Embarked/Destroyed Crews: Objectives

[womble]

In CM2 I regularly find that levels of victory can change radically eg: from Major Defeat to Total Victory on the basis of a couple of guys KIA. It's frustrating and unsatisfying.

Sounds like you're not interpreting VCs very well, to me. Since the victory level depends on the ratio of the higher score to the lower, there is literally no mathematical way, apart from a mahoosive "friendly casualties" VC (which aren't exactly common) that a couple of casualties on their own can change a result from defeat to total victory. If those casualties were the ones that pushed the AI to Surrender (which, since all their troops count as destroyed and all the VLs count as captured by the player, could turn even the most unfavourable apparent tactical situation into a victory) then you might have a point. There's something odd going on, though, if you're failing to take any VLs or achieve any other VCs while reducing the AI's morale to Surrender levels.

[Michael Emrys]

Erwin might have meant a situation where a couple of high value VLs are being held by a couple of guys who if then are KIA or even seriously WIA, could swing the outcome of the battle entirely. My answer to that is if such an outcome occurs, the VLs were being tenuously held in the first place. In short, it is not obvious to me that this argument provides a compelling reason by itself to change anything in the code.

Michael

[postfux123]

If I understand the new system and remember the old right I think it is easier now to wrestle a VL from your opponent and more difficult to claim it for yourself.

In CM1 you would sometimes just bunch as many troops as close as you could and hope it would be yours. Didnt work for me every time.

Right now you can only own it if you take the VL, clear it and make sure not a single scout sneaks back in. Works for me.

I think in the end it is a matter of scenario design. Given the new system is more flexible i see it as an improvement.

[Erwin]

"Right now you can only own it if you take the VL, clear it and make sure not a single scout sneaks back in. Works for me."

I don't get that argument. It actually encourages gamey play where you can make a major change to the victory level by "sneaking in a single scout" that in RL would make no difference.

[postfux123]

I don't get that argument. It actually encourages gamey play where you can make a major change to the victory level by "sneaking in a single scout" that in RL would make no difference.

1. I think in RL a single scout can make a huge difference.

2. If he can sneak in a scout, he could sneak in a squad = area definetly not secure.

3. The scenario designer can define a small area as VL making our discussion obsolete. If he decides a large area is your objektive, secure it.

[Erwin]

You still have to convince me that a single enemy soldier hiding in a cellar vs a company or battalion or even larger occupying the town would nullify the concept of "control/secure the town" in RL.

[womble]

You still have to convince me that a single enemy soldier hiding in a cellar vs a company or battalion or even larger occupying the town would nullify the concept of "control/secure the town" in RL.

If he's got a radio, the town's not secure. If he's got a sniper rifle, the town's not secure.

This has been done to death over and over: it's a map design issue. "Occupy" is what it is. If a designer just paints one single massive VL, then it starts to stretch the boundaries of credulity. Simple solutiuon: several small VLs. Then one "straggler" can't determine control of the whole stick of terrain VPs.

[Erwin]

"Simple solutiuon: several small VLs. Then one "straggler" can't determine control of the whole stick of terrain VPs."

In game terms, that certainly helps.

But by your logic, you could have a whole division occupy the town and one guy cowering in a cellar makes the "town" insecure. That just doesn't make sense. By that logic, none of us are "secure"... ever... Since in all of our communities TODAY there is probably a nutcase cowering in a cellar with his rifle.

[womble]

The problem is that the game cannot determine whether a pTruppe is in a cellar or at the top of a bell tower. I.e. it cannot determine the threat posed by the un-neutralised enemy.

And pretty much any analysis of VP conditions pushed to the extreme makes a nonsense of them: if there's a whole division in that one town, why isn't it being flattened by every piece of arty within range?

You're given orders to clear an area. Clearance is binary. You know what you have to do. Live with it. If the designer hasn't recognised that painting half the map as one VL is a bonehead thing to do, that's a problem with the designer, not the nature of "Occupy".

[postfux123]

You still have to convince me that a single enemy soldier hiding in a cellar vs a company or battalion or even larger occupying the town would nullify the concept of "control/secure the town" in RL.

If he has a means to communicate Mr. 15cm and his friends might make a difference espacially if there is a battalion bunched in the VL. A single rifle opening up on your superior officer in an area declared secure by you might make a difference for your personal career. Historically a stupid horse sneaked into a VL bristling with enemy troops made a difference to turn a major defeat into a major victory. Trojans didnt get a chance to play gamey Greeks again.

