Something Very Wrong with LOS Through Trees

[slysniper]

Very well put JasonC.

I would assume we all know that the game does not come even close to all these aspects of how visual works.

I would be really amazed if someone did some day figure out some programming that could mimic what happens in real life when it comes to spotting things and how things can be hidden because of camouflage.

The game does a few things to catch a little bit of it, but it could always do more.

The game does factor in spotting units once they fire, and does it pretty good. It factors in moving units being spotted easier, I feel it does this semi good, but lacks at times.

But we know it does nothing for outlining, shadows  and these type of items. It would be interesting to see if they could program something that would take into account the terrain the unit was in and just make it harder to spot by that basic fact that it was in that terrain. Just because a line of sight can be drawn to it, the terrain should still be factoring in something as to it being seen or not seen.

At this point, this is where the game really lacks. (simple proof, a tank in open ground should have a modifier that makes it almost impossible to miss within the first viewing test of a turn.) But what do we have, tanks that can be invisible for way to long in open ground and that is at distances that make it a joke. So we have grumbled about this since this game engine has come out, but again, I do not see this changing until they try a new engine  and take a little bit of a different approach and try to program something that factors these things in.

 

Edited January 26, 2016 by slysniper

[Fizou]

Its a good idea to use one of the mods with semi transparent foliage. This will help you avoid situations like these. I cant recall any wtf moments since I did. 

[A Canadian Cat]

Oh now that is an interesting idea.  I never liked the look of that one myself but if it gives you less "why the face" moments then that gives it a new value.

[ASL Veteran]

From what I understand, the Original Poster was complaining because an enemy tank located in the forest spotted and destroyed his tanks that were sitting out in the open.  The issue being complained about is apparently not whether the SU85 could be spotted by the player, but whether the SU85 could spot and engage the player's tanks that were sitting out in the open.  The complaint being that tanks sitting in open ground apparently should not be spotted and engaged by tanks sitting in the forest because too many trees were in the way in the opinion of the original poster. 

[Michael Emrys]

This is fundamentally a camouflage effect, not a physical LOS blockage effect.  But it is very real.  And it causes a sighting asymmetry - it is far far easier to see something out in that snow covered field than it is to see something inside the tree line from the field.  Photons are intercepted or not, either way, so you might expect the LOS effect to be reciprocal and equal.  But it isn't, not with the camouflage effect included.

If a tank were sticking out of that far tree line it would certainly be visible.  But 10 or 20 yards back inside of it, not moving, and it would have great camouflage effect compared to the same out in the middle of the snow white field.  The tank inside the treeline would see the one out in the field ages before the reverse happened.

What the game seems to be underrating is this camouflage type effect.  Otherwise put, the spotting routines pick up marginally "visible" targets far too rapidly and reliably, in the absence of the movement or fire that actually attracts the eye and enables us to pick out a particular shape at range in such uniform visual environments.  The routine probabilities seem fine for large objects standing out against a skyline or a uniform field background of high contrast to the vehicle itself.  But that situation is an outlier of best spotting chances on real battlefields, not the rule.

These are the same points I was making in another thread a few days ago. What I have found over and over again in playing games on real terrain is that it is far easier for a spotter immediately behind some foliage to see out than for someone farther away to see in.

Michael

[Michael Emrys]

But programming that sort of AI thinking is beyond the scope of the overall game.

This reminds me of a thought I have had many times over the years, which is that as players of computer games we are pretty much stuck with whatever the designer/developer hands us. One of the advantages of paper games was that if a rule was found to be unsatisfactory, it was a relatively simple matter to change it, either in consultation with the original designer or without it. Win some, lose some.

Michael

[kevinkin]

That brings back memories of my school daze. You are right about board games Michael. For the current case, has anyone systematically looked at LOS through various forest types?

Kevin 

[simon21]

The human brain only needs ~5-9 photons to give a signal.  This game is too lenient.

http://math.ucr.edu/home/baez/physics/Quantum/see_a_photon.html

I totally agree with all you guys that spotting is an issue.  It's just a very hard issue to resolve successfully given how the game works.  To fix the issue an approximation would probably have to be made.  I know you guys probably won't like the reference, but in Warhammer 40k, back in early 2000's, models couldn't see through greater than 2" of woods.  This prevented certain forms of game-y play.  Down the road 40k went over to 'true line of sight' including a laser pointer with the logo on it, and this caused no end of issues with players claiming insane keyholes in both high and low level play.      

