Biggest shortcoming of CM system

[flamingknives]

When I first started playing CM, I would have readily agreed about the 'sudden' changes of elevation. Then I remembered that each square is 20m to a side. The nearest flat area to me IRL is a deliberately flattened area of flood plain, and is maybe 100m across (5 squares) and, ignoring fences, buildings etc, is about 300m long. the rest is quite sloped, at varying steepnesses. Admittedly, I do live in a range of hills, but then this would compare roughly to 'moderate hills' in the map generator. The main difference between a QB map and RL is the lack of continuous features, such as a river valley (albeit a river too small to represent)

[BlackVoid]

Just played a game that is a good example. The enemy had a PZ4J, behind a patch of trees. I advanced 2 tanks (1 Sherman, 1 M10) on a road, where some of my infantry was fired upon the previous turn.

As my tanks moved forward hunting, the M10 got LOS and started to engage shots. Moving paralelly, the Sherman advanced and then stopped. NO LOS to the PZ4J. I planned a 2 vs. 1 fight and ended up in 1 vs. 1. Too many dancing on the edge of LOS in my opinion.

Moments like this decide games way too many times. Where one guy finds that spot with perfect LOS, but the other does not...

Wont be around till next Friday....

[Redwolf]

Originally posted by BlackVoid:

Just played a game that is a good example. The enemy had a PZ4J, behind a patch of trees. I advanced 2 tanks (1 Sherman, 1 M10) on a road, where some of my infantry was fired upon the previous turn.

As my tanks moved forward hunting, the M10 got LOS and started to engage shots. Moving paralelly, the Sherman advanced and then stopped. NO LOS to the PZ4J. I planned a 2 vs. 1 fight and ended up in 1 vs. 1. Too many dancing on the edge of LOS in my opinion.

Moments like this decide games way too many times. Where one guy finds that spot with perfect LOS, but the other does not...

Wont be around till next Friday....

I see the situation exactly the other way round with regards to modeling problems: in real life the stationary tank would have shot a whole platoon of enemy tanks coming around the same corner of the same house. At least after hitting the first, all others would fall victim to zero'ed in shots and standing versus having to stop and then aim.

In CMBO you often see a stationary tank miss and the delay for the moving-in one is only minimal, in the majority of cases they end up with 1:1 chances pure duel with no advantage for the stationary tank. In addition, if several tanks come around the same corner, there is no zeroing in for the stationary shooter, each time it selects a new target it starts zeroing in from scratch, even if the target is exactly in the same spot where the previous one was at the last shot.

And moving multiple tanks into LOS for one enemy tank is in my opinion a lot easier in CMBO because you can shoot through friendly tanks, you just stack them. I don't know why you Sherman stopped without LOS to the Pz IV, if you use hunt right into very clear LOS places that can't happen.

Overall, I think it's pretty obvious that the modeling shortcomings in CM even out quite a bit, otherwise this thread wouldn't be so long.

[CMplayer]

Originally posted by BlackVoid:

Just played a game that is a good example. The enemy had a PZ4J, behind a patch of trees. I advanced 2 tanks (1 Sherman, 1 M10) on a road, where some of my infantry was fired upon the previous turn.

As my tanks moved forward hunting, the M10 got LOS and started to engage shots. Moving paralelly, the Sherman advanced and then stopped. NO LOS to the PZ4J. I planned a 2 vs. 1 fight and ended up in 1 vs. 1. Too many dancing on the edge of LOS in my opinion.

Moments like this decide games way too many times. Where one guy finds that spot with perfect LOS, but the other does not...

I think what you are describing is a side effect of the development of the WeGo system, and as such is a small price to pay for such an improvement over the old way of taking turns moving.

One could argue that these little 'problems' are actually good things. Why? They counteract somewhat the overreaching, unrealistic level of coordinated control the player has. Control freaks are better off playing chess.