Seriously, a VL that takes a company or even a battalion to secure seems to be a design flaw or at least a scenario not very funny to play. You can cover a lot of ground with a single squad. Most scenarios use multiple VL spread around the map and even some touch objectives. If there is a big VL you should put some effort into securing it.

[Erwin]

Ok, I understand the concept...

So, really what we're debating is the identity of the single guy hiding in town and the fact that the AI can't tell a real threat from a non-threat.

If he's a WIA desperate guy with a pistol or rifle he's not realistically "contesting" that town. However, if he's a sniper or an arty spotter he is dangerous. (Although, while a sniper is an irritation, I still wouldn't say he is "contesting" that town that may be occupied by a large number of one's troops.)

However, do arty barrages mean control is really contested? Just cos a town is being shelled doesn't mean it's not controlled by the side occupying that town. The FO doesn't even have to be in the town after all.

I am talking about a RL situation not so much the way the game works - which in this case is a bit silly due to AI limitations. And that gives rise to unrealistic and often frustrating victory results.

[Michael Emrys]

...that gives rise to unrealistic and often frustrating victory results.

I think womble has put his finger on it: the problem lies not so much with the mechanics but with scenario/map design. With a large number of VLs scattered around the map, the control or not of any one of them is not going to tip the final outcome very much. Toss in the points assigned to things besides objectives and the effect becomes very small indeed.

Michael

[Erwin]

I agree that the workaround works (sorta). But, it is a workaround. I still prefer the way CM1 calculated who held a particular VL.

[A Canadian Cat]

But it is not a workaround. If the scenario designer has a whole village as the objective then you know you have a lot of work to do - you must clear the whole village completely and probably more so no one can sneak back into the village. If the scenario designer sets up three or four key buildings in the village as objectives then you know you have less work to do and you do not need to secure the whole area just some key points. You can then ignore stragglers in the village cause the designer told you they will not count.

The CM2 way makes things clear - you own it or you don't. This give designers flexibility in deciding how much control you need to have over an area.

The CM1 way was fuzzy - you never new what you controlled or what points you would get. This offered less flexibility to scenario designers.

That's how I feel anyway.

[Pelican Pal]

The current occupy objective is woefully designed to handle how it is actually used and constantly results in gamey battles.

If you need to occupy a whole village then it is possible that a lone wounded crewmen in some shack on the corner of town will contest the town from your good order company who has secured it. That is blatantly absurd.

If the occupy objective is only on very specific "key" points then the battle becomes some absurd king of the hill where two opposing companies might be trying to push a squad into a single building in a village. Holding that single building does not mean that you own the village.

The occupy objective as it stands is really a sweep and clear objective. Having a proper occupy objective would be a lot better than the current system.

[postfux123]

If the occupy objective is only on very specific "key" points then the battle becomes some absurd king of the hill where two opposing companies might be trying to push a squad into a single building in a village. Holding that single building does not mean that you own the village.

Exactly that could happen with the old system. There was an incentive to push troops close to a victory flag with the single reason to be close to it.

The new system gives you points when you are there and he is not. If you achieve that it usually means you have controll over an area beyond the VL.

I agree that a large VL can be annoying but that is a design decision.

[womble]

If the occupy objective is only on very specific "key" points then the battle becomes some absurd king of the hill where two opposing companies might be trying to push a squad into a single building in a village. Holding that single building does not mean that you own the village.

The point of this approach isn't to have just one outhouse in the middle of the village as the only vital location, but to have several (3? 5? more?) VLs on several buildings/areas, so that you're correct, pushing the other guy out of that single building doesn't mean you have control of the village, just, possibly, now, more of the relevant parts of the village than they do.

It remains a matter of designer's usage of the tools they've been given. Like having an occupy objective on a bridge is militarily nonsense: that bridge isn't held or safe in any meaningful way when the enemy's armour can sit hull down 400m away and snipe anything crossing it; there need to be occupy objectives on the salient overlooking terrain points.

The "occupy" objective was not "designed to handle how it is actually used". It was designed, and it gets used with increasing finesse as designers figure out how its actuality affects gameplay. See MarkEzra's (relatively) recent comments on having more VLs in the QB maps for RT. Perhaps the effects should have been more obvious from the outset. Perhaps they could have been more rapidly assimilated into the general corpus of knowledge of how to make splendid scenarios, but they do all right most of the time in spite of the potential for generating absurd situations. There are all sorts of ways of doing that well before you get to the point of assessing victory...

[Hellas]

I think the problem is, that the game engine has no capability to judge the tactical situation. All the problems result from this.

