This is unacceptable.

[Holman]

I think it's necessary to accept that we still don't have 100% modeling of every single aspect of real-life warfare.  In this case those aspects include every blink, distraction, and turn of the head of every crewman, what you might call FOW at the facial level.  Sometimes the enemy spots your unit when you (with your omniscient, zoomable, WASD'able view of the battlefield) think you should have seen him first; sometimes it's the other way around.  That's probably more realistic (in terms of what-happens-in-the-world) than strictly perfect simultaneous sight modeling would be. 

I've played CM for years and years, and I have a very strong sense that it handles these things more fully, more realistically, and more fairly than any other wargame.  The results in the long term are very consistent with reality.  One thing to remember, too, is that we never even notice the times we're on the winning side of these edge-case equations; that's when we congratulate our pixelguys on their sharp eyes and quick shooting.

[Pak40]

I am not convinced.  There are two things to keep in mind. First chance: there is chance involved in all these encounters and something that happens only rarely can still happen and if you recalculated the turn multiple times something else might happen all the other times you try it.  Stuff happens, get over it, we all have to.  Second look at the conditions here.  The Panther does have optics damage - sorry but that does make spotting harder.  also there is a burning tank to the right and a tree partly in the way.  It could very well be that the commander has an obscured view, the loader has smoke in the way an the driver is trying to keep the tank facing the right way according to the best the gunner can figure the incoming fire is coming from.  Suddenly it is not a huge surprise that the crew took 20-30s to figure out what was what.

While there may be infinitesimally small chances that one of these issues is the cause, I'm going with the odds and saying it's a bug. It should be treated as one until BF can prove that it's not. Whether it's a bug cause by an error in code or a "bug" caused by a smoke value or a "partial tree" blocking the view - it needs to be fixed. There's no logical explanation why a unsuppressed crew facing an enemy tank wouldn't see it from 30m away, especially after taking several rounds from it. If it's chance, then that Panther would have had to roll 01 on a D100 several times in a row. Extremely unlikely. And if that is the case then BF needs to re-evaluate how it calculates these obvious close encounters.

 

Edited January 18, 2016 by Pak40

[Vanir Ausf B]

So a buttoned Panther with damaged optics takes 15 seconds to spot a Sherman that is partially obscured by a burning vehicle and foliage? Am I missing something? It's an edge case but not really that hard to believe.

[Battlefront.com]

While there may be infinitesimally small chances that one of these issues is the cause, I'm going with the odds and saying it's a bug.

This is like saying that "going by the odds I played Powerball this week"   By this point in CM's lifespan, there's pretty much no bugs left to find in the underlying systems.  At least not ones that really matter.  Now, is there a "corner case" here that theoretically could use a bit of specific massaging?  Perhaps, but it's not worth pursuing unless the issue comes up more than once in a blue moon.  We only have so many hours in the day, so why spend it trying to fix something that might come up once every couple of thousand games when we could instead spend the time making something that may positively affect every game played?

There's no logical explanation why a unsuppressed crew facing an enemy tank wouldn't see it from 30m away, especially after taking several rounds from it. 

No logical reason?  You must have missed a bunch of posts above because there's plenty of logical reason for this happening.  There's also the LOS/LOF check clock that I mentioned.  I don't know why you haven't picked up on that because Vanir Ausf B came in late to the discussion and has a pretty good grasp of the whole discussion.

 

So a buttoned Panther with damaged optics takes 15 seconds to spot a Sherman that is partially obscured by a burning vehicle and foliage? Am I missing something? It's an edge case but not really that hard to believe.

Yup, that's a very good summary except.  Add to that the hardware limitations which impose a limit on how frequently LOS/LOF checks can happen and that further reduces the reasons to suspect "bug" or even "flaw".  I think the worst possible situation here is that we have a "limitation" due to hardware realities.

Steve

 

[Boomkow]

What are the LOS/LOF check clock timings? I think knowing those might help some of us experiment with our micromanagement skills.

For instance, in relation to timing a turn around a corner? (I want them focused)

[weapon2010]

So a buttoned Panther with damaged optics takes 15 seconds to spot a Sherman that is partially obscured by a burning vehicle and foliage? Am I missing something? It's an edge case but not really that hard to believe.

is optics a factor at 30m?

