Why infantry combat in CMx2 is so different

[skelley]

I have been having a lot of issues with the assault command causing troops to end up in a tight circle at the last waypoint. The other commands keep them fairly spread out (not enough but better than assault). But when assaulting, being bunched together is a huge hazard. Have there been any advances in this area for CM:BN?

[futon river crossing]

Let him play his boardgames LOL.

[Michael Emrys]

If I may butt in with an opinion here, I think it is a great and fundamental mistake to compare CM to boardgames in the first place. CM is possibly the greatest ever tabletop miniatures game created for the computer. It's most natural audience is among people who love to play miniatures. Its secondary audience would be found among boardgamers who prefer to play tactical-level games, ASL as a classic example. But they in turn are trying to some extent to replicate the miniatures experience with cardboard.

Operational and strategic level boardgames are a whole different kind of animal.

Michael

[Gpig]

Totally agree, Mr. Emrys.

It's starting to feel more and more like a cool little diorama with my scale models ripping all over the place. Sometimes I find myself checking out the interiors and wishing I could watch the crew load and fire the main gun.

[Battlefront.com]

The higher up one goes in terms of organization, the less an engineered solution is necessary or even advisable. The reason is that as individual soldiers and weapons cease to have importance as such, the more abstraction is possible. Design for effect at some point becomes the primary desirable methodology and engineered the minority. And of course, as I have been saying, the opposite is true. The lower down you go the more the little stuff matters.

Regarding the Assault Command's behavior... we haven't made any changes yet, but I still hope some things can be tweaked with it. When it works it works great, but agreed it sometimes produces sub-optimal situations.

Steve

[DaveDash]

I think we can get bogged down in details until the cows come home, but the fact is for me and many other dedicated CMx2 players, many of us are very happy with the outcomes the game produces now. This is in no small part due to guys like c3k's constant bug finding, the BETA testers dedication, and BFC's dedication to the engine.

The game now does not ruin the suspension of disbelief, and acts in a predictable and believable manner. The immersion factor blows away anything CMx1 can provide, and the ability to use and apply real life tactics, as demonstrated by a few of our military posters, is higher than CMx1 due to the nature of better fidelity and the new C2 and spotting systems.

The big "gotcha" moment for me is that CMx2 taught me to use significantly better tactics that CMx1 because the higher fidelity makes it far less forgiving to error.

Yes modern weapon lethality plays a part in this, but it doesn't tell the whole story. The game -can- be far more lethal if YOU the PLAYER, do stupid things, and that has taught me overall to be a much better tactician, especially when it comes to terrain (which is much better modelled in CMx2), squad level tactics, and Command and Control. I went and read up on real life tactics, applied them to the game, and those real life tactics worked. So somehow it can't be that bad of a simulation.

Looking forward to the improvments Normandy brings.

Steve: One note about the assault command - I'd love to see the fireteam providing securty and/or support NOT be suppressed when the assault team is. That's a bit of a big bug for me at the moment, and I believe you guys know about it?

[sand digger]

Some amusing comments on how great the infantry is/was in CMx1, it was rubbish. I forget how many times that infantry in attack fell to the ground and wriggled around until they were eventually killed. That was it, if the going was too tough thats what they always did, no exceptions.

But the thing that sticks in my mind was the use/misuse by the British of the Bren LMG, it could not be fired unless the user was prone on the ground! No 'marching fire' with WW2's ultimate assault weapon no matter how many times it had been used for that purpose in real warfare!!!!!

So I am not a CMx1 fanboi but it was relatively good perhaps because it had bugger all competition. So the infantry aspect is interesting, I see in the featured AAR that there is some US infantry surrendering when perhaps fight and flight may be more appropriate? Its obvious to say that in similar situations infantry may react in different sometimes inexplicable ways, hope that this is what CMx2 incorporates.

[Michael Emrys]

Some amusing comments on how great the infantry is/was in CMx1, it was rubbish. I forget how many times that infantry in attack fell to the ground and wriggled around until they were eventually killed. That was it, if the going was too tough thats what they always did, no exceptions.