For example: claiming to control a crossroad, while it is in range of heavy enemy weapons in reality doesn't work. You can occupy the road, but you don't have tactical control over it as long as the enemy can shoot at it (with sufficiently strong firepower).

Currently the game cannot determine how much a spot is threatened. As long as that is the case, plausible results must be limited.

At the end of a game it should not be the most difficult thing to perform a LOF check from every unit on the map to every single action square within a VL: calculate the potential firepower from each unit to each spot within the VL. Every spot within the VL would receive a firepower-value from each unit.

Summing up the firepower of all the squares within the VL (for each side separately) should give a nice simple metric indicating how much firepower (at the given moment) each side has on a VL. Therefore it should allow a tactical judgement how well each side controls the VL!

Example1:

Blue has a one legged lousy crew member with a pistol within the VL.

Red has a full squad within it and three tanks are placed around it with LOF into most part of the VL.

The calculation with this system would deliver probably something like a 99% control of VL for red, since the firepower of a single pistol and covering only a small part of the VL, would be weighted against the firepower of a squad plus the potential firepower of the tanks torwards most parts of the VL.

Tactically sound result.

Example2:

Blue has cleared a village, which shall be a big VL, and holds it with a platoon, well spread over the village.

But Red still has three tanks with high explosive ammo somewhere outside the village and these tanks have LOF on a few houses, only small parts of the village. So blue obviously was not able to remove this threat completely.

Tactically it would be wrong to give 100% of the points to blue sitting on the VL, since red still can threaten heavily a few houses with his tanks and is dominating these spots in the VL.

My suggested system would calculate strong fire values for the few action spots the red tanks can shoot at, while the blue platoon, being well spread over the whole VL is able to shoot at almost all squares within the VL and therefore controls a much bigger part of the VL - and so the weaker infantry firepower was shomehow compensated with a much wider controlled area.

The numerical result would reflect that blue was not able to secure the whole VL, but Red would also collect a few percentage points from this VL.

Tactically sound result.

Example 3:

A VL in a wide open area without cover.

Blue has a platoon infantry in it.

Red has three tanks around the VL.

The tactical situation of the infantry: awful. It's only chance was teh removal of the enemy tanks, otherwise it is doomed.

This system would calculate:

Blue: infantry firepower on all squares within the VL.

Red: firepower of three tanks on each square within the VL.

Although red was not sitting like a duck on the VL, the system would correctly calculate that red clearly is dominating the VL without even sitting on it like duck.

Tactically sound result.

[A Canadian Cat]

I think the problem is, that the game engine has no capability to judge the tactical situation. All the problems result from this.

For example: claiming to control a crossroad, while it is in range of heavy enemy weapons in reality doesn't work. You can occupy the road, but you don't have tactical control over it as long as the enemy can shoot at it (with sufficiently strong firepower).

Currently the game cannot determine how much a spot is threatened. As long as that is the case, plausible results must be limited.

Sure cool thoughts. Frankly I am not sure if it is worth doing though. What you are talking about is a fair amount of work and the results would be potentially confusing: I can hear the questions now "why don't I get victory points for the crossroads"? So, to fix that you would have to make it easy to see what enemy had LOF to the objective etc. etc.

A much more efficient way would be for scenario designers to create the crossroads objective such that is included the actually crossroads and the two main ambush points (as an example - modified to fit any real scenario). That way the player has to control the crossroads itself and the places that cover it.

There, now your scenario realistically depicts control of the crossroads and BFC have several developer months of programming effort they can use doing something else rather than cranking up the complexity of the victory points calculations.

[Erwin]

A work around is for the designer to create many small VL's at a location so the occupation of one or two VL's by a wounded crewmember with a pistol won't affect the overall result too much.

[Pelican Pal]

The point of this approach isn't to have just one outhouse in the middle of the village as the only vital location, but to have several (3? 5? more?) VLs on several buildings/areas, so that you're correct, pushing the other guy out of that single building doesn't mean you have control of the village, just, possibly, now, more of the relevant parts of the village than they do.

In which case you have to leave men in positions that may be tactically unimportant in order to get all the VPs. The more discrete VPs you have the less flexibility you give the player and the more men you attrite away for administrative tasks.

[Hellas]

Sure cool thoughts. Frankly I am not sure if it is worth doing though. What you are talking about is a fair amount of work and the results would be potentially confusing: I can hear the questions now "why don't I get victory points for the crossroads"? So, to fix that you would have to make it easy to see what enemy had LOF to the objective etc. etc.

The game itself is capable to check where a unit can shoot, correct?

Where is the problem for the programmer to write a code that begins with the first unit on the map and the first action square in the first VL and make a LOF check?

