It's got to go.

[sburke]

http://www.battlefront.com/community/showthread.php?t=111250&highlight=attached+targeting+order&page=6

Ian found this out above which worked. BF deemed it gamey and eliminated it in a patch. I tried to search around but could not find any evidence of people complaining about this being a "gamey" feature. If anything I think they created a problem which you guys are discussing here. If they would have left well alone everything would not be perfect but satisfactory in my opinion. This little hint also allowed you to fire through that little invisible forcefield called "smoke". So if this is deemed gamey then using smoke should be eliminated from the game as being just as "gamey". Who was the idiot the complained on this being gamey in the first place? just curious as to why BF changed it as I did not see many people complaining.

That "idiot" might just be Steve - your price for RT just doubled

Actually I think the issue here wasn't this particular use, IIRC it might also have been tied to a mortar being set to a target then moved to where it no longer had LOS. Would you consider that to be gamey enough?

[eniced73]

That "idiot" might just be Steve - your price for RT just doubled

Actually I think the issue here wasn't this particular use, IIRC it might also have been tied to a mortar being set to a target then moved to where it no longer had LOS. Would you consider that to be gamey enough?

Guess I was a little harsh by name calling there. I apologize for that one. But NO, I would not consider that to be enough. I see more people complaining about the issue at hand rather than the mortar issue. In fact that is the first I heard of it. Would I be worried about playing an opponent that used this gamey tactic. NO. I really do not think that would turn the tide of a game. Just my opinion. I vote to return to the original way seeing that the AS limitation is not going to be solved anytime soon.

Oh. Would you consider putting down smoke on a street not allowing my tank to fire through it gamey? If you would then Ian's find would be useful to counter this.

[Vinnart]

This is where I think a 1:1 action square (2m x 2m) system like Sudden Strike used would eliminate this problem. I also see other advantages of flexibility in squad formations too in such a system. Sudden Strike was a less complex game, but had many similarities to CM in structure. Both games use layers of action square grids to control the game although CM’s 8m x 8m is bigger. Perhaps CM with current hardware is too complex at this time for a 1:1 action squares, but it is fun to think of the possibilities.

[A Canadian Cat]

to a mortar being set to a target then moved to where it no longer had LOS. Would you consider that to be gamey enough?

Yep, exactly. More accurately have your mortar safely behind your lines give it a move order near where you wanted to hit, issue the target order then change the move order to right next to where the mortar already is. Instant mortaring of any where on the map. No delays, no spotting rounds. Yikes!

Even I did not notice the connection between using my technique and what you could do with mortars. I payed around in testing and it was actually pretty scary what you could do. Someone actually PM'ed me and pointed out the problem and I just stopped mentioning the technique after that. Made me sad:( but leaving it in would have just been crazy.

[A Canadian Cat]

This is where I think a 1:1 action square (2m x 2m)

But that is still 16 times bigger LOS tables plus many more combinations of tests. Just the kind of heaving computing load BFC are trying to avoid.

[Battlefront.com]

That is not what he means. He is suggesting that buildings always have a "unit-type thing" that can be targeted. So when faced with an unobstructed view of a building you could target multiple points on each floor on the face of the building instead of targeting the centre of the building (which I guess you could so as well).

Then this is akin to what you suggested and what I originally designed back in 2004. Not practical.

Which is the problem.

I know

You have expressed the issues with computational power etc that mean your options are limited to improve things. I get that. We are just shooting ideas around in the hopes that it inspires a way to solve the problem. I know, wishful thinking, since we don't have the detailed knowledge of how the system works.

It is highly doubtful that there's a clever way around this. As I've said, I recognized this problem before the game was even coded. It was coded with the understanding that this would crop up. It's been 8 years since people first experienced it within the game itself (CMSF Beta). We've smoothed the edges a bit, but there's no simple solution for it.

As computers get faster and people start having 8GB of RAM standard, maybe something will happen. But it will likely not happen until then. Even tiny changes to this part of the system involves potentially exponential increases in resource usage.

Thanks for spending some time thinking about this and sharing with us some of the issues.

No problem.

Steve

[Battlefront.com]

Guess I was a little harsh by name calling there.