There was also the famous Dance of Death. I was playing the US defending in CMBO one night. There was a double clump of trees in front of my line that I was sure the Germans would use as the jumping off point for their attack. Sure enough, that's where a whole platoon of infantry headed. So I targeted one clump with 105s and the other with 81 mm. Being trees, I got a lot of nice deadly tree-bursts. Well, the Germans proceeded to run back and forth between the two clumps until they were utterly annihilated.

This kind of thing, where a unit under fire would perform in an utterly irrational manner (not even in the irrational manner that is often encountered among soldiers in the abnormal stress of combat), was quite common in CMx1.

Michael

[Calibration]

This kind of thing, where a unit under fire would perform in an utterly irrational manner (not even in the irrational manner that is often encountered among soldiers in the abnormal stress of combat), was quite common in CMx1.

Michael

To be fair, the application of good (i.e. sensible and realistic) tactics tended to mean these situations could often be avoided in the first place, frustrating though they were at times. How do you accurately model rational/irrational behaviour in times of high stress anyway? Some of the things I've read people do in RL would seem unacceptable in a game.

[Barkhorn1x]

There was also the famous Dance of Death...Well, the Germans proceeded to run back and forth between the two clumps until they were utterly annihilated.

I have seen this behavior myself many times. A simple case of:

- Under fire at point X, run to point Y

- But point Y is also under fire, move back to point X

- But point X...etc.,etc.

- End result = dead infantry

...stemming from the limitations of CMx1 infantry treatment.

My experience w/ CMx2 is rather limited (I am a WWII guy) but I noticed a much more realistic, IMO, treatment of foot and am really looking forward to CM:BN and a historical and verdant atmosphere.

[Barkhorn1x]

To be fair, the application of good (i.e. sensible and realistic) tactics tended to mean these situations could often be avoided in the first place...

True for the player(s), but us SP only guys were forced to watch as the AI caught up in this situation fairly often.

[Michael Emrys]

True for the player(s), but us SP only guys were forced to watch as the AI [got] caught up in this situation fairly often.

Exactly. I don't recall my troops ever doing the Dance of Death, although they did at times perform almost equally stupid acts, but the AI would do this one from time to time.

Michael

[Calibration]

Fair enough again. One's perceptions of the strengths and weaknesses of the game rather depend on who you are playing against. I stopped playing versus the AI years ago so I've swapped dodgy AI behaviour for gamey human underhandedness

[Battlefront.com]

The Dance of Death, as it became known, was a pretty serious problem we ran into because of the lack of fidelity. It's a perfect example, actually, of where JasonC's presumption that it's easy to get a "counter" can work well in all cases at this level of warfare.

The inherent problem we had in CMx1 is the Squad was a single unit with no simulated soldiers except for their weapons. As the unit got hit more and more the ability to Fight vs. Cower started to swing more towards Cower. This reduced the outgoing opportunities to fire, which in turn reduced the chances of getting the enemy off its back. If this situation remained basically unchanged the Squad getting hit would cease being able to Fight at all, and instead would Cower.

A corollary effect happened for the enemy units doing the firing. To the extent nobody was shooting at them was the extent they had plenty of opportunities to fire at the Cowering unit. Their Morale, which might have initially been whacked, would recover and make them more robust. Which increased the chances of outgoing fire from them, which made the punishment cycle on the friendly unit even worse.

At some point a beat on Squad would pass over a fuzzy logic line where it went from Cower to Flight. Sometimes they would go straight from Fight to Flight! Regardless, once the unit had decided to give up trying to Fight and instead relocate it lost its ability to even try returning fire. Now it was fully engaged in moving. Which, due to enemy fire and terrain conditions, could result in a situation where no movement options were very good and it would go back and forth between more than one. Logic added to the game over time reduced this, but it was impossible to eliminate it because it would screw up other situations where a change in destination WOULD result in a better outcome.

The best way to avoid the Dance of Death (as noted above) was to not get caught in a situation where the enemy had the ability to shoot without being shot at. This is where good Combined Arms tactics could keep the extremes from happening. But luck is not something that can always be counted on, so even the best players ran into situations where the Dance of Death was a frustrating reality.