[MikeyD]

Inside a buttoned tank with a burning wreck billowing smoke in front of it... yeh.

[Vanir Ausf B]

is optics a factor at 30m?

If you wear glasses and your glasses are cracked it's going to be a factor at any range.

[weapon2010]

If you wear glasses and your glasses are cracked it's going to be a factor at any range.

I mean why would they be looking through the optics at all?, an not  the vision ports?

[Battlefront.com]

What are the LOS/LOF check clock timings? I think knowing those might help some of us experiment with our micromanagement skills.

For instance, in relation to timing a turn around a corner? (I want them focused)

Units randomly request LOS/LOF checks so that they aren't all asking for checks at one time.  Once a unit has requested a check it has to wait a length of time before it can request another one.  However, it continues tracking and updating LOS/LOF information of any unit it has already spotted, which allows it to monitor location and engage (if that's what it is supposed to do) between checks without any sort of penalty.  If the target is really close (I forget the range) the unit is temporarily allowed to request checks far more frequently to reduce the chances of being unrealistically blindsided.  The latter feature was added at some point after the initial release of CMBN because the nature of hedgerows and urban terrain lent itself to accidental knife fights fairly frequently.

Because the system is random there is no way to use it to your advantage through micromanagement because you have no idea when units check or don't check.

I mean why would they be looking through the optics at all?, an not  the vision ports?

Actually, now that I come to think of it... this is likely a major factor in this situation.  In fact, this is probably even more of a factor than the LOS/LOF check timing.

Look at the video again and you'll see a LOS/LOF line snapping and unsnapping to the Sherman and you hear excited crew chatter.  Normally these behaviors go along with a target being spotted, yes?  Yet in this case the Sherman isn't spotted.  Anybody curious about that behavior?  I mean, it is a little bit unusual, isn't it?  Even I was a little puzzled by it until someone mentioned the optics were damaged.  Put that into the equation, add in some knowledge about the game mechanics, and then everything makes sense.

A weapon can not shoot at something until the GUNNER of that weapon can see the target.  In this case it appears that other crew members, peering through vision ports, spotted the Sherman but the gunner was not so quick because what he's looking through isn't functioning as well.  Hence the Sherman not being meaningfully spotted right away.  The delay in this case is an abstraction of the gunner either switching to his own vision port and guiding the gun onto target or was guided in by the TC after initially bumbling around ("I don't see it!"  "It's 2 meters left! No your OTHER left!" :)).  This adds a few seconds to the targeting process.

OK, so here are the issues with the Panther acquiring a target instantly:

1.  The target was not previously known to the Panther

2.  Target partially obscured by dead, smoking PzIV

3.  Target partially obscured by foliage (at least one tree)

4.  The Panther is buttoned, which reduces situational awareness significantly

5.  The gunner's optics are damaged, further reducing situational awareness

These are all compounding factors which together created a situation where the Panther took a handful or two extra seconds to identify the target.  Once it did, it fired quickly and nailed the Sherman.  This despite the Sherman having the drop on the Panther for some reason (impossible to say why).

As far as I'm concerned we've gone from alarmist calls of "BUG!!!!" to justified shouts of "WOW!!!  No other wargame does that!!" all in one thread.  Well, at least that's what I shouted

Steve

 

Edited January 19, 2016 by Battlefront.com

[weapon2010]

Steve,

 I thought if one member of the tank crew spots the tank it will show in the game,whether the tank will fire is a different issue, Like when an unbuttoned TC spots another tank on a crest but the gun is too low to shoot.So when you say,"In this case it appears that other crew members, peering through vision ports, spotted the Sherman", should not the Sherman appear in the game ?and the panther just not be able to  fire due to all of the other factors you mentioned?

[weapon2010]

also in the video there seems to be some smoke/dust in front of the Sherman that has nothing to do with the burning tank, maybe its the muzzle smoke

[RockinHarry]

Haven´t read through all postings here, but some additional info I´d like to know would be the Panthers crew experience, suppression/morale state and commander rating. Knowing the same info for the Sherman could likely help. When buttoned up, the gunner would have little use for his optics anyway at such close range (magnification and possibly narrow field of view).