Well, if by "idiot" you meant someone who has been designing computer games for 20 years and is extremely well versed in what constitutes an abuse and what constitutes a legitimate game function, yet spends his precious time on this earth catering to an extremely small number of gamers who complain at the drop of a hat EVEN when they are happy... then I definitely am an idiot. I'd say I'm a HUGE idiot. If, on the other hand, you mean someone that blindly flails around adding or subtracting features seemingly without a clue what he's doing, then I do beg to differ and instead accept your apology.

But NO, I would not consider that to be enough. I see more people complaining about the issue at hand rather than the mortar issue. In fact that is the first I heard of it. Would I be worried about playing an opponent that used this gamey tactic. NO.

You should be if you are concerned about two things:

1. Realism

2. Not falling victim to someone who is willing to "cheat" to win

No matter how you slice it, that feature was a cheat. I can say that for sure because the game system is founded on the principle of limiting unreasonable "gamey" micromanagement. Therefore, the minute it was pointed out that you could get LOS information from a point where your unit isn't, it was an instant no brainer to pull it.

Fundamentally units in real life HAVE NO CLUE what LOS is from a given point to a given point until they are right there in person. Allowing people to place a waypoint in the EXACT optimal position to within a fraction of a meter, perhaps hundreds of meters from where the unit is, absolutely without a doubt is "gamey". It isn't just about the mortar thing, it's about EVERYTHING. To argue that a unit would have this sort of second sight is to lose the argument before even starting.

Oh. Would you consider putting down smoke on a street not allowing my tank to fire through it gamey? If you would then Ian's find would be useful to counter this.

This is a whole 'nother ball of wax. In short, the player has too much information about what/where the units behind the smoke are. He can, therefore, basically neuter the effects of smoke to a degree that is totally unrealistic. So we do not allow people to shoot through smoke because if we did that we might as well not have smoke at all.

Trying to work around the inherent flaws of having a single player in control over the game, with the ability to second guess everything he does ahead of time, is an extremely difficult and definitely imperfect art. The smoke example is definitely more questionable as there is a legitimate argument against the behavior we have. There is no legitimate argument in favor of precise and exact LOS information ahead of actually being there. Or more exactly, whatever argument someone could make I can (and have on more occasions than I can count) crush it with counter arguments.

This is where I think a 1:1 action square (2m x 2m) system like Sudden Strike used would eliminate this problem.

No, it would make CM unable to run on your computer Even going to something smaller like 6x6 would likely crush your computer. Plus, the "oblique" problem would still exist, but it would likely be extremely rare.

Sudden Strike was a less complex game, but had many similarities to CM in structure.

A Yugo has a lot of similarities to a Ferrari, but I know which one I'd rather own and drive. Er, if someone else was paying for it of course

As far as I know there is no other game out there that has an ELOS system. Heights are not taken into consideration directly, but instead through modifiers. If we didn't have to take various heights into consideration we could go with a smaller grid. However overall the quality of play would suffer big time as the numbers of problems complained about would skyrocket. I can assure you of that because CMSF initially didn't have ELOS (v1.05 introduced it, IIRC) and it was absolute misery trying to round off all the rough edges that resulted from a single fixed height.

Steve

[sburke]

Yep, exactly. More accurately have your mortar safely behind your lines give it a move order near where you wanted to hit, issue the target order then change the move order to right next to where the mortar already is. Instant mortaring of any where on the map. No delays, no spotting rounds. Yikes!

Even I did not notice the connection between using my technique and what you could do with mortars. I payed around in testing and it was actually pretty scary what you could do. Someone actually PM'ed me and pointed out the problem and I just stopped mentioning the technique after that. Made me sad:( but leaving it in would have just been crazy.

Exactly and I think that is sometimes the problem with ideas we come up with. They may be great, but they also may have unforeseen side effects that are worse then the problem they are intended to resolve.

For what it is worth eniced73, I agree, I would love to be able to fire into smoke. It is frustrating in that smoke isn't just concealment, but also effectively cover now. There are some things you can try, but overall nowhere near what you should be able to do.

Still the mortar issue was way over the top and apt to have far more impact in the game than not being able to fire into smoke.

EDIT

This is a whole 'nother ball of wax. In short, the player has too much information about what/where the units behind the smoke are. He can, therefore, basically neuter the effects of smoke to a degree that is totally unrealistic. So we do not allow people to shoot through smoke because if we did that we might as well not have smoke at all.

