I fear the blocker to significant changes may be the implications for the TacAI - stopping to spot / shoot, stalking behaviour etc.

[Battlefront.com]

Findings are still preliminary, but it doesn't appear there is anything significantly wrong. Two things we are still looking at:

1. The strength and/or range at which the driver and radio operator's contribution to spotting drops off to effectively nothing. There is some indication that they should be less valuable to spotting than they currently are.

2. Possible need to tweak the qualifications for when a tank is Hull Down. Due to the limitations of computer hardware, the determination for Hull Down is not granular (subtle) but more-or-less binary. This effects whether or not the vehicle gets a concealment bonus that reduces the chances of being spotted. There is some indication that larger tanks have a harder time meeting the criteria than they probably should.

Now, if you add these two together you can see how they could conspire to skew spotting results.

Notice that this has absolutely nothing to do with defensive bonuses from being behind something. Since the game does literally trace a round's trajectory, if you have a tank that is 1/3rd behind cover it will be proportionally more difficult to hit than if it was behind no cover at all. It doesn't matter one iota if the game considers that tank Partial or Full Hull Down or not.

Steve

[Amizaur]

So, in the first test the tank A (set in hull-down position) possibly was still treated by the game engine as fully exposed, same as the B (set in the open terrain) tank ?

But the tank A being partially hided, had less "eyes" available for spotting (because hull vision ports were obscured) so having effectively the same visual "signature" as fully exposed B tank it performed worse in spotting competition ?

[John Kettler]

Reconstructing parts of my research.

Panzer IV/G driver's view. 2:04

Panther G cupola assembly. Smack at beginning.

Tiger 131! Bow MG and driver positions at beginning and positions in turret later.

Panzer IV/G exterior of drum cupola clearly showing height of optics above turret (right at beginning).

Panzer IV/G in Tunisia. Exterior view of drum cupola with viewing ports open 2:18. TC in the cupola of a moving Panzer IV/G 2:23, followed by shot through an open port.

http://www.youtube.com/watch?v=1gOOfY40wyg

Regards,

John Kettler

[A Canadian Cat]

...Two things we are still looking at:

1. The strength and/or range at which the driver and radio operator's contribution to spotting drops off to effectively nothing....

2. Possible need to tweak the qualifications for when a tank is Hull Down...

...

Now, if you add these two together you can see how they could conspire to skew spotting results.

Indeed.

Investigation points sound good. With #1 the tank in the open would, at a certain range, have effectively the same number of eyeballs looking for threats that the hull down tank does. That puts them on equal footing in terms of eyes looking for threats. After a certain range. Again sounds good.

The only concern I would have then is with how #2 effects things. If we now have only the turret seated crew looking for a hull down tank and the reverse - only the turret seated crew looking for the tank in the open. My expectation would be for the hull down tank to have a significant spotting advantage because it would usually be harder to see. When you say that being hull down is binary rather than granular for CPU load reasons I get that too. When you say larger tanks currently have a harder time meeting this criteria red flags pop up in my head just because I do not consider Shermans or PzIVs as larger tanks. Where do you draw the line here?

What I get from your explanation is that it is possible to have a tank in the open and a tank behind some obstruction, but not meeting the game's hull down criteria, and be on equal footing in terms of concealment. I am OK with that as long as we can tell when a tank is hull down and when it is not (yes, to a specific area of the battle field).

I have no trouble with a tank in the open and a tank behind a low wall being equal in terms of how easy they are to spot. But at what point do we flip to the hull down concealment advantage? Is it when the blue line says our tank is hull down? Is that message based on infantry looking at our tank? Is that message based on a tank of equal size looking at our tank.

I care about this because if we get the hull down position wrong then the spotting it just as if we left the tank parked in the open. So, I want to be able to tell when I get hull down right.

Notice that this has absolutely nothing to do with defensive bonuses from being behind something.

I personally never saw anything that made me concerned about that.

[akd]

When you say larger tanks currently have a harder time meeting this criteria red flags pop up in my head just because I do not consider Shermans or PzIVs as larger tanks. Where do you draw the line here?

Steve means taller.