Basically when you play your armor in CM, the movement orders are extremely detailed, more detailed than what would be given to a tank commander. So you are in a sense 'playing' the driver and TC. BUT you suffer from order delay, as though the orders are coming from highers up, and an inability for the tank to know that its mission is, in your case, to engage an enemy tank at a known position. So the AI won't move it that extra milimeter to enable it to engage.

The result is that the tank drives along a minutely planned route, but sometimes can't react properly to contingencies. You're sometimes schizophrenically stuck between a FPS tank and a tank under the purposeless command of the AI. And your chances to tweak orders come at one minute intervals, which can be synched or unsynched with when the crucial junctures in the battle arise.

These problems and uncertainties, through their chaos, make the game more of a wargame, imo, even if they artificially arise out of the logic of the game, rather than being modelled on real life chaotic factors in C&C. Ultimately of course we long for the latter, but such as they are, they work to throw more of that randomness and 'oh ****' factor into the game.

One very nice detail of play, is making sure to have some extra waypoints on units at the end of a turn. These can be dragged and changed to get movements without command delay in the next turn.

But suppose two tanks 'arrive' at a destination, and one has a milimeter of waypoint left, but the other doesn't. If they are in danger, you can drag that remaining waypoint to a safe zone and switch to reverse or fast move. He's outta there. His buddy tank who hasn't done anything different is stuck for 13 second. There is no un-forced real world explanation for this. It's just a result of how the ordergiving system is designed. But you can say to yourself he couldn't get the tranny into first, or something like that...and watch the brewup with detachment. Get some marshmallows.

Anyway, until computers have more processor power, and more time is put into conceptual design issues and programming we'll have to live with this. The game still represents an advance, a giant advance, over previous games afaik, and that's all we can really ask for, right?

[CMplayer]

Redwolf,

Very insightful constructive criticisms which show why CM (or rather the typical CPU) really isn't yet able to model many very real tactical problems, especially wrt to disposing of troops and vehicles to lay down maximum fire without putting friendlies in a crossfire. And trying to manuever the enemy into the reverse, disadvantagous situation.

In addition, both for tanks and infantry, the advantage of being stationary is underrepresented in the game.(as you mention wrt to tanks and guns) But I wonder whether that too might not be a good thing for gameplay. Real life battles could go on for hours over meters. Maybe by not penalizing movement so much in CM you encourage a more mobile, and thereby interesting game.

[BlackVoid]

------------------------------------------------

I see the situation exactly the other way round with regards to modeling problems: in real life the stationary tank would have shot a whole platoon of enemy tanks coming around the same corner of the same house. At least after hitting the first, all others would fall victim to zero'ed in shots and standing versus having to stop and then aim.

-------------------------------------------------

This is true, but does not relate to the problem I am describing. The situation could have been the other way round: my tanks stationary and the enemy moving. My tanks were beside each other not after each other (maybe that would have worked better).

This small difference caused one of my tanks to spot, the other not to spot. As most human players always play on the edge of LOS for armor, this situation happens fairly often AND DECIDES BATTLES. It can be very frustrating....

[Dirtweasle]

Originally posted by OGSF:

</font><blockquote>quote:</font><hr />Originally posted by LeeW:

On the question of troops having to exit buildings to move from one to the other. This seems real to me unless you have time to blow a hole in the wall.

lee

Which is one of the things they did. Used bazookas to blow holes through adjoining walls so they could avoid the street. Doubler describes the practice being used by U.S. troops in Achen, in "Closing with the Enemy".

The other tactic was to throw smoke grenades into the street before makeing the dash for the next building.

Neither of these tactics are modelled in the current CM game engine, but they were used.

OGSF</font>

[Redwolf]

Originally posted by CMplayer:

[QB]Redwolf,

Very insightful constructive criticisms which show why CM (or rather the typical CPU) really isn't yet able to model many very real tactical problems, especially wrt to disposing of troops and vehicles to lay down maximum fire without putting friendlies in a crossfire. And trying to manuever the enemy into the reverse, disadvantagous situation.