Steve

Huh, didn't really think of it from that angle. Good point.

[Battlefront.com]

Huh, didn't really think of it from that angle. Good point.

And probably you didn't also consider that the units popping the smoke can not put covering fire through it to the other side either It might be unrealistic at the lowest level, but it is at least even handed.

When one goes to design something, anything in fact, the first thing to do is figure out what you want to achieve with your design. The next thing you do is identify all limiting factors that are beyond your ability to change. Then you start designing your way towards your goal while always keeping your limiting factors in mind. As you go you probably find more limiting factors you hadn't thought of, which then causes you to adapt the design to address them. The more complex the design goals, the higher the bar being set, the more likely you'll have to compromise along the way.

With a realistic wargame you start out with some of the absolute worst factors to work around. One person in charge of thousands of soldiers and vehicles simultaneously? Yeah, that's a big one. A player that doesn't have to worry about demotion of physical injury? Another one. A player that has access to perfect intel (like the map) before he makes any decisions of note? Yup, that's one too. A player that can... well, the list goes on and on.

Many of the design decisions we make are not necessarily defendable in a vacuum, but are never-the-less helping the game achieve it's overall goal of realistically portraying tactical warfare. Since the overall goal is what we're all interested in, our design decisions must be taken within that context.

Steve

[Vinnart]

What you say there Steve regarding design is very true. There are always compromises to be able to achieve a happy medium in all designs no matter what it is including creation.

A classic product example is the Sansa Clip MP3 player I got not long ago. The cost is $35 so affordable to most. It sounds very good, It is very light, compact like the size of a car’s remote, clips to everything, expands to 32g, great features especially for one who enjoys audio books. It allows one to pause one chapter then go to another, then go back and pick up where you left off. Sounds perfect right? Well, the cons to being compact are the screen must be small, and a little tough to read, and must be held a certain way by the finger tips, and can slip out of your hand, and is not as loud and doesn’t sound quite as good as a bigger Sony Walkman I have. $35 is not as good as $10.

So, there will always be cons to achieve an overall good design. This is the thing people should realize to understand the design process. It can’t be prefect in EVERY way. The good thing though to look forward to when it comes to technology based things though such as computers is the constant increasing of power. Can you imagine what could be done with a game concept such as CM on a computer 25 to 50 years from now? It is exciting to think about.

[Michael Emrys]

OT (but I seldom let that stop me): I just can't figure people wanting to watch movies on their smart phones. Do such creatures actually exist? The phone makers and service providers apparently think so.

Michael

[womble]

I asked:

So how can a unit spot a unit in an AS which it can't fire at unless there's a unit there? [puzzled]

...as I remember it, this isn't possible.

What can happen is Friendly Unit has knowledge of Enemy Unit because either they were previously in two Action Spots that could see each other or the information was passed on by another unit. In either case there is now a "special relationship" between the two units which supersedes the Action Spot center requirement. As soon as the two units break contact that relationship disappears and now it's purely Action Spot related.

Is that special relationship created bilaterally: if unit A in an occluded building AS can see enemy unit B, does that make A eligible for spotting by B?

It cross references all the heights.

...When the soldier is prone (Height 0) the LOS Map is consulted...When the soldier pops his head up the LOS Map is consulted again, but using a different criteria (Height 1) for determining what can be seen...the calculations are not done on the fly, only the decision which set of data to use is. Huge difference in terms of processing time.

Aha. Thanks. That's given me a handle to grasp on the lifeboat of understanding. The LOS map has a z axis. Units' eyes are not initially used, only the LOS map. That explains clearly (if I've got it right, anyway).

And just to expand slightly for practical purposes: when you're tracking the Targeting tool across the map, LOS is being looked up using the centre of the AS your unit is in, at the (possibly approximate; there can't be LOS table entries for the height of every vehicle vieing port... can there? ) current height of the eyes of your unit at one end, and the centre of the AS at ground height at the other end.

No, that can't be complete, because I can set an area target to one portion of an AS and the target line anchors to the centre of that AS which showed as "no target" when my cursor was over it... How can I modify the above statement to account for that feature?

[lofty12]

Very interesting discussion. It is a bit disappointing that there is a technical solution that would help solve the main issue in this thread but because battlefront deem there are enough people with old systems it wont be implemented.