[Dan Dare]

http://information2share.files.wordpress.com/2011/05/tank-size-comparison-chart.jpg

[Battlefront.com]

So, in the first test the tank A (set in hull-down position) possibly was still treated by the game engine as fully exposed, same as the B (set in the open terrain) tank ?

Yes, it is entirely possible. In fact, it is most likely the reason.

Hull Down status can be tested by using the Target line from the shooting unit and/or from the target unit. Line color and text gives you some information about Hull Down status.

But the tank A being partially hided, had less "eyes" available for spotting (because hull vision ports were obscured) so having effectively the same visual "signature" as fully exposed B tank it performed worse in spotting competition ?

Possibly. It could also be one tank has better spotting capabilities (optics) than the other. It could also be that one is easier to spot than the other. Other factors too.

One thing we know is that at 1000m or greater the hull crew members don't appear to add significantly (if at all) to the spotting chance. At that range for sure it's only turret crews doing the spotting.

Steve

[Battlefront.com]

The only concern I would have then is with how #2 effects things. If we now have only the turret seated crew looking for a hull down tank and the reverse - only the turret seated crew looking for the tank in the open. My expectation would be for the hull down tank to have a significant spotting advantage because it would usually be harder to see.

Yes, when everything is equal (both tanks are identical, neither has hull crew spotting, neither moving, same facing, same experience, etc.) the chances of two tanks in the open should be identical. Put one behind some terrain and the chances one will be spotted should be lower than the other. That definitely is how it works in the real world.

When you say that being hull down is binary rather than granular for CPU load reasons I get that too. When you say larger tanks currently have a harder time meeting this criteria red flags pop up in my head just because I do not consider Shermans or PzIVs as larger tanks. Where do you draw the line here?

As AKD noted I meant "tall". Tanks are generally categorized as either "Tall" or "Very Tall". The PzIV is considered "Tall" while a Panther is considered "Very Tall".

What this means is that it is harder for the Panther to achieve Hull Down because it needs more terrain to do that with.

What I get from your explanation is that it is possible to have a tank in the open and a tank behind some obstruction, but not meeting the game's hull down criteria, and be on equal footing in terms of concealment. I am OK with that as long as we can tell when a tank is hull down and when it is not (yes, to a specific area of the battle field).

As stated above, the Target tool gives this information relative to a specific point on the map. And that's critically important to keep in mind because Hull Down is not a universal condition. It's always point to point specific. Which is why there's such a hit to the CPU for this sort of thing.

I have no trouble with a tank in the open and a tank behind a low wall being equal in terms of how easy they are to spot. But at what point do we flip to the hull down concealment advantage? Is it when the blue line says our tank is hull down? Is that message based on infantry looking at our tank? Is that message based on a tank of equal size looking at our tank.

It is generic information, of course, because there's no way for the system to tell you all the possibilities for a given point based on every conceivable situation.

I care about this because if we get the hull down position wrong then the spotting it just as if we left the tank parked in the open. So, I want to be able to tell when I get hull down right.

In real life a tank would go into hull down position relative to a specific anticipated area of engagement. And even then rarely did they have the chance to double check from that position to ensure they were truly hull down. Which is to say we have no intention of making a system which gives the player 100% reliable and foolproof information. However, players should be able to achieve reasonable hull down results using the information we provide. We're checking into this to make sure that we do.

I personally never saw anything that made me concerned about that.

The other thread appeared to be calling this into question, I think. It was tough to figure out Whatever the case is, our collective testing indicates that there is no problem here. And of the two elements this is the one that is more important. Especially because in game terms it's unlikely that you'd have a hull down vehicle facing a non-moving enemy without the hull down vehicle engaging it.

This is a good time to remind people that just because something works one way in an artificial test doesn't mean it comes up in the average game situation. Some are definitely more likely to than others. The testing simply tells us there is a possibility of something happening within the game, not that it absolutely does or to what degree it does.

Steve

[HistoryLover]

The other thread appeared to be calling this into question, I think.

No. The other thread you closed down mentioned, that the RESULT is wrong, and I also gave the correct explanation: because HD tanks are spotted earlier than they can spot themselves. I explicitly mentioned the importance of spotting first to hit first!