While I go on to defend CM, I have to point out that I see no good excuse for a command "move along this path and stop when gaining LOS to THIS spot" missing. I my opinion it should be mandatory for every WEGO game, but not even Tacops has it (which otherwise has a SOP system which is very attractive and a better solution to some of the TacAI problems, although on a slightly different game scale). But TacOps also has an absolute/free LOS tool.

In addition, both for tanks and infantry, the advantage of being stationary is underrepresented in the game.(as you mention wrt to tanks and guns) But I wonder whether that too might not be a good thing for gameplay. Real life battles could go on for hours over meters. Maybe by not penalizing movement so much in CM you encourage a more mobile, and thereby interesting game.

That is a good point.

I think that a compromise might do: extend the effect for zeroing in to an area around a shot that was fired, like -say- 15 meters. If a new target appears within 15 meter of the place where the last shot landed, let the old zeroing in be still in effect. The same applies if the shooter moved, but less than 15 meters. Right now the zeroing in is always resetted instantly and completely at any time there is no continuous targetting in effect. Even retargetting the same targets resets zeroing in.

Note that while I say 15 meters on the ground, a more sophisticated criterium based on degrees in gun elevation or similar might be used.

If in addition you get a command to limit the number of shots in a turn, you can area fire two or three rounds into the ground to zero in at a place you suspect targets to appear. Targets then appearing at that corner are hit with very high probablity. As a nice side effect, this doesn't develop into a instant deathtrap since the opponent will get a chance to spot your firing vehicle and see where it zeros in. A nice tradeoff for the player to think about.

Doing this would also help defenders and make things like Nashorns and 88mm gun more useful, you can zero them in. And since CMBO only models hasty defenses, you have to do it while possibly being observed. To round it up, the defender may be able to buy direct-fire target reference points which put the 15m zeroed-in area around them and also would be useable as ambush markers to shoot at anyone entering the high-precision zone.

[ April 01, 2002, 06:10 PM: Message edited by: redwolf ]

[Redwolf]

[double post]

[ April 01, 2002, 06:11 PM: Message edited by: redwolf ]

[Renaud]

It would be nice if vehicles were LOS obstacles for direct fire weapons. This might discourage the practice of tank mobs operating tread-to-tread. Of course it might result in a severe handicap for the AI which tends to cluster tanks together in tank-blobs. I also wish vehicles could not meld into one another...that their 3d forms were more material rather than incorporeal. Hurts the immersion factor when you see a bunch of vehicles melded together into an unrecognizable blob. Ever notice how long tank guns never bang into walls, other tanks, trees? Really a tank with a long gun (88, 75LL, 76) would not be able to swivel right through an adjacent building or heavy woods. I guess this would drive the AI crazy though.

Ren

[Redwolf]

Originally posted by Renaud:

It would be nice if vehicles were LOS obstacles for direct fire weapons.

Actually I am not so sure anymore. yes, it is unquestinable a big win for the realism of infantry moving behind a tank or riding on it, using the turret for cover.

But for tnk-to-tank combat you get all the problems with LOS and not enough control that make this page in this thread so long, but twice as bad. So you have to move besides the tank to shoot alongside it instead of behind - then what? When you plot a move you don't have enough control over it to ensure the tank ends up with clear LOS, where in reality ever drive would be clever enough to move until sight is clear.

This might discourage the practice of tank mobs operating tread-to-tread.

It would be far better to punish these guys with zeroing in that applies to all of them at once. Bang Bang Bang Bang, platoon gone.

I am not sure why you want to discourage that, it is realistic, and party punished in CMBO right now because an attacker doesn't has to turn hull or turret,

I also wish vehicles could not meld into one another...that their 3d forms were more material rather than incorporeal. Hurts the immersion factor when you see a bunch of vehicles melded together into an unrecognizable blob.

You are sure you use magnification "1"? Yes, there is some oversizing left, but pretty minimal, IMHO.