That's not a criticism of battlefront by the way, just the frustration of someone who has a mid to high end system. Why don't battlefront do a forum survey to find out exactly what level of hardware most of the playerbase have? Maybe you would be surprised.

On a side note,

I made a thread last week asking if it was possible to check los from a future waypoint and was told that it was possible. However, earlier on this thread steve stated that this was a gamey feature and was removed.

Is this possible or isn't it?

[akd]

You can check LOS from a waypoint, but unfortunately the line still draws from the unit. The information shown, however, is for waypoint to target.

What was removed was being able to place an area target from a waypoint, then move the waypoint without cancelling the target.

[Rocky Balboa]

There is something called an "LOS Map" which is precalculated when the map is made and it is loaded into RAM for rapid access during the game. This basically tells CM, ahead of time, which Action Spots can see which Action Spots. There are *no* on-the-fly calculations going on. And that's why your Target tool works snappy under all circumstances.

Steve

Interesting ....

Steve, have you given any thought to adding FOW terrain which essentially would block out action squares based on the LOS map? It seems to me that this could alleviate a lot of the confusion surrounding this issue. In addition I think it would also help a player visualize terrain relief as well as eliminate the gamey use of object deformation (ie ... broken fences and walls ) to track otherwise unseen movement ...

[Battlefront.com]

Is that special relationship created bilaterally: if unit A in an occluded building AS can see enemy unit B, does that make A eligible for spotting by B?

Absolutely. CM is coded, from top to bottom, to be a slave to reciprocal behavior. At least inherently. Battlefield conditions, however, mean that spotting is situationally dependent. The rare times when someone claims the system is "broken" are getting confused by inherent possibilities vs. what is appropriate for the given situation. We had a lot of problems with such perception problems early in CMx2's life, but we seem to be past that now.

Aha. Thanks. That's given me a handle to grasp on the lifeboat of understanding. The LOS map has a z axis. Units' eyes are not initially used, only the LOS map. That explains clearly (if I've got it right, anyway).

I forget the technical term for this, but it's kinda like a progression of tests designed to quickly deflect as many queries as possible, as quickly as possible, that do not merit more intensive calculations. The first one, in this case, is "Can the Action Spot you are in see the Action Spot being pointed to?" The answer "no" stops the analysis right there, which is good because CM is asked this question probably millions of times during the course of a game. If the answer is "yes" then a whole series of progressively more specific questions are asked before arriving at a definite conclusion of exactly to what extent a unit in Action Spot X can see something in Action Spot Y.

And just to expand slightly for practical purposes: when you're tracking the Targeting tool across the map, LOS is being looked up using the centre of the AS your unit is in, at the (possibly approximate; there can't be LOS table entries for the height of every vehicle vieing port... can there? ) current height of the eyes of your unit at one end, and the centre of the AS at ground height at the other end.

Not quite. It's from the eyes of your unit to a fixed position of 1m off the ground.

No, that can't be complete, because I can set an area target to one portion of an AS and the target line anchors to the centre of that AS which showed as "no target" when my cursor was over it... How can I modify the above statement to account for that feature?

If I see the circle with the line through it, I can't get the line to stick.

Very interesting discussion. It is a bit disappointing that there is a technical solution that would help solve the main issue in this thread but because battlefront deem there are enough people with old systems it wont be implemented.

Whoa... that's not quite true There MIGHT be a way to reduce the number of oblique angle problems for higher end systems at the expense of a lot more RAM and hits to the framerate. At least in theory there are. But whether it's actually practical to do it, we don't know and we're not going to try any time soon because for sure we'd be cutting out large swaths of our customer base.

Why don't battlefront do a forum survey to find out exactly what level of hardware most of the playerbase have? Maybe you would be surprised.

Because the Forum participation represents a tiny fraction of our total customer base and is likely to be disproportionally "higher end". It's like asking a dining room full of people at a resort hotel if they use food stamps (or some other government assistance program). The answer coming back wouldn't surprise me at all and it wouldn't be a good basis to make a decision that affects hundreds of thousands of people.

I made a thread last week asking if it was possible to check los from a future waypoint and was told that it was possible. However, earlier on this thread steve stated that this was a gamey feature and was removed.