Because of this error, the whole tank combat model has not worked for years, because not being HD and therefore being smaller is key to success, but making the enemy tank button down and therefore destroy his spotting ability is the key.

At least in the meantime it should be clear to everyone that the size of a target can be compensated with the amount of shots fired at the target to achieve the same or better hit probability in the game. And exactly that is the problem I described from the beginning, that spotting behaviour of HD tanks is so bad, that they are hit more often in duels than tanks in the open.

It's enough that you lock up threads with excuses, but at least you should refrain from intentionally misrepresenting what others wrote.

[Battlefront.com]

No. The other thread you closed down mentioned, that the RESULT is wrong, and I also gave the correct explanation: because HD tanks are spotted earlier than they can spot themselves.

Not accurate. This MIGHT happen in some situations and it MIGHT be unrealistic when it does. And that is more likely to happen in artificial tests than in the game. So technically what you said is factually incorrect. As is this...

Because of this error, the whole tank combat model has not worked for years, because not being HD and therefore being smaller is key to success, but making the enemy tank button down and therefore destroy his spotting ability is the key.

The WHOLE tank combat model? If you want to be taken seriously, you can start by dialing back the hyperbole and substitute a statement that has at least some relation to the facts.

At least in the meantime it should be clear to everyone that the size of a target can be compensated with the amount of shots fired at the target to achieve the same or better hit probability in the game.

It's not clear to me because I don't even know what you're saying. Not that I think it matters much because you're so totally convinced of things which either aren't true or are taken out of context.

And exactly that is the problem I described from the beginning, that spotting behaviour of HD tanks is so bad, that they are hit more often in duels than tanks in the open.

Except you presented absolutely no data to back that up. A flawed test in vacuum doesn't have any relation to what happens in a real game. As I said above, generally speaking the non-hull down tank is moving. That changes everything right there.

It's enough that you lock up threads with excuses, but at least you should refrain from intentionally misrepresenting what others wrote.

If you reread what I said I wasn't really sure what your point was. I'm also not the only one that couldn't figure out what you were going on about. The bits that are clear aren't very useful. And locking up a redundant thread is sensible. Especially when there is a much more useful one going on.

Steve

[Sgt Joch]

Because of this error, the whole tank combat model has not worked for years.

you have been here what, a month?

Why don't learn how to actually play the game before making silly comments like that. You have obviously no clue how the game actually works.

The "issue" you are so worked up about has no practical impact on the game. If it did, it would have been noticed before. If you actually played the game, you would know that.

What Steve said is clear and our internal tests show the same, being hulldown gives both a concealment and a protection bonus. Should it be tweaked? maybe, but there is nothing fundamentally wrong with the tank combat model.

[HistoryLover]

What Steve said is clear and our internal tests show the same, being hulldown gives both a concealment and a protection bonus. Should it be tweaked? maybe, but there is nothing fundamentally wrong with the tank combat model.

I just corrected Steve from Battlefront and you repeat it outright again?

I did not say hd tanks had no concealment or protection.

I said that spotting of at least the PzIV hd tanks is that off, that the hd protection is more than compensated by their lost ability to spot adequately and I said (and prooved that later) how by forcing them to button up, they can be put at an even bigger disadvantage than the opponent in the open.

You can keep on repeating it that everything was not that bad, but if the buttoned up PzIV, HD, with his 75/L48 looses around 80% of the duels @1500 m with the same opponent, buttoned down, in the open, the model delivers implausible results.

Did I say that it can't be corrected? Or that CM was not fun? Or not a great game? No. But in regard to realism in this situation of hd and at least the PzIV, the model is NOT slightly off. If a model delivers an inverted result to reality, then it is totally off. I don't know what is so hard to accept with that.

And secondly, I'm sure you do a decent work as beta tester, but as someone who obviously has not recognized that the performance of (at least the) PzIV tanks in hd is way off, you should not claim that I had no clue.

[Battlefront.com]

I did not say hd tanks had no concealment or protection.

Correct. What you said was "Because of this error, the whole tank combat model has not worked for years." How does making inflammatory, and obviously factually incorrect, statements help make your game experience better?

