Carrier design

[Mad Russian]

Jason, I should back up. When I say that the problem is with the gamer, I meant that the system allows for us to do the very things you point out.

There are lot's of things that CM gets wrong. The list is well known...

God view and God control being at the top of the list as far as CCC modeling is concerned.

But to say that CM doesn't get the casualty part of war right I think is going the wrong direction. IMHO, it is one of the things that the game gets right.

Charge a HGM over open ground and get shot to pieces. That is accurate.

Move to within 200 meters of a functioning

gun/tank and for the most part get hit. That part is pretty accurate.

What is not accurate is the advance without regards to casualties. As you point out this is a game function. Depending on what the battle/operation level that is chosen by the designer there should be some real differences in what an assault, attack and a probe look like on the ground. An assault is what you get in every CM situation. You should expect to take heavy casualties and you will. An attack the computer should be less tolerant of losses and a probe should accept very low rates of loss.

Today that doesn't happen.

Sorry for misstating my position earlier.

[JasonC]

CM has per shot effect about right. Engineer types typically think analytic-atomically, that if they get the lowest level incident correct, the results will scale to whole battles. Doesn't work that way.

First a few quibbles about per shot stuff. When I assault an HMG over open ground no I do not get all shot up. I kill the HMG easily for the cost of a few casualties, usually less than in the HMG team. But the procedure I use is not "charge", it is slow advance to exploit rally power.

Artillery gets too much effect per shell overall, largely because the effectiveness against men who have already gone prone is overstated, particularly for the large shells. Those also get too much benefit from exact placement - it should plateau at within 10m or so. (The targets are not 2m points, in reality).

Snipers miss far too often (they should hit 20-40% of the time, depending on target cover), and in general light fire at men with poor cover is too unlikely to hit, while the morale effect is about right. Same for MG fire at moving men, and HE burst effect at longer ranges from the impact point.

The flip side of all that is just being prone should provide significantly more protection than it does. Largest effect against HE, strong effect against infantry fire, least effect against the most aimed fire (e.g. snipers). That would encourage going stationary and staying that way, and use of crawl-sneak. Right now there is no point in it.

Overall, the reason casualties are far too high - and they are - is not these per shot tweaking issues, but overmodeled bravery. Units recover too rapidly and too fully from pins. A unit that cycles through that several times should have its morale capped at alerted, eventually capped at cautious. Rattled behavior should be more common and spread when overall losses mount. Rally power through moderate fire is a bit too strong as it is. These again are per unit tweaks, useful but not the main issue.

The main issue, to repeat, is lack of morale impact of fire on larger units. Nothing above the squad cares what has happened to all the other squads. That is completely unrealistic, and it is exploitable. Units can be shoved into situations where they are completely used up, then fresh ones sent. The remnants of the first can rally completely, and fight "ugly" under higher HQs later in the fight.

The result is a large formation is significantly more robust than its constituent parts. No modeling of higher level morale effects means break-away squads and fearless battalions. The fearlessness of the battalion sets the overall ferocity of the engagements, and with it casualties. The reason real units do not inflict such high losses on each other is the men will not collectively stand it, giving out and refusing to continue the mission aggressively, far sooner than our toy soldiers.

As I said, it is a known design issue with grand tactical step loss games. They regularly get high level robustness wrong for correct values of individual engagement variables. Only higher level morale and cohesion systems can get the higher level robustness variable right.

Note this arises in grand tactical games because those put 1-2 extra echelons under player control, for the overall level of the engagement. Which means pushing the game primitive, per unit morale effects well below the level the commander cares about. This insulates the commander from real morale effects, if no aggregate morale system is added as well.

Concretely, my CM plans involve company attacks and small scale maneuvering by platoons, neither of which has any directly modeled morale state.

It arises in step loss games in particular because those add another layer further to modeled direct effects of combat. Shots do not eliminate units, they dock them slightly while leaving them effective. Per unit morale modeling can create a half way to eliminated state in morale terms, while allowing recovery. Which can be exploited by cycling subunits. Thus, enemy actions ding replacable parts but spare the actual components of any reasonable plan.

The atomistic expectation is that repetition of those enemy actions is supposed to add up to appreciable scale effects against the unit scales the players are actually manipulating. Each shot barely dings one squad, but n units fire 5 times a minute for 10 minutes, and it is supposed to all add up. Against particular subunits it does. Against the whole force, it only succeeds in dislocating an enemy plan when half or more of his individual pieces have been thoroughly trashed. Which is wrong, empirically.

The details of a solution one can quibble over, and many alternatives are possible. But the issue is real, its cause is known, it is a game design issue not people "playing it wrong" nor focusing on special war events. And the sort of corrections that work are known, though naturally their scale has to be tweaked and tuned.

The correction formula is some broad morale impact to units beyond ones actually hit directly, from cumulative damage to other subunits. This can come from contagion effects, command and control effects, direct morale reductions, impaired morale recovery, loss of allowed actions or orders etc. Many possible mechanisms, one goal - to impair the functioning of the whole as its parts break.