Sorta. AKD is right, but there's some limitations on what he said. You can click a Target from a Waypoint and you will get LOS information so that you can cause the Target to lock in. This is a little bit of a "cheat", but it is necessary because the system doesn't allow you to click on things you don't have LOS to (though I suppose Charles could have changed that). It would be very frustrating to have clicks ignored until you found a legitimate target spot. However, what you can not do any more is click down a Target Order and slide the Waypoint around to optimize the movement position relative to the target.

Think about it this way. The old system allowed you to target a specific area and then allow you to select the spot from which to shoot. That's definitely "gamey". Now you have to select the spot from which to shoot and then the spot to target. For sure a determined player can futz around to find the most optimal position to fire from to a specific spot, but now it's a PITA to do. Hopefully that is enough to keep this sort of "gamey" behavior down to a dull roar. Certainly makes it impractical for RealTime play. And it also takes care of the indirect fire problem mentioned.

ISteve, have you given any thought to adding FOW terrain which essentially would block out action squares based on the LOS map?

It was, in fact, part of the original design for CMx2. So the answer is "yes" Unfortunately there are a HUGE host of unintended consequences that come from such a design. The biggest one is how do you plot units "into the black" if you can't see the terrain? In real life terrain is revealed as you move and you make adjustments as you go. That is not possible to do in WeGo and it is practically difficult to do in RealTime (i.e. because you can't babysit each unit). Plus, then we'd have to account for maps and oral descriptions giving the specific moving unit some hints of the terrain in front. For example, "when you come to a fork in the road, head left and that takes you to the bridge". Not possible to convey that knowledge to the player when the terrain is blacked out.

And that is only the pointy tip of the iceberg. It gets much worse the more you think about the mechanics of it.

Oh, and no way to hide the map contours. That's technically not possible to do. Which means the blacked out terrain already has a huge limitation before you've even started looking at the specifics.

Steve

[womble]

Not quite. It's from the eyes of your unit to a fixed position of 1m off the ground.

The actual eyes? Not the generic AS LOS map 3D matirx height?

I asked:

... I can set an area target to one portion of an AS and the target line anchors to the centre of that AS which showed as "no target" when my cursor was over it...

If I see the circle with the line through it, I can't get the line to stick.

Naturally, if the cursor is that icon. The case I'm offering, though, is when you're mousing around looking for a place to aim, you get the yellow "Area Target" cursor, click, and the targeting line (as expected) anchors to the nearest AS centre. Yet, if you hit "Target" again, and hover over the end of the drawn targetting line, it says "No LOS" and has the red line through it. Your troops will fire according to the line; it's not some chimeraical thing like shooting through the edge of foliage...

I was thinking about this again earlier. I have another situation that occurs often which I don't comprehend how it converts from "No LOS in the LOS map" to "Two units spotting each other", viz:

• Houses lining either side of a street.
  
• One side of the street there's an infantry unit static, looking out of the windows facing the street.
  
• The other side, an infantry unit moves into a building obliquely across the street
  
• If you put a straight edge between the two AS centres, the party walls of the adjacent buildings would prevent any LOS between the two buildings.
  
• Neither unit has been seen by any other opposition unit.
  
• How do those two units end up spotting each other stood in the windows if the LOS map says the two AS can't see each other?
  

[Battlefront.com]

The actual eyes? Not the generic AS LOS map 3D matirx height?

Not sure what you mean. There are 5 ground heights in the game in terms of LOS, LOF, and Spotting:

Prone

Kneeling

Standing/Small Vehicle

Tall Vehicle

Very Tall Vehicle

These are the only values that are used because to do anything other that would fudge up everything.

As an aside, other games use modifiers to simulate height differences. MUCH quicker that way, but also not nearly as rich and robust as more points.

Naturally, if the cursor is that icon. The case I'm offering, though, is when you're mousing around looking for a place to aim, you get the yellow "Area Target" cursor, click, and the targeting line (as expected) anchors to the nearest AS centre. Yet, if you hit "Target" again, and hover over the end of the drawn targetting line, it says "No LOS" and has the red line through it. Your troops will fire according to the line; it's not some chimeraical thing like shooting through the edge of foliage...

Yes, in EXTREMELY limited, marginal circumstances where a partial meter here or there makes a difference between Partial LOS and No LOS the system SOMETIMES allows you to pop a point down and have it snap to the center of an AS that you can not directly target. This is because the system is sophisticated enough to understand marginal circumstances and "round up". But if you try to go directly to that center point you MIGHT not get it.

