Units should be less willing to move/manoeuvre.

[FredSanford3]

I've noticed this, too.

ORIGINAL: fleischer

I think one of the factors that contribute to units "milling about" is that the AI often seems to be very poor at taking units' current position into account when deciding who goes where in a formation. Perhaps this is even a bug of some sort, because often the AI will "flip" the current formation around when given a new order(i.e. the leftmost unit changes place with the rightmost unit, and so on). Consider this situation, for example:



Notice that the unit at the bottom is given the topmost position in the new formation while the unit at the top is given a position at the bottom. It is as if the AI does the opposite of what would be logical. And this example is just the first and best I could find, and far from a worst-case.

I imagine this has quite an impact on fatigue too, especially with foot units in larger formations.

[Arjuna]

The allocation routine for formations does factor in proximity, but it factors in a lot of other things as well. There is certainly scope to tweak this further. But if we go too far it then will leave ineffective units on the line just because they are there already, when in fact they should be swapped out of the line.

[wodin]

I would like to see things like Inf Gun support units going in with infantry support than moving in on their own...most of the time they run into an Inf coy and get wiped out quick when they are moving in...if they came in just behind an Inf coy this would help....I also think Mg coy's should be placed further away tham they are now...I've started micromanaging them...

Not sure if it's me but it appears it makes more boobies when there are alot of units...

[Arimus]

It is really to early to say for sure, but it appears that battalion formations do really well. Larger formations do some strange things. For now, I am sticking with giving orders to battalions and reinforced battalions.

[fleischer]

ORIGINAL: Arjuna
The allocation routine for formations does factor in proximity, but it factors in a lot of other things as well. There is certainly scope to tweak this further. But if we go too far it then will leave ineffective units on the line just because they are there already, when in fact they should be swapped out of the line.

Arjuna, this problem is far too consistent to have anything to do with outswapping of fatigued units etc. In this particular example, the two units in question were almost identical, both in strength and type, and there were no enemy units nearby.

The problem seems to happen alot when formations make alot of direction changes, i.e. assault in one direction, retreat in the opposite direction and then defend in the first direction again, and also when units with different facings are attached together. It is as if the AI rotates its formation instead of just mirroring it when it changes direction.

[Chief Rudiger]

Also,

When Regt/Bde are given orders they don't always form their sub units up logically. In my screen shot (upfiles/32312/4D128F81D0964F4886E5CB93B571BCD0.jpg) in this thread (tm.asp?m=2489614) you can see the six companys of the 753 Regt, to the North, are formed up in two discrete bodies, with their respective Bn HQs, whereas the companys from the Regt are mixed up. I didn't check at the time whether Coys 1,3,7 were now under control of II Bn, and ditto for I Bn, or not, but it means that if you want to re task one of the Bns (say to a flank attack) the Bn will not have its companies near at hand, and the other Bns attack will have to be replanned! Fair enough a replan will be required now the the Regt has only half the Bns but the effect would be less if the two Bn's companies were kept seperate.

What good reason has the AI for not keeping companies under their repective Bn HQs?

[henri51]

I am not defending nor complaining, but we should remember that in real life, commanders and sub-commanders DID make mistakes, some of them serious.

It is not for nothing that Bradley fired more commanders than anyone else in the US Army, and Patton did not hesitate to fire some either (not to mention Hitler, who fired just about every top German commander including Guderian and Manstein for incompetence - and often rehired them again.

As a matter of fact, I would like to see a greater effect of HQ competence (although I have no competence how it works at the present time). Clearly a unit commanded by Panzer Meyer should perform better than one commanded by Obersthochtoifelsturmfuhrer Dumbkopf...

Henri

[wodin]

I have found playing the bigger scenarios you don't notice all these problems as much because your concentrating on the bigger picture...and it usually goes the way you want....

It's when your playing the smaller scenarios so your really taking notice and analyising the units that you see odd behaviour which usually in the end sorts itself out but it's alot more noticeable.