Ever notice how long tank guns never bang into walls, other tanks, trees? Really a tank with a long gun (88, 75LL, 76) would not be able to swivel right through an adjacent building or heavy woods. I guess this would drive the AI crazy though.

Right, the big cats should have serious problems in narrow strees and narrow passages in woods. Jagdpanther and Panzer IV/70 much more than Panther and Tiger.

However, I think you misunderstand the problem. The real problem is not that CMBO doesn't compute the gun traverse limitations. The real problem is that the terrain patches are 20x20m. Every street in CMBO is guaranteed to be wide enough so that even the longest tank can turn, every passage through woods is guaranteed to be wide enough.

In addition, real-life tank drivers are not stupid. If you are in a 10 meter lane in woods, with a 12 meter tank gun you cannot just turn as such, but you can use a combination of turret traverse and hull movement to move the gun cleanly into the woods and turn that way. No way coding this into the TacAI can be a worthwhile investment.

[Michael Emrys]

Originally posted by flamingknives:

</font><blockquote>quote:</font><hr />

quote:

------------------------------------------------------------------------

There is the slight problem due to the resolution of the ground tiles, but I hardly think it's worthwhile moaning about.

------------------------------------------------------------------------

What problem is that?

As each 20m terrain square is divided up into a number of polygons, changes in elevation can sometimes lead to "butresses" protruding from the side of a hill, where IRL the hill side would be smooth. The little pockets formed in this way are great for positioning a unit in, but a little unrealistic.</font>

[Caesar]

What I meant by bumbiness is the many chaotic elevation changes on a map. Real life terrain tends to be a bit simpler: smoother hillsides, etc. Go to any location in your area and you will see that terrain most of the time is simpler than in CM. Positions that have good LOS coverage are quite readily apparent. In CM it is often tricky to find good positions because of the 'bumpiness'. I would like to have an option in terrain setup that would set this 'bumpiness' level.

A cleaner terrain with more objects (ditches, railroads, etc) would be more realistic than the current one.

BlackVoid, I guess it must come down to where you live and the terrain around there. Just last week while I was stuck in traffic on the way home I started to notice the terrain around me from the perspective of a CM game (Hmmm - is this game beginning to get to me). With the exception of straight along the motorway, there was no clear LOS for anything more than a couple of hundred metres. You could see some more distant patches of land on the tops of rises but there were virtually no clear open spaces and what there was had dips and bumps that would have easily hidden me and perhaps vehicles. It was this that struck me at the time. Even the farmland further on was not actually all that open with the hedges, drains, creeks etc. I agree that there needs to be more ditches and other such terrain objects but really these are largely eye candy.

If you want really good maps, try the scenerios. I;ve just begun playing them recently and they add a whole new dimension to the game.

As far as the bumpiness goes though, does anyone know if CM models men (in open ground) as being under partial cover when in the human equivalent of hull down i.e. partially behind a rise.

[Wilhammer]

"It was an example used to support the argument that too much time in CM is spent depriving the enemy of LOS or trying to achieve LOS yourself."

Ummmm. I fail to see a problem here. Terrain is EVERYTHING in modern warfare. In existance are many accounts about units being engaged by well hidden opponents ("We did not know a German Tank was their sometimes until a bush moved").

In the close range combat seen in CM, the scurrying rats thing is realistic, IMO. The battle we see on TV of footage of expose troops and armor corsiing the open is generally BEFORE engagemnt ranges we see in CM. Baiscaly, except for artillery, at over 1.5 km, weapons were not effective, and thus moving fully exposed to the enemy was not inherently dangerous (excepting the intel given by moving in the open).

But:

1. I despise the idea that I can send my men out in front, and deliver supporting fire THROUGH them to the enemy without risk of my troops getting hit. A new tactic I am using is to send out a platoon to take some woods, supported by area fires through my friends into tree lines. One would think some of your men would fall due to friendly fire. (This might sound gamey, but it is a valid historical tactic.)

2. On the edge "non WYSWYG" LOS problems. I remember a specific example in (scenario name withheld) when a PZ4 took out a fast moving M-18 in one shot, and one could swear that where and when the round made its kill, a crest line blocking LOS existed between them.