The LOS map is sophisticated enough to take into consideration micro elevations within an Action Spot when determining how well a specific Action Spot can draw LOS to a specific Action Spot. But again, this is determined ahead of time. You can always draw a Partial LOS to that particular Action Spot and that's that.

I was thinking about this again earlier.

You are definitely thinking about this too much

How do those two units end up spotting each other stood in the windows if the LOS map says the two AS can't see each other?

Same as above. It's an extremely marginal situation which could go either way. It's just enough for units to spot each other, but not enough to effectively target.

If you want to think about this in real world terms, one guy is peeking out the window and catches sight of some motion going into the neighboring building. Maybe he sorta sees a sleeve of one guy. But it's such a difficult shot as to be ergonomically improbable shot that he only thinks to take it when there's a known reason to do so.

You really need to stop over thinking this. Seriously. There's nothing going on in the game engine now that can address the oblique area fire problem. There is no inconsistency within the paradigm that handles LOS and LOF. There are, however, marginal circumstances which are not 100% perfect due to the limitations we are working with.

Honestly, you shouldn't be wondering how CM can do things better, but instead wonder how it can do what it does so well on a PC. At least that's what a programmer would think

Steve

[womble]

This is going to be my last post on the subject.

Not sure what you mean. There are 5 ground heights in the game in terms of LOS, LOF, and Spotting:

Prone

Kneeling

Standing/Small Vehicle

Tall Vehicle

Very Tall Vehicle

I'm having real difficulty reconciling some of the things I think you've said here and in the past. I'm certain it's down to communication, and it really isn't worth either of our times trying to clear up the disconnects.

Yes, in EXTREMELY limited, marginal circumstances...

Which occur several times in most games I play...

The LOS map is sophisticated enough to take into consideration micro elevations...

I don't think I'm (only) talking about micro elevation.

But again, this is determined ahead of time.

Yes, I've got that. What I don't get is how "there can never be LOS between these AS" from the LOS Map gets converted to all the myriad cases (and they certainly aren't limited in number, though they're probably often marginal, since "margins" are where engagements can be won or lost, so players seek them as a matter of routine) where, actually, if there are units there, they can shoot at each other.

Honestly, you shouldn't be wondering how CM can do things better, but instead wonder how it can do what it does so well on a PC. At least that's what a programmer would think

Obviously it's escaped your attention that I've been asking that very question. "How does it work?" so that I can understand it. I've stopped suggesting ways to improve (this aspect of) the game, I'm picking holes in what I understand about it, since it's obvious that my previous understanding was inadequate to the cause (and that remains the case). This medium isn't going to change that in any practical timescale.

[A Canadian Cat]

This is going to be my last post on the subject.

Yeah, I think we have read the knowledge and understanding that you have available to give. Thanks for your explanations and entertaining our hair brained schemes (that, it turns out, you have already thought about ). It seems that there is not much to be done and that is too bad because it is the one place where I feel a disconnect with the game.

Which occur several times in most games I play...

I would not quite go as far a most games but the thing is if I play a game that has any kind of built up area then it occurs turn after turn. For example right now I am playing Men with Suspicious Hats and running into this problem of not being able to area target the face of buildings that my men can fully see (not partial but fully). They cannot see anything but the face of the building so the AS is not visible. But the enemy in the front of the building can see and fight my men in the street and then duck. Fighting in the streets becomes very difficult because I have to send men out into the street without proper suppressive fire against known enemy locations (by that I mean we saw guys in the windows 5s ago). Let alone add some speculative fire to other building fronts.

So, while I understand that there is nothing practical that can be done at the moment, please understand that this issue is not marginal or rare. I hope one day you can make modifications that will make progress.

[Battlefront.com]

This is going to be my last post on the subject.

I'll follow up, and of course you can do the same if you think it's worth the time.

I'm having real difficulty reconciling some of the things I think you've said here and in the past. I'm certain it's down to communication, and it really isn't worth either of our times trying to clear up the disconnects.

Probably. As I said, though, there are 5 fixed heights (relative to the ground, obviously) in the game vs. the industry standard 1. This makes the game infinitely more robust and flexible, but it's still an approximation of real life.