If the check is positive it calculates the firepower from the unit that could reach this spot.

The value is stored for the square and the next square in the VL is checked.

Once the loop for the first VL is done, the next VL is is checked. Until all VLs are done.

Then comes the next unit on the map.

And after the last unit of blue, the same is done for all units of red.

Where do you see a problem? It's a very simple programming task.

A much more efficient way would be for scenario designers to create the crossroads objective such that is included the actually crossroads and the two main ambush points (as an example - modified to fit any real scenario). That way the player has to control the crossroads itself and the places that cover it.

This still does not give the game a capability for make a numerical judgement of the tactical situation.

I think my system in principle is not only very simple, but it could offer capabilities the AI currently can only dream about: if it was not only restricted to VLs, but would be used for the whole map, and running on a separate core, the AI would have tactical knowledge available that could help to find better paths (which could be helpful when units are running away in panic or where paths must be decided).

There, now your scenario realistically depicts control of the crossroads and BFC have several developer months of programming effort they can use doing something else rather than cranking up the complexity of the victory points calculations.

How important is it for a realistic tactical simulation to be able to make sound tactical judgements who controls a VL? I think this has the potnetial of a game changer to the next level.

Since the calculation would take place not during the action but after the battle has ended, the code could be simply appended.

How much time would it take for their experienced programmers to program three loops: 1) for all remaining units on the map, 2) all VLs and 3) all squares within the VLs and attach each square four or six doubles carrying the firepower values (small arms, HE, AT)?

How much time would it take to define an algo that combines these values intelligently into one and write another loop to add up all these values per VL and then calculate the control-ratio for the VL?

Maybe I have overlooked something with the concept of firepower as a tactical judgement-tool.

But programming wise: how more simple can things be, while they are offering so dramatic benefits to realism?

I have read that maybe in the future they would reduce the size of the action squares. This concept works independently from the size of the action squares. Future proove.

Recently I had done my first map, hilly Italy. And one of the tactical key locations is in LOS of several hilltops. There is no possibility for the game to take this into account. Using additional VLs can not reflect the creativity of players. And good players often are very creative to bring the firepower where they want it. And how would another VL with points occupied by a blind and one legged crew reflect the tactical control this spot offers with a gun or mortar torwards the other VL?

It doesn't work and no designer in the world can overcome this problem. One cannot measure length with a scales.

And even if the designer can magically read the intentions of the players (some are very good at that!), he can never know, if player X prefers to spread his long range weapons over many hilltops or if he thinks it's better to concentrate them on one location.

I think designers are already forced to use many VLs to overcome the deficit, that the game has no ability to measure tactical control of a VL. It would be better, if the game would have a mechanism, that allow it to intelligently calculate who controls a VL tactically.

Additionally the problem that the game needs to be able to judge tactical control over a distance would remain unsolved.

What, if it seems suicidal or tactically bad for the player to sit physically at a VL? Maybe he prefers to control it from the distance while keeping his ability to advance somehwere else?

If you have three guns controlling the whole battlefield from the heights, why should you be forced to place units on a VL if any enemy would be dead there without taking out the guns first? The game should intelligently be able to determine who controls VLs.

Being forced to occupy a VL physically (right now a blind and one legged crew is enough) instead of tactically controlling it (by firepower), to me already seems to have been a workaround from the beginning, because there was no algorithm how the game could tactically evaluate the control of an objective.

This method to calculate the control of a VL ofcourse could be offered as an additional option - nothing would change if the scenario designer wants to use the established formula how the control of a VL is determined. But if the designer wants, he could select this algorithm to calculate the ratio, which would determine how the points from this VL are split between the two sides.

[A Canadian Cat]

I think your idea is pretty cool and it could be used as a basis for various interesting things as you suggest. I am just not convinced that a) it would be easy or necessary.

Where do you see a problem? It's a very simple programming task.

How much time would it take for their experienced programmers to program three loops: 1) for all remaining units on the map, 2) all VLs and 3) all squares within the VLs and attach each square four or six doubles carrying the firepower values (small arms, HE, AT)?

How much time would it take to define an algo that combines these values intelligently into one and write another loop to add up all these values per VL and then calculate the control-ratio for the VL?

LOL you must be a manager. I can hear you now "that should take you a day how hard can it be". That is my way of teasing you to say that it is much more complex than you think - probably harder than I think.

[Hellas]

I think your idea is pretty cool and it could be used as a basis for various interesting things as you suggest.

Now the question is, how do the developers become informed about this brilliant idea? :cool: I assume they do not read every thread?

[Hellas]

Anyone knows?