You can keep on repeating it that everything was not that bad, but if the buttoned up PzIV, HD, with his 75/L48 looses around 80% of the duels @1500 m with the same opponent, buttoned down, in the open, the model delivers implausible results.

And we're right back to where we started saying that your tests were flawed and don't prove what you say they do. Does your test, and other tests, indicate there is room for improvement? Perhaps, and we're looking into what can be done.

Did I say that it can't be corrected? Or that CM was not fun? Or not a great game? No. But in regard to realism in this situation of hd and at least the PzIV, the model is NOT slightly off. If a model delivers an inverted result to reality, then it is totally off. I don't know what is so hard to accept with that.

Because you are showboating and therefore it's difficult to take you seriously.

The real situation, the one that can be scientifically demonstrated, is there's probably a minor issue some of the time in some situations with some vehicles. And that minor issue may, or may not, produce a tangible effect within an actual game. What is important now is identifying exactly where improvements can be made. And statements that "the game is totally broken, BFC fix or do somefink" doesn't really help the process any.

And secondly, I'm sure you do a decent work as beta tester, but as someone who obviously has not recognized that the performance of (at least the) PzIV tanks in hd is way off, you should not claim that I had no clue.

No, the claim is that you are actively choosing to be combative and unhelpful. That's a different thing.

Look, if you want to have a POSITIVE impact here then don't pick fights with people who want to get to the bottom of REAL problems and get them ACTUALLY fixed. Making unsupportable and easily disproven wild claims, then getting all steamed up about being called on it, doesn't do anybody any good. Least of all whatever case you want to make.

Steve

[Vanir Ausf B]

Finished the Panther, Sherman and Pz IV long range spotting testing.

All tests were done spotting Sherman M4A3w Sherman 76 (early) tanks. Range 1200 meters. All tanks in the open (not hull down). All tanks buttoned.

Data points:

Panther A late:320

M4A3w Sherman 76 early: 300

Panzer IV H late: 300

Average spotting times in seconds

M4A3w Sherman 76 early: 127.8

Panzer IV H late: 128.5

Panther A late: 139.7

Well, so much for the much-heralded German optics The Sherman and Panzer IV are practically equal. The Panther is worse than both by a small but probably statistically significant margin (74% likelihood t-test, 64% Kolmogorov-Smirnov).

[Jock Tamson]

post error

[Sgt Joch]

You can keep on repeating it that everything was not that bad, but if the buttoned up PzIV, HD, with his 75/L48 looses around 80% of the duels @1500 m with the same opponent, buttoned down, in the open, the model delivers implausible results.

It should be obvious even to you that the pzIV in your tests were not "hull down" in game terms. Also your tests completely ignore the "movement" factor which has a huge impact on spotting in actual games.

That is why we said your tests are intriguing, but have no practical impact in game terms.

you should not claim that I had no clue.

well.....

You are not the first new player to have trouble with the game since it is a complex simulation. You will find players here are very helpful if you give them a chance.

However, if you just want to dig yourself into a deeper and deeper hole..well, we all love a good spectacle also...

[A Canadian Cat]

As AKD noted I meant "tall". Tanks are generally categorized as either "Tall" or "Very Tall". The PzIV is considered "Tall" while a Panther is considered "Very Tall".

What this means is that it is harder for the Panther to achieve Hull Down because it needs more terrain to do that with.

OK makes sense.

As stated above, the Target tool gives this information relative to a specific point on the map. And that's critically important to keep in mind because Hull Down is not a universal condition. It's always point to point specific. Which is why there's such a hit to the CPU for this sort of thing.

Thanks for clarification. Not only is it point specific but observer height specific. Do we know to what target height the target tool is using for reporting hull down status? In other words is it for a solider standing, a short tank, a tall tank etc.?

In real life a tank would go into hull down position relative to a specific anticipated area of engagement. And even then rarely did they have the chance to double check from that position to ensure they were truly hull down. Which is to say we have no intention of making a system which gives the player 100% reliable and foolproof information. However, players should be able to achieve reasonable hull down results using the information we provide. We're checking into this to make sure that we do.