Which occur several times in most games I play...

Depending on how much you use Area Fire and how much you fight in built up areas, of course the issue will come up more/less frequently. What I meant to say is that given the number of fire opportunities that crop up within the average game for the average player the oblique angle problem is extremely small percentage of them. If you were playing CMSF Beta you'd have a better appreciation for how much worse it could be. It was maybe as high as 10% of the time when you went to do Area Fire.

I know that gamers focus on the exceptions and take the rest for granted, therefore I was just making a friendly reminder

Having said that, I fully understand how frustrating the oblique angle issue is. As I've said, when I designed the game on paper I deliberately planned a way to minimize the potential for the problem. It was not viable to code and therefore other means were undertaken to minimize the undesirable behavior. I'm just not sure when we'll be able to take further steps to reduce the occurrence even further.

Yes, I've got that. What I don't get is how "there can never be LOS between these AS" from the LOS Map gets converted to all the myriad cases (and they certainly aren't limited in number, though they're probably often marginal, since "margins" are where engagements can be won or lost, so players seek them as a matter of routine) where, actually, if there are units there, they can shoot at each other.

I've done the best I can to explain it. I think at this point you should just accept that the game's behavior is internally consistent and is not offering a magical way out of the oblique angle problem. There's no quick or easy fix to the oblique angle problem. Sounds like you have gotten that out of this discussion so good deal.

Obviously it's escaped your attention that I've been asking that very question.

You misunderstood I was making a crack that no system is perfect and CM's certainly isn't. But it is so vastly more complex and detailed than anything out there it's a wonder it can operate with only a few (though certainly frustrating) negative issues.

"How does it work?" so that I can understand it.

And I've been only too happy to help do what I can to bridge the knowledge gap.

Again... the bottom line here is we fully understand and appreciate how frustrating it is when an oblique angle problem crops up in a game. Often times it's not that important, and I'd say never influences the overall game outcome, but it's always frustrating. Frustrating is never a good thing so we have all the motivation in the world to fix this. When we can.

Steve

[Battlefront.com]

I would not quite go as far a most games but the thing is if I play a game that has any kind of built up area then it occurs turn after turn.

For sure built up areas highlight the problem big time since it has the exact conditions that make it a) more likely to happen, less likely to have an acceptable workaround, and c) more likely to have tactical significance.

So, while I understand that there is nothing practical that can be done at the moment, please understand that this issue is not marginal or rare. I hope one day you can make modifications that will make progress.

I hope so too. I never wanted this sort of problem to exist in the first place

Steve

[PhilM]

What I don't get is how "there can never be LOS between these AS" from the LOS Map gets converted to all the myriad cases ... where, actually, if there are units there, they can shoot at each other.

The word "never", in the above example, was part of my problem in understanding this.

Where I've now got to on trying to grasp this (Steve, if what follows is rubbish, by all means disillusion me, but please don't take away my gold star: that's too precious!) is ...

Action squares are (in the context of the game) inanimate objects; they themselves do not / cannot "have" LOS to other AS, but rather offer the opportunity of LOS for units within them to other AS or units.

So the LOS map will "never" allow LOS to certain other AS: it is always completely blocked to a specific other AS, regardless of which unit is where in either AS, because, say, of a large hill in the way.

Similarly, some AS pairs will always be in LOS: say, two adjacent AS, on the same altitude and both level, just bare earth. Even the lowest, prone, unit viewpoint will always give LOS into that next AS.

But in between those extremes, there are many AS where LOS is potentially available for a unit to another AS or to a unit in that other AS, dependent upon which unit, in which "stance", is in one or both AS. The examples that puzzle us come in this third category of AS.

But only some of the factors that affect the outcome are illustrated in the graphics we see: e.g. an unbuttoned TC in a tall vehicle visibly gets a different view from the point of his AS where he is positioned than does a kneeling figure.

But what about those many factors NOT portrayed nor explicitly represented in the calculation: how tall is the TC, or the kneeling figure? Does a squad member lean out of a window to look across a road? Etc, etc.

To cope with all of these grey areas and fuzzy edges, they are abstracted into a LOS calculation that sometimes allows marginal LOS to occur - reciprocally - and sometimes doesn't. But, only in those calculations that fall into my third category of AS relationships: if the LOS map says (truly) "never" or "always", then the fuzzy edges are immaterial.