[Vanir Ausf B]

I mean why would they be looking through the optics at all?, an not  the vision ports?

Depends on what you mean by vision ports. From the Panther Ausf A forward there aren't really any forward-facing open air vision ports. The driver does have his vision port in the hull which can be opened and closed, although it has a protective glass block so is not immune to damage (this port was deleted in the Ausf G). The driver also has 2 fixed periscopes (changed to a single traversable scope in the Ausf G). The radio operator has his machine gunner's sight and a fixed roof-mounted periscope, although the periscope is not directly forward-facing in Ausf A and G. The loader has a single turret roof-mounted fixed periscope (Ausf A and G). The commander's cupola has open air ports on the Ausf D but these are switched to fixed periscopes on the A and G. Critically, the gunner only has his gunner's sight, unless you count the pistol port on the turret side.

Which of these are possible to damage in the game and exactly what effects that has is unknown to anyone outside BFC. All that is known for sure is that it is impossible to damage optics to the point where the tank is unable to fire it's weapons and that optics damage does increase average spotting time.

Edited January 19, 2016 by Vanir Ausf B

[Pak40]

This is like saying that "going by the odds I played Powerball this week"   By this point in CM's lifespan, there's pretty much no bugs left to find in the underlying systems.  At least not ones that really matter.  Now, is there a "corner case" here that theoretically could use a bit of specific massaging?  Perhaps, but it's not worth pursuing unless the issue comes up more than once in a blue moon.  We only have so many hours in the day, so why spend it trying to fix something that might come up once every couple of thousand games when we could instead spend the time making something that may positively affect every game played?

No logical reason?  You must have missed a bunch of posts above because there's plenty of logical reason for this happening.  There's also the LOS/LOF check clock that I mentioned.  I don't know why you haven't picked up on that because Vanir Ausf B came in late to the discussion and has a pretty good grasp of the whole discussion.

 

 

Yup, that's a very good summary except.  Add to that the hardware limitations which impose a limit on how frequently LOS/LOF checks can happen and that further reduces the reasons to suspect "bug" or even "flaw".  I think the worst possible situation here is that we have a "limitation" due to hardware realities.

Steve

 

I saw the posts, Steve, None of it was logical to me with the current visuals that are present in the videos supplied:

1. The tree is obviously not blocking LOS. If the tree is blocking LOS in the underling code, then something needs to be tweaked with how the trees are rendered. Although it's not a bug by coder's definition there's definitely something wrong in that video, i.e. not realistic, not intended in programming. 

2. If smoke from the KIA tank is blocking LOS then something needs to be tweaked with how the smoke is rendered.

3. If the LOS/LOF clock is the issue then please explain how it took 44 seconds in the second video for the Panther to acquire LOS. I can understand the Panther not acquiring LOS with 7 seconds or whatever the clock is but I know it's not a 44 second clock. And this is after 4 rounds fired by the Sherman which should bloody give it's position away.

[edit] Steve has since explained the LOS clock system below..

4. Optics being knocked out: I'm guessing that an unsuppressed crew, even if buttoned, will be able to spot a big tank 30 meters directly in front of it. Optics wouldn't help at 30 meters anyway. Using vision slits shouldn't take much time for the gunner to acquire a close tank.

 

So a buttoned Panther with damaged optics takes 15 seconds to spot a Sherman that is partially obscured by a burning vehicle and foliage? Am I missing something? It's an edge case but not really that hard to believe.

It's completely believable situation but the visuals in the game don't show it. The tree is hardly blocking any part of the tank. It suddenly pops into view as if by magic.

I've played this game and CMx1 long enough to know that many things in these games are abstracted but USUALLY things such as smoke and foliage give you visual clues that they are blocking LOS. In this particular case though, there isn't a shred of visual evidence that would explain why this tank could not be seen. If the visuals matched the outcome then I'd be fine with everything. I'm sure the OP  would be too.

A weapon can not shoot at something until the GUNNER of that weapon can see the target.  In this case it appears that other crew members, peering through vision ports, spotted the Sherman but the gunner was not so quick because what he's looking through isn't functioning as well.  Hence the Sherman not being meaningfully spotted right away.  The delay in this case is an abstraction of the gunner either switching to his own vision port and guiding the gun onto target or was guided in by the TC after initially bumbling around ("I don't see it!"  "It's 2 meters left! No your OTHER left!" :)).  This adds a few seconds to the targeting process.