Basically, we want reasonable keyholes to exist, but foliage to block LOS.  I don't think that will be possible without putting bubbles around the end of tree limbs that will make close in shots that look good impossible.  Basically the short range game would take a hit.  No more shooting between two trees.  

 

Edited January 27, 2016 by simon21

[JasonC]

Board wargames can design for effect and get this sort of thing right pretty effortlessly.  The spotting rules in Jim Day's Panzer from GMT, for example, work extremely well, well enough that I've been "porting" them to older game systems that didn't include the right effects or give them the right strengths.  And yes board wargames are easier to "patch" - Panzer needs some "area fire" options are previous contacts, for example, that have been lost in the meantime because one side or other went prone, and similar.

As usual, the engineering approach that promises greater realism from bottom up modeling instead, doesn't actually deliver on that promise until it is nearly perfect in its handling of every relevant factor.  Get the photon interception stuff perfect (we aren't there with CM, as several posters above have pointed out, the issue being "transparent" concealment) and you still have to get the camouflage and object detection stuff perfect too, or you get worse not better realism than was available with a simpler "design for effect" approach.

I understand why CM is committed to the bottom up method, and that's fine.  It just carries with it a commitment to fix every issue anyone notices, and not to "plead" the endlessness of those fixes as a reason not to do any of them.  If you don't want endless fixes to get every piece right, you should stick with design for effect.  If you want engineering realism bottom up, then every time someone finds another issue, it has to be improved.  Don't expect to get off that treadmill soon.

[simon21]

They do break down and use 'design for effect' on some stuff.  There was a pretty big showing of it with the SMG's a few months ago.  

[Battlefront.com]

Good discussion.  Picking up on the last train of thought... design for effect vs. engineer for effect...

Design for effect is very predictable, but vastly less diverse because each and every real life possibility has to be specifically catered to (either supported or blocked).  The more variables you want the game to take into account, the more difficult design for effect becomes.

Engineer for effect is in some ways more predictable, in some ways less.  The object is to have the increased predictability lend itself to things which design for effect has problems with (e.g. small terrain elevation differences) while the stuff that is less predictable aspects should allow for rare battlefield experiences to happen even though they are not specifically planned for (e.g. keyhole LOS on a specific map between two specific points applicable only to specific unit combinations under specific conditions).  Although engineer for effect can handle vastly more variables and allow for far more diversity than a design for effect system could ever hope to achieve, it definitely is more challenging to handle the "corner cases" than it is with design for effect.

CM uses both approaches because it is absolutely impossible to have a simulation of tactical combat that is purely engineered for effect.  We'd need years and teams of programmers as well as a cloud of super computers to crunch the numbers.  Since we have neither, we absolutely have to do some design for effect work.  This is one of those situations.

The game does make some gross generalizations about foliage in order for it to work.  The big one is the same one that is pervasive throughout the game.  Specifically that LOS is not pixel by pixel, millisecond by millisecond.  There is no home computer on Earth that can do that and be a viable game.  Which means the LOS is determined by more-or-less the same sort of rules that one expects to see in a paper and dice game.  More sophisticated and nuanced, for sure, but inherently similar.

The way it works is the LOS line is "degraded" as it is drawn from point to point.  The more cumulative crap in the way the less strong the line becomes.  The quality of the spotter, the less restrictions on view, etc. give the line a higher starting value than a LOS line drawn from a unit with negative factors.  Some of the factors are specific (restrictions on range of view or height for example), others are general (optics of X type vs. eyeballs is the best example).  Each piece of terrain has ratings which determine how much the line is degraded when it comes to that piece.  At some point the line is so degraded that it is considered "blocked". 

The 3D graphics of the foliage plays absolutely no role in LOS drawing since that is simply impossible to do because the computers are totally not up for it.  This is the primary reason for potential disconnect between the visual representations in the game and the ability to spot.

LOF, which is not the same as LOS, is a little different.  LOF does pay attention to the pixels on the screen at least to some extent.  Foliage is one of the things that is a "no" except for the trunks.  This is possible to do because shots are fired in low enough quantities that there is enough CPU power to allow for more calculations.  Plus, LOF is a binary decision about if something should keep going or stop, not the sort of conditional quality stuff going on with LOS. 