[Wreck]

The problem with LOS in CM is in part due to the borg sighting. But in large part it is due to the modelling of all units as points. In particular this is true of the gun/tank problem, where tanks get to kill guns safely via area fire. In real life this did probably happen. But not as frequently as in CM, because of lack of knowledge to fire (no borg sighting in real life), and because the gun would be just as likely to be able to see part of the tank (but not its gun), as the tank would be able to see near the gun (without seeing it). In fact, given the fact that guns are much smaller than tanks, the reverse situation (guns killing tanks that could not see them, at least not down the barrel) probably happened a lot more than the CM-style tank-kills-helpless-gun.

BTS should forbid tanks from area firing on spots which they have more than a modest amount of LOS obscuration to. That would be easy. A bit harder would be at add one exception: allow them to area fire at any spot they can currently, *if* they are in range of a friendly infantry HQ that sees the spot (i.e. they act a bit like mortars). And they should be in-range of infantry only if unbuttoned.

The best fix would be to give all units an area, especially tanks. Of course this would require a lot of work (and extra CPU horsepower) in/on LOS, which is why BTS doesn't do it.

[Louie the Toad]

Dear BlackVoid,

Regarding one tank being able to draw a line of sight to something and the tank next to it not being able to do that.....

Isn't that just like real life ? How often have you said to the person next to you " Look over there, do you see that?" and lo and behold

they can't see it!. There are many real life reasons for this happening.

20/20 ..... Toad

[arax3]

BV: The key point here as elsewhere in the gaming industry is the word: "GAME." You could always sign up with the US armor troops if you want the real thing! Right? Now, I don't know about how the AI would run if CMBO used a Cray Computer but that is not practical. L

[CMplayer]

Originally posted by arax3:

Now, I don't know about how the AI would run if CMBO used a Cray Computer but that is not practical.

15 years ago the computing power sitting on my desk would have taken a mega mainframe to achieve. In 15 years I'm sure we'll have the equivalent of the custom deep blue machine in our wristwatches.

Watch out..._then_ CM is gonna really be something.

[CMplayer]

Originally posted by Caesar:

There is no cover bonus for being behind the crest of a rise for infantry. Tanks can get hull down status as you mention.

[CMplayer]

Originally posted by Louie the Toad:

How often have you said to the person next to you " Look over there, do you see that?" and lo and behold

they can't see it!. There are many real life reasons for this happening.

The point is that if it's important, the person next to you can move to where you're at to look, and doesn't need to wait a minute for orders, then 13 seconds for a command delay.

In RL if those tanks were tasked to move to engage, then they would find good spots. Since you can't check LOS from spots you aren't at, you can't always plot the proper move to get it. Perhaps a few good simple SOPs which influence the 'hunt' command would solve it. Such as being able to do a hunt which _excludes_ engaging specified units or unit types.

(as it is in 90% of cases there is some clever way to use the existing commands to get a good result. For instance a combination of fast move, move and then microshort hunts forward will usually get you to a nice hull down with LOS without you stopping on the way to take potshots at some out of ammo mortar on a hill 800 meters away.)

So the issue has to do with the WeGo system and the tacAI, and how they function together in the game, not simply that keyholed LOS occurs.

[ April 02, 2002, 01:28 PM: Message edited by: CMplayer ]

[Michael Emrys]

Originally posted by Caesar:

As far as the bumpiness goes though, does anyone know if CM models men (in open ground) as being under partial cover when in the human equivalent of hull down i.e. partially behind a rise.

So far as I know, the bumpiness does not provide partial cover, though a test could be run to determine whether that is true or not. What it does do is in some cases absolutely block LOS. I often make use of that fact when moving troops through long stretches of open terrain where they might come under fire. I try to fast move them in bounds of 50 meters or less and drop into hiding behind a bump, there to rest until I am ready to move them again. Seems to work okay for me.

Michael