The inherent reason for the Dance of Death is that once the unit went to Cower mode it had no chance at all to knock the enemy off its back. In CMx2 the chance exists because individuals can continue to engage the enemy even if some portion of the unit is Cowering or (to some extent) moving. Therefore, the chance exists that you can take the heart out of the enemy's ability to fight, for example killing a Machinegunner with a single aimed rifle shot from the one soldier still in Fight mode.

Having said that, in CMx2 you still find situations where you've put one of your units in an impossible situation. You don't get the Dance of Death so much as you just get the Death part. And for the most part when that happens it seems to be a fairly realistic end result.

Steve

[Vark]

I liked the crawl of death, where the AI would often send your men closer to the unit killing them because cover was closer to them. This meant the last 20m or so of an assault often resulted in squads attempting to mime hapless synchronised swimmers. Talking of terrain, I take it that in CM2 using folds of the ground offer cover from fire, not just LOS obstruction. Putting infantry units on the crest of a hill, in CM1, was a death sentence. Instead of offering a reduced target profile, all it meant was that more units could shoot at you, found that out the hard way in the Graffenhoer? practice mission! Luckily my butchered platoon 'swam' the right way because there was a clump of cover at the base of the rise.

Interesting that Steve said infantry in CM had more time spent on them than armour, because often they felt like an add on. Tanks would sometimes engage in soome quite realistic AI behaviour, whereas infantry seemed to follow some bizzare code of practice which was hard to reconcile with reality, as though you were watching a combat mime workshop! Or the, we have grenades but we cannot use an area fire command, even though the weapon used was specifically designed to be used speculatively. Loss rates in urban fighting become unsustainable, without heavy tank support, talking of which can vehicles enter building/factory hexes in CM2?

[Barkhorn1x]

You don't get the Dance of Death so much as you just get the Death part. Steve

Leave it to BF to take the romance out of everything.

[Battlefront.com]

Talking of terrain, I take it that in CM2 using folds of the ground offer cover from fire, not just LOS obstruction.

Yes, though we still have some fudge factor in there because even the increased fidelity of CMx2 isn't as fine as real life. The idea is to let the directly simulated stuff do the best job it can, then abstract from there if necessary. That's a common philosophy for both CMx1 and CMx2. It's just that the point of abstracted started much earlier for CMx1 than CMx2 in most everything.

Interesting that Steve said infantry in CM had more time spent on them than armour, because often they felt like an add on. Tanks would sometimes engage in soome quite realistic AI behaviour, whereas infantry seemed to follow some bizzare code of practice which was hard to reconcile with reality, as though you were watching a combat mime workshop!

That is a symptom of two problems:

1. Infantry is a MUCH harder thing to "get right" compared to vehicles. Which means, all else being equal, infantry will always have more issues than vehicles. Again, all else being equal.

2. The design for effect portions of the CMx1 modeling, which were largely dictated by technological limitations (i.e. couldn't simulate 1:1), required proportionally more work to enhance, tweak, and fix than the more engineered vehicle side of things. And since infantry behavior is inherently more complex than vehicle behavior, the amount of effort needed went up even further.

Again, this is the primary problem we found with using design for effect models to represent extremely complex interactions. At some point it becomes more work, with less success, than an engineered solution. CMx2's approach, on the other hand, was to spend a HUGE amount of time on the engineering of infantry modeling (and the complimentary pieces like better resolved terrain) up front. This means that as we move forward, in and out of various fronts and time frames, the basics hold true and therefore don't require any significant reworking. Behavior improvements are also easier to add because the whole structure is easier to work with. Which means at this point we've probably put in much more time into infantry modeling than armor even for CMx2, but here on out they should be rather similar in terms of needs.

Steve

[sand digger]

What makes me optimistic is that the infantry problems have been recognised and the effort to fix them has been, apparently, put in.

On this infantry tank support all arms thing, I just hope that the success or failure of infantry is not irrevocably tied to it having tank support. It goes without saying that all things being equal infantry benefits from having tank support but in the real world that support was not always available for many reasons. Yet infantry was still able to get the job done and advance under fire, like they did in WW1.

[Michael Emrys]

I liked the crawl of death, where the AI would often send your men closer to the unit killing them because cover was closer to them.

Yeah, that was another one I had in mind but failed to mention. Sometimes troops would start crawling in utterly the wrong direction for their own survival. During the orders phase I would cancel that movement and give them orders that had a chance of getting them to safety, but then when the turn started, the AI would countermand my orders and have them crawling out to where they could be slaughtered...which they were. Very frustrating that.