Where we get into tricky situations is that sometimes there is LOS to a target and the LOF is not all that great.  There is logic in there to try and advise the shooter "yes, technically you can see the target and technically you can hit it, but it's a long shot and therefore don't give it a go".  The problem with this logic is where to draw that line?  How variable should it be?  These are very difficult questions to answer because there's a lot of variables to consider and those things can be weighted very differently depending on circumstances.  For example, if a Stuart has a really poor quality shot at the rear armor of a King Tiger, in real life that gunner would take the shot way more often than not because the chances of getting that sort of shot again are low and it's the only chance of destruction the Stuart has against the King Tiger.  But switch the two vehicles and the King Tiger gunner might roll the dice in favor of a better shot.  It's all very twitchy.

A design for effect game system doesn't have to deal with this sort of stuff as much because it is inherently less nuanced.  Like CM there is a line that has to be drawn if it is to be allowed at all.  Like CM it's tricky to know exactly where to draw that line.  Unlike CM a design for effect system has a much smaller range for tailoring the decision to conditions.  Either way, a paper and dice game leaves the decision to the player to determine if it is worth taking the shot or not.  With any type of computer game that doesn't have explicit player decision making on a shot by shot basis that decision has to be made by the AI and that means more programming and more lines drawn and more twitchy situations. 

Neither design philosophy is without it's drawbacks.  Overall CM is a far more diverse and exciting game system because philosophically it is engineered for effect.  The very high quality of the engineering is also why the game system is so engaging to people.  However, no system is perfect and so there are areas within the game that could be improved given various factors (time, computing power improvements, etc.)  We do intend to keep improving things as we go along.

Steve

Edited January 27, 2016 by Battlefront.com

[Cpl Steiner]

Thanks Steve for your detailed reply and for giving us a look "under the hood".

I totally take on board what you say about foliage and LOS. However, when it comes to fields of tall crops LOS seems to degrade extremely quickly - almost as soon as it reaches the field. From your explanation, is it right to say that tall crop tiles are weighted to give a higher degradation effect on LOS than tree tiles (ignoring for the moment seasonal effects on foliage)? If so, would a possible solution to LOS being traced too far through tree foliage simply be to give such tiles a greater degradation weight? It seems to me that crops block LOS very well in the game even though by your own explanation each individual ear or husk of corn is not being taken into account. Could the game not just treat tree foliage like a field of tall crops hovering just above the ground? Perhaps it does already, in which case maybe the weighting of how much it degrades LOS needs to be increased?

Just some thoughts on what might be a quick and simple solution to these sorts of issues (well, quick and simple compared to other ways of tackling it). I did mention this idea in my original post but I don't think anyone has so far responded to it. I also do accept that the example I posted might not have this issue as the trees are tall pines with foliage high off the ground, but I have definitely seen LOS traced a long way through forest canopy in various battles - far more than would be the case for a corn field.

 

 

Edited January 27, 2016 by Cpl Steiner

[Battlefront.com]

Thanks Steve for your detailed reply and for giving us a look "under the hood".

I totally take on board what you say about foliage and LOS. However, when it comes to fields of tall crops LOS seems to degrade extremely quickly - almost as soon as it reaches the field. From your explanation, is it right to say that tall crop tiles are weighted to give a higher degradation effect on LOS than tree tiles (ignoring for the moment seasonal effects on foliage)?

No, that is not correct.  Mostly because it is an overly simplified question which doesn't really have a straight forward answer since there is no such thing as "tree tiles".  There are a bunch of different tree species that can be mixed in various densities along with different combinations of ground conditions.

If so, would a possible solution to LOS being traced too far through tree foliage simply be to give such tiles a greater degradation weight?

Sure, all weights can be tweaked.  However, that doesn't mean there wouldn't be unintended consequences.  There most likely would be.  Which is why we've pretty much stopped mucking around with the weights several years ago (except for new stuff being added).

It seems to me that crops block LOS very well in the game even though by your own explanation each individual ear or husk of corn is not being taken into account. Could the game not just treat tree foliage like a field of tall crops hovering just above the ground?