This picture helps me to come to terms with what is going on, and for me makes it easier to accept the outcomes we get, but don't always like!

[Battlefront.com]

The word "never", in the above example, was part of my problem in understanding this.

Where I've now got to on trying to grasp this (Steve, if what follows is rubbish, by all means disillusion me, but please don't take away my gold star: that's too precious!) is ...

You earned that one, so you keep it. You also get a second one

Action squares are (in the context of the game) inanimate objects; they themselves do not / cannot "have" LOS to other AS, but rather offer the opportunity of LOS for units within them to other AS or units.

So the LOS map will "never" allow LOS to certain other AS: it is always completely blocked to a specific other AS, regardless of which unit is where in either AS, because, say, of a large hill in the way.

Yes. Think of it like this. You are a unit and you are in Action Square 3244. Without any new calculations CM knows which Action Spots you can theoretically see based on conditions and which ones you can NEVER see from that location. By dividing Action Spots into a "maybe" category and a "never" group CM can avoid having to make extremely costly repetitive checks of places which are 100% irrelevant.

Without this approach CM would have to treat every Action Spot, no matter how ridiculous, as a potential place to spot. Which means constantly asking dumb questions over and over and over again like, "can I see Action Spot 1023 that is 4000m away and on the other side of 10 hills? No? OK. Hey, how about now can I see it? No? OK. WAIT! How about now?". How well do you think this would work on a map that has over 300,000 Action Spots? I think if CM were like this I'd still be sitting around waiting for my first turn of CMBN from 2010 to be finished

OK, so what about Action Spots which are in the "maybe" category and therefore might be visible to a unit in Action Square 3244. To those Action Spots, and only those, CM does dynamic checking based on the current height of the Spotter and the center of the valid Action Spots. This is done from the spotter to the Action Spot Center of each "maybe" Action Spots. If that works then various other things come into play here such as cumulative terrain effects, lighting, distance, binoculars, Morale, Experience, etc. If the check survives all of this, then you have the ability to shoot to whatever degree is appropriate.

But in between those extremes, there are many AS where LOS is potentially available for a unit to another AS or to a unit in that other AS, dependent upon which unit, in which "stance", is in one or both AS. The examples that puzzle us come in this third category of AS.

Yes. That's because variables are at work which are more subtle. Nobody questions why a unit can't spot another unit on the other side of a hill. That's pretty obvious. But why can't a tank spot another tank in a particular set of conditions? Sometimes that too is obvious (one's moving the other is in a forest), but toss in one more variable and sometimes it is not so obvious.

But only some of the factors that affect the outcome are illustrated in the graphics we see: e.g. an unbuttoned TC in a tall vehicle visibly gets a different view from the point of his AS where he is positioned than does a kneeling figure.

But what about those many factors NOT portrayed nor explicitly represented in the calculation: how tall is the TC, or the kneeling figure? Does a squad member lean out of a window to look across a road? Etc, etc.

To cope with all of these grey areas and fuzzy edges, they are abstracted into a LOS calculation that sometimes allows marginal LOS to occur - reciprocally - and sometimes doesn't. But, only in those calculations that fall into my third category of AS relationships: if the LOS map says (truly) "never" or "always", then the fuzzy edges are immaterial.

Yes. It is simply impossible to create a visual world that is as diverse and subtle as the real world, so there's some degree of "fuzzy" edges deliberately programmed into the game. It's what gives CM a more overall realistic feel than other games. Including CMx1, which basically was all "fuzzy" edges but without 1/10th the visual feedback.

This picture helps me to come to terms with what is going on, and for me makes it easier to accept the outcomes we get, but don't always like!

The only significant problem we've found is the oblique angle issue. That happens when theoretically Action Spot A can see Action Spot B, but due to the type and position of Friendly unit the system determines the angle to hit that Action Spot is too marginal for Area Fire. However, if an enemy unit in Action Spot B can shoot at the Friendly unit in Action Spot A, because it is in a better position to see into Action Spot B, then CM changes its ruling so that there is reciprocal behavior. Once the Enemy unit moves the situation reverts to how it was before... too marginal for Area Fire. Enemy unit moves back to engage Friendly unit... exception is once again reinstated.

Steve