I'm sure that the AI was just following the logic that had been programmed into it, which in the majority of cases worked fine. But it just wasn't sophisticated enough to recognize the rare case when that logic led to disaster.

It's easy enough to say, and true too, that in real life there were dumb officers and NCOs who led their men into bloody disasters as well as troopers who did so without any prompting from their leaders. That much I am fine with. It's just that the peculiar quirks we are discussing at the moment were typical not of human misjudgment on the battlefield, but failures of machine logic. Hopefully, we can now wave goodbye to all that and in future when our troops screw up our beautiful plans, they will do so in typically human fashion.

Michael

[Vark]

Another intensely frustrating experience, in CM1, was the deceptive nature of the terrain. You'd pan around to see a good approach route and, using terrain masking, plot your lead platoons advance towards your first objectives, hit the play button and watch. Except you would not see a careful advance but a massacre as the "yellow lines of death" would suddenly curve over the shielding terrain and cause havoc. What are the LOS calculations, in relationship to terrain, like in CMBN, can you accurately predict what is a safe route, or do the enemy still appear to have earth penetrating munitions and x-ray optics?

[Rokossovski]

I'm concerned about whether the infantry in CMBN will be helpless if they run out of specific anti-tank assets. Do infantry in CMBN have any way to fight armor absent dedicated anti-tank weapons?

In CMBO/BB/AK there was the ability to mob armor if it came too close to infantry, although of course that was abstracted. (It would show the infantry tossing grenades, and the explanation we were given was that it represented a close assault by one or more members of the squad.) It had the (to my mind realistic) virtue of forcing armor to treat hostile infantry with at least a little respect, especially in dense terrain. Is there anything similar in CMBN?

If not, how will things be handled with, for example, Soviet infantry or early and mid-war incarnations of the game where bazookas/panzerfausts/piats are not a factor? Will armor simply be impervious to enemy infantry?

[YankeeDog]

If they're close enough, infantry will already use grenades against armor in CMSF if they don't have any "AT Specials." It's not very effective, but it can occasionally cause some damage. While assaulting armor (and especially a full-out MBT) with infantry is generally not a good idea, armor is certainly not "impervious" to the effects of brass balls and grenades in CMSF.

Plain old fragmentation grenades were never very effective against armor -- not now, and not in WWII, either. But AIUI WWII-era armor was somewhat more vulnerable to grenades than modern armor is, so infantry with nothing better than grenades should stand a better chance of KOing armor in CMBN than they do now in CMSF. At the least, WWII-era armor lacks infantry spotting gizmos like IR optics, so even without any other changes to the game engine, infantry should survive for a little longer when close assaulting armor in CMBN than they typically do in CMSF.

Cheers,

YD

[LongLeftFlank]

The ol' Molotov on the engine compartment, or the SPR sticky bomb in the tracks? Both long shots but technically feasible.

[sand digger]

There was the British Gammon AT grenade which I believe the US used sometimes.

[YankeeDog]

Gammons were issued to U.S. Paratrooper infantry prior to D-Day. I think it's a pretty safe guess that Moltovs (if they exist at all in CMBN; they're more of an Ost Front phenomenon) and Gammons, grenade bundles, etc. will be considered "AT specials" in the game, and treated accordingly. CMSF already allows units with Demo charges to "throw" their Demos at AFVs (I assume this is somewhat abstracted; I don't think anyone could actually throw a demo charge very far; they're heavy and unwieldy). Infantry with demo charges are actually pretty effective in CMSF against AFVs if you can get the infantry close enough to the AFV; it's the getting them close that's the hard part. So I think it's a pretty safe guess that short range Infantry AT weapons like Moltovs and Gammons, which can be thrown a short distance and/or placed on a vulnerable spot by some brave (or foolhardy) soul, will be similarly modeled.

Improvised solutions like "sticky bombs" I have no idea... I suspect there won't be anything in particular done to model stuff like this. These are tools of desperation, and if the scenario designer wants to model these they can always give a certain unit a demo or Gammon to represent a "sticky bomb;" it's not really that different.