OK, so here are the issues with the Panther acquiring a target instantly:

1.  The target was not previously known to the Panther

2.  Target partially obscured by dead, smoking PzIV

3.  Target partially obscured by foliage (at least one tree)

4.  The Panther is buttoned, which reduces situational awareness significantly

5.  The gunner's optics are damaged, further reducing situational awareness

These are all compounding factors which together created a situation where the Panther took a handful or two extra seconds to identify the target.  Once it did, it fired quickly and nailed the Sherman.  This despite the Sherman having the drop on the Panther for some reason (impossible to say why).

As far as I'm concerned we've gone from alarmist calls of "BUG!!!!" to justified shouts of "WOW!!!  No other wargame does that!!" all in one thread.  Well, at least that's what I shouted

Steve

 

That's the best explanation yet but it still doesn't make much sense when you say a "handful or two extra seconds" because the second video clearly shows it took at least 44 seconds! Also you forgot to factor that the Sherman fired 4 rounds which should drastically help the Panther spot the tank, no? 

Whether it's a "by the book" definition of a bug or just some values that need "massaging", something in that video just isn't right.

Edited January 19, 2016 by Pak40
new posts

[eniced73]

This is like saying that "going by the odds I played Powerball this week"   By this point in CM's lifespan, there's pretty much no bugs left to find in the underlying systems.  At least not ones that really matter. 

Steve

 

http://community.battlefront.com/topic/96083-tree-force-fields/?page=3

Really?  I still cannot seem to shake this one.  No matter how much Jack I drink to kill it.  Still there!

 

 

[Vanir Ausf B]

 

4. Optics being knocked out: I'm guessing that an unsuppressed crew, even if buttoned, will be able to spot a big tank 30 meters directly in front of it. Optics wouldn't help at 30 meters anyway. Using vision slits shouldn't take much time for the gunner to acquire a close tank.

Keep in mind, there really are no vision slits for the gunner.

To reiterate, the only proven effect of optics damage in-game is to increase spotting time. Given that spotting times in CM are very random to begin with it's difficult to draw conclusions from single events. I can put two undamaged tanks 1 km from each other and in one instance a tank may spot the other instantly, and in another instance it could take several minutes.

If the contention is that a tank should never ever take 44 seconds or however long to spot another at x distance then that would probably require a significant change to the CMx2 spotting model to make it more deterministic, either through a general reduction of the random variables or the placement of hard caps on spotting time.

Edited January 19, 2016 by Vanir Ausf B

[Battlefront.com]

Steve,

 I thought if one member of the tank crew spots the tank it will show in the game,whether the tank will fire is a different issue, Like when an unbuttoned TC spots another tank on a crest but the gun is too low to shoot.So when you say,"In this case it appears that other crew members, peering through vision ports, spotted the Sherman", should not the Sherman appear in the game ?and the panther just not be able to  fire due to all of the other factors you mentioned?

Yes, I do believe it should have popped up.  This is where I do believe there is a "corner case" involved where the crew is almost getting an ID on the tank but it isn't quite strong enough for about 3-4 cycles.

also in the video there seems to be some smoke/dust in front of the Sherman that has nothing to do with the burning tank, maybe its the muzzle smoke

I saw that too but I'm not sure if it would have an affect or not.  But it could be yet another factor.

 

I saw the posts, Steve, None of it was logical to me with the current visuals that are present in the videos supplied:

 

Which is part of the problem.  I've been doing this for a very, very long time and I can't tell you how often there's something not in the video that's relevant.  A picture might be worth a 1000 words, but pictures can still be misleading.

 

 1. The tree is obviously not blocking LOS. If the tree is blocking LOS in the underling code, then something needs to be tweaked with how the trees are rendered. Although it's not a bug by coder's definition there's definitely something wrong in that video, i.e. not realistic, not intended in programming. 