We could make a single birch tree block LOS completely.  But that doesn't mean we should

Have you seen a wheat field at full height?  Visibility is about a foot even if you are standing right up against it.  This is not how forests work, therefore you are comparing two things together which should not be.

Just some thoughts on what might be a quick and simple solution to these sorts of issues (well, quick and simple compared to other ways of tackling it). I did mention this idea in my original post but I don't think anyone has so far responded to it. I also do accept that the example I posted might not have this issue as the trees are tall pines with foliage high off the ground, but I have definitely seen LOS traced a long way through forest canopy in various battles - far more than would be the case for a corn field.

The problem with your approach is that you view things as if they are fairly simple, but they are not.  While I do not doubt there could be some changes made to reduce some instances of seeing too far into a wooded area, I am sure they would not be as easy as tweaking a value here or there.  I suspect such simplistic number pushing would result in negative side effects elsewhere.  I suspect what is needed is a major modification to the LOS system to account for something that we currently aren't accounting for.  That isn't something we have on our planning board at the moment, though I have made a note that we should look into it at a later time.

Steve

[Cpl Steiner]

I have made a note that we should look into it at a later time.

Steve

Much appreciated Steve.

 

 

[Placebo]

Steve, always interesting to see how some of the game mechanics work in the game.

I wonder if some of these explanations/game mechanic insights should be added to the manual or held somewhere as a bring up guide for newbies.  The learning curve is steep so might stop some of the early frustrations of "why did that happen", "why does that not work". 

 

 

[RockinHarry]

Good candidate for making this a "pinned", as LOS questions pop up too frequently and this thread contains all what is needed to know.

[A Canadian Cat]

@Battlefront.com that was very interesting.  I do not believe we have seen that discussed before (I am referring to the discussion about the LOS being degraded as it goes).  Thanks

@Placebo putting that into the manual might cause too much information for my brain for people.  On the other hand it would be good to have it available some where.  I wonder if the wiki is still active...

[kevinkin]

I was thinking the same way guys. These technical explanations are always great reads but are scattered over the forums. It would be nice to have them in one place (read only). 

Kevin 

[Bulletpoint]

I think the major problem here is that the game treats the trees as semi-transparent, but to the player's eyes, they completely block LOS. This makes it nearly impossible to anticipate lines of sight across a landscape.

Of couse in real life, you also get surprised sometimes, but generally you have a good intuitive impression of what will be visible at any one point, because if you can't see through it, chances are your opponent won't either.

Many trees in real life block vision 100 pct. It's not like you sit and watch a huge oak tree for some minutes and suddenly it becomes more transparent and you spot a panzer behind it, then suddenly it reverts back to being nontransparent. Even in strong winds, you won't be able to see through the foliage.

Edited January 28, 2016 by Bulletpoint

[c3k]

Some trees near the shooter are ignored. This simulates the ability of the unit to optimize its location. Aiui.

[Michael Emrys]

Some trees near the shooter are ignored. This simulates the ability of the unit to optimize its location. Aiui.

That's interesting. Would you please tell us what is your source for this information?

Michael

[sburke]

Some trees near the shooter are ignored. This simulates the ability of the unit to optimize its location. Aiui.

I am not so sure that is true even if it is intended.  I have CMBS save with a sniper pointing his barrel right into a tree trying to target an enemy soldier.  He fires repeatedly and it ricochets off the trunk. If that behavior goes against what the game is supposed to do I can submit a ticket on it.  I can submit anyway as a request to change behavior.  Just need to know which it is.

[Battlefront.com]

Oh, that is most definitely not something we want happening.  You know the drill

The LOS/LOF issue gets into a really thorny problem of how much do we allow an individual soldier or vehicle to maneuver for a clear shot to a specific target.  Cascading unintended consequences that wind up with angry villagers with pitchforks and torches are the sorts of things we try to avoid

Steve

[c3k]

That's interesting. Would you please tell us what is your source for this information?

Michael

I haven't the foggiest notion of my source, nor am I sure it's correct, hence the "aiui". 

[Michael Emrys]

I haven't the foggiest notion of my source, nor am I sure it's correct, hence the "aiui". 

Sooooo...taken with your Me-262 misinformation, your credibility is taking some hits this week.

Michael