Good. I will continue to just use the target tool knowing that a real TC would be making an educated assessment as well.

Thanks for looking into this and providing us with useful feedback. I appreciate it - and so do most others I am sure.

[Sgt Joch]

I just did a quick test.

setup:

-4 PzIV H late, unbuttoned, regulars positioned hull down behind the crest of a ridge.

-4 enemy AFVs, also unbuttoned, regulars, go over a rise and move towards the PzIVs with a "Hunt" command.

results:

-All 4 PzIVs spot first, fire first and knock out their target;

-none of the enemy AFVs fire, 3 never spot the PzIV firing at them, 1 gets a question mark and spots the PzIV just before it is knocked out.

I will do some more runs this weeked and will also try with Panthers, but at first glance, the "tank combat model" appears to be working fine. That does not mean, as Steve said, that the Hull down parameters/crew weighting will not be relooked at and maybe tweaked.

p.s. - never forget that "movement" has a huge impact on spotting. It is easier to spot a moving unit and conversely a moving unit has a harder time spotting other units.

[Battlefront.com]

Well, so much for the much-heralded German optics

From past discussions you probably are aware that we don't put much stock in the "much-hearlded German optics" position

The Sherman and Panzer IV are practically equal. The Panther is worse than both by a small but probably statistically significant margin (74% likelihood t-test, 64% Kolmogorov-Smirnov).

Seems fine to me. The Panther is a significantly larger tank and therefore is relatively easier to spot than the smaller Sherman or PzIV. At least in this instance where the three tanks in question have effectively the same spotting capabilities.

The other thing to keep in mind is that this sort of test doesn't offer much useful information. In real life there was never a case, ever in the course of the entire war, with two tanks magically appearing on the opposite sides of a long, flat, unobstructed field with neither one moving and both facing each other square on with no distractions. Therefore, whatever the results are they don't have much applicable value to the gameplay. And since the gameplay is the only thing that matters... well, you can follow the logic

What tests like this can do is establish basic understanding of baselines going into more applicable tests. In other words, tests like yours establish some basis for evaluating tests like this...

results:

-All 4 PzIVs spot first, fire first and knock out their target;

-none of the enemy AFVs fire, 3 never spot the PzIV firing at them, 1 gets a question mark and spots the PzIV just before it is knocked out.

Note... I wrote something which I felt wasn't accurate enough to help the conversation, so I removed it. No good comes from muddying waters

I will do some more runs this weeked and will also try with Panthers, but at first glance, the "tank combat model" appears to be working fine. That does not mean, as Steve said, that the Hull down parameters/crew weighting will not be relooked at and maybe tweaked.

It is a rare day when I ever claim that anything, at all, in the game is working 100% as it should be. It's almost as rare for me to claim that something in the game theoretically couldn't be made better. Unfortunately it is common to hear me say that changes aren't anticipated for one or more reasons. In this case I can't say when/if anything will change, just that we're at least looking into the potential need for specific tweaks.

p.s. - never forget that "movement" has a huge impact on spotting. It is easier to spot a moving unit and conversely a moving unit has a harder time spotting other units.

Adding to that, one must generally understand that "billiard table" tests need to be constructed very carefully to answer a specific question. Two cases in point:

Vanir's test establishes a basic spotting parity between three specific vehicles in one specific artificial setting. Since the test is both limited and artificial, the results can not be taken too far beyond limited and abstract conclusions.

Sgt Joch's test establishes a basic concept that a tank hull down and a tank moving is heavily weighted in favor of the non moving hull down tank. This is a little more applicable to a real game than Vanir's test, but it too is quite limited as it only tests two vehicles in one specific condition. Could be that the results would be the same even if the PzIV wasn't hull down simply because the Sherman was in motion and the PzIV not. Another test would be needed to test that hypothesis.

Vanir's test, while itself not very applicable to the game, could help draw some conclusions from a PzIV and Sherman matchup in a test similar to Sgt Joch's. And that is that two tanks with similar inherent spotting capabilities produces extremely different results when at least one variable (movement) is changed for one side. This gives us reassurances that the game's basic modeling, as seen in the games we play, is functioning as expected for similar situations.

Steve