Again, you are speaking with waaaaaaaaaaay too much certainty and not taking into account other factors.  The visuals in Combat Mission are pretty darned close to 1:1, but they are not perfectly 1:1 because it's not possible for computers to handle.  When things don't make visual sense the first thing to do is try to figure out what factors might be at play and not presume "bug".  As I said, we've had you guys pounding on this game for a very long time.  The chances that there's any significant bugs remaining with the underlying system is quite low.

In this case it is necessary to remember that LOS/LOF is drawn between center mass of target and center mass of spotter.  I can not tell form the video but it is possible the tree is in exactly the wrong spot, so to speak.  There is definitely code in the game to prevent a tank being able to hide behind a tree, but with the other factors mentioned it could be tipping the balance.

 2. If smoke from the KIA tank is blocking LOS then something needs to be tweaked with how the smoke is rendered.

Ideally, sure, but you have to accept some visual abstractions for the game to work at all given CM's scope.  Remember, thousands of things going on... not just a couple of tanks or a couple of guys.  We do not have the luxuries that World Of Tanks or Battlefield have, even if we had millions to invest in programming.

 4. Optics being knocked out: I'm guessing that an unsuppressed crew, even if buttoned, will be able to spot a big tank 30 meters directly in front of it. Optics wouldn't help at 30 meters anyway. Using vision slits shouldn't take much time for the gunner to acquire a close tank.

See previous explanation.  Plus, keep in mind that the Panther *DID* spot the Sherman.  It simply wasn't instantaneous.

 

It's completely believable situation but the visuals in the game don't show it. The tree is hardly blocking any part of the tank. It suddenly pops into view as if by magic.

 "Popping" into view is the only way to have Relative Spotting.  You either see it or you don't, there's no possibility for an inbetween condition.

I've played this game and CMx1 long enough to know that many things in these games are abstracted but USUALLY things such as smoke and foliage give you visual clues that they are blocking LOS. In this particular case though, there isn't a shred of visual evidence that would explain why this tank could not be seen. If the visuals matched the outcome then I'd be fine with everything. I'm sure the OP  would be too.

Yet I looked at the same video as you did and came to very different conclusions.

 

That's the best explanation yet but it still doesn't make much sense when you say a "handful or two extra seconds" because the second video clearly shows it took at least 44 seconds! Also you forgot to factor that the Sherman fired 4 rounds which should drastically help the Panther spot the tank, no? 

44 seconds might seem like an eternity, but it's not.  People have far, far, far too high expectations for what happens on a battlefield.  Imperfect stuff happens on battlefields all the time.  So yeah, if every single last encounter resulted in near perfect reactions then we, the developers, would have done a poor job simulating real life.  Which means that sometimes you will see things that don't look or seem right.  That's not necessarily grounds to yell BUG!

Think of it this way... if tanks were routinely not seeing each other for 44 seconds at this range then we'd have a problem to look into.  But here we are some 5 years after CMBN was released.  We've released several other CM games and lots of expansions.  Ask yourself, honestly, if this one situation is a mountain or a molehill.

As for the Sherman firing, you bet that is a factor in spotting.  For all we know the Panther would have taken even longer to spot the Sherman if it had not been shot at.

 

Whether it's a "by the book" definition of a bug or just some values that need "massaging", something in that video just isn't right.

I think there is more wrong with the interpretation of the video than what is actually in the video.

As for our definitions... BUG = something that is not working as it should.  I don't see any evidence to suggest that is the case here.  Is there something that could use massaging?  Perhaps, but it's not worth our time chasing down a single instance even if we had access to a save file.  Without a save file it is absolutely not worth bothering with.

Steve

[MikeyD]

I recall reading when Panther first saw action an after-action report urgently requested an additional wide-angle viewing periscope for the gunner to augment his high-power coaxial sight. That request was never acted on. One big disadvantage of a high magnification optics is its like peering through the world through a soda straw. Especially bad at close range. What's the magnification of the Sherman gunner's sight, x3? Plus the gunner also has a normal viewing periscope on the roof to help orient himself. Every little bit counts.

I recall some time ago a bug was found in Tiger I - the (invisible) buttoned TC was sitting sideways in his cupola not peering though his (invisible) forward vision slit. There's a LOT going on under the hood in this game that we can only guess at.

[Warts 'n' all]

It's those "How can you not see that?" moments that make this game so realistic, rather being a reason for wanting my money back. I've lost count of the amount of times I've ranted and raved at my men, only for my ravings to be followed by "Ah, right, that's why you couldn't see it".

[5th SS Div Wiking]

If the issue is one of CPU power, as has been suggested, maybe an obvious solution would be to get this game running on all the available cores of a given CPU? I asked about this in another thread and was told CM still uses only one core in any multi-core CPU. Give the game more CPU power and couldn't these LOS checks etc be ramped up such that situations like this don't occur?

I played the older games a hell of a lot in the past and it always made me scratch my head how a tank could frequently fail to see another tank but had an uncanny ability to spot a slowly crawling bazooka guy in a hedgeline or a spotter in a window, which it would immediately fire at, having previously ignored the MG crew (or whatever) in that same window. I always assumed these were game-balances intended to check the 'uberness' of bazookas and spotters.

If somebody would like to tell me how to extract the saved game-file I'll upload it to my server and then the experts here can grab and analyse it.

[5th SS Div Wiking]

It's those "How can you not see that?" moments that make this game so realistic, rather being a reason for wanting my money back. I've lost count of the amount of times I've ranted and raved at my men, only for my ravings to be followed by "Ah, right, that's why you couldn't see it".

The power of positive thought. Unfortunately I consider those moments to be so frequent that I've amused myself by thinking of issuing all my soldiers with white sticks. At least now the reasons have been explained: lack of CPU power vs LOS check-cycles. In effect all the soldiers in the game are partially sighted. And I have no desire to be able to unrealistically surprise an enemy unit any more than I want my units to suffer it themselves, it destroys the suspension-of-disbelief factor that is vital to games (the power of positive thinking not withstanding).

[Rokko]

Multithreading is not someting casual that you just turn on or off. It requires careful consideration and can cause some serious (logical) headaches throught an entire project and its code. Adding multithreading to a 10 year old engine that wasn't designed for multithreading in the first place is probably not feasible unless their programmer (I think it's really only one) decides to live in a cave for a year and do nothing else but code.

[5th SS Div Wiking]

Multithreading is not someting casual that you just turn on or off. It requires careful consideration and can cause some serious (logical) headaches throught an entire project and its code. Adding multithreading to a 10 year old engine that wasn't designed for multithreading in the first place is probably not feasible unless their programmer (I think it's really only one) decides to live in a cave for a year and do nothing else but code.

From what I've read about multi-threading in the past, I think you're right. But then we get into the territory of claims made for the game vs its relationship with modern hardware and if something isn't stated up-front a purchaser might be forgiven for reading the claims and making an erroneous assumption in that regard. I guess BF are caught between a rock and a hard place, because any solution to that issue is going to require considerable monies (I guess).

And let's be clear and fair on that issue. You cannot attempt to excuse a game's limitations by referring to lack of PC power when the game in question isn't able to take advantage of multi-threading. That's like excusing a race-car's failure to win a race by saying it was using only one cylinder in it's V4 engine.

Edited January 19, 2016 by 5th SS Div Wiking

[RockinHarry]

I recall reading when Panther first saw action an after-action report urgently requested an additional wide-angle viewing periscope for the gunner to augment his high-power coaxial sight. That request was never acted on. One big disadvantage of a high magnification optics is its like peering through the world through a soda straw. Especially bad at close range. What's the magnification of the Sherman gunner's sight, x3? Plus the gunner also has a normal viewing periscope on the roof to help orient himself. Every little bit counts.

I recall some time ago a bug was found in Tiger I - the (invisible) buttoned TC was sitting sideways in his cupola not peering though his (invisible) forward vision slit. There's a LOT going on under the hood in this game that we can only guess at.

Thought of that close range problem for high magnification, small FOV optics as well (despite the optics beeing "damaged" anyway). Can´t hurt to check good old CMAK game manual about optics modelling at times. So it was a veteran Panther crew (should have read the first page entirely ) and suppression from the dead Pz4 and repeated pounding from the invisible Sherman negligibly. Anything else that can be ruled out? A "-" (minus) rated tank commander maybe? I can´t get the info from the videos.

I´ve yet to stumble into any such game situation, but at last bits of Fubar is part of any war and any halfway realistic war game me thinks.