[Chief Rudiger]

I gave up on Bde attacks in COTA because the AI seemed to do ridiculous things too often. I avoided the larger scenario is COTA because of this as i hate managing anything below battalion level, but enjoy watching the details, and like to issue as few orders as possible. Because of this i find BFTB a hell of a lot better, but it still has its disfiguring bugs.

In the tutorial, when i gave CCA an order to move to St Vith and attack at its disgression on the way, and it did it, i was amazed. When the bugs already identified get fixed, which i have no doubt they will, then resolving little issues, such as the muddling of subordinates, will really put a cherry on top.

[wodin]

ORIGINAL: Chief Rudiger

In the tutorial, when i gave CCA an order to move to St Vith and attack at its disgression on the way, and it did it, i was amazed. When the bugs already identified get fixed, which i have no doubt they will, then resolving little issues, such as the muddling of subordinates, will really put a cherry on top.

I've said exactly the same thing a few times....enough infact to have got me into trouble[;)] Just trying to put it into perspective....things are much improved.

[Arjuna]

Re allocation of companies to Bn HQs. For BFTB we overhauled the ForceAllocation() at least five times that I recall. For HTTR we had a top down approach which more often than not did keep companies assigned to their organic Bn HQs. However it didn't cut the mustard when it came to apportioning smaller forceGroups to tasks that didn't warrant a whole Bn. In COTA we tried various methods that were better but still not really good enough. For BFTB we initially went to a bottom up approach, allocating company units first and then allocating their Bn HQ second. This worked a lot better at handling multiple objectives. Over ensuing revisions of this code we have added code to favour Bn companies staying together. But it doesn't always work out the way if some other suitability factor ( eg proximity, fatigue, supply, cohesion etc ) is given precendence. If you fail to do this then you get other serious problems like one of the companies being isolated by a river or being too tirted to move. If you don't allocate him then you often need to allocate some othe unit to provide the necessary combat power/capability. Hence you get cross attachments. It's not perfect and I consider this a work in progress. We'll endeavour to refine it further.

BTW if you read the unit histories you will see that on many, many occassions part of a Bn was hived off to another force and other units were cross attached. This was the norm, not the exeption. Why? Because operational warfare is not a series of discreet actions where everything can be coreographed just so. Rather it is a web of actions, some starting together others not. In this semi chaotic framework commanders are ordered to grab whatever forces are effective and to hand. Often they have to let part of their force continue with what they are currently doing ( eg defending X or attacking Y ) while they take the other part plus some ring-ins from neighbouring forces to do an urgent job like containing a breach or exploiting a gap. In this way Bns often get intermingled and only regroup when they can draw breath at some pause in the battle.

[DanOppenheim]

I've been training myself out of the neat-and-orderly mentality that I've always had when playing strategy games. I can't help but try to form them up into parade ground order. Disarray causes dismay! I think this style of play was allowed and encourage by other games. Now I'm starting to form much more heterogeneous task groups, in which I'll assign tank companies to infantry battalions and vice versa or bulk up a battalion with companies from a sibling battalion and use the lighter one as the reserve. I find there's little point saying "I'll send a battalion to do X", when I really need to be asking "what is X, what's it going to take to get it done and what can I put together to do so?".

All of which reminds me, I need to go and bang my head against Manhay Crossroads again.

[daft]

DanO,

Some excellent points there in regards to force composition. Trying to analyze what it is you actually want to achieve and why. Then compose a force based on this and the most current intel on the enemy forces at or close to the objective. If available, I also set aside forces to exploit any breakthroughs or to occupy the objective after it is taken so that the assault force can move on to the next objective. I hate letting the enemy reorg and complicate further advances by my forces.

[wodin]

Manhay...shudder....twitch...twitch....blink....shudder....Crossroads....twitch.....

[Chief Rudiger]

Like DanO i've also been trying to get out of the mentality but, in the example i gave in my original post, the attack was a set piece affair that i'd waited three days to assemble and rest my troops for. The troops moved to the pictured FUP from a distant bivouac rather from say from an already assembled defensive formation, in situ, so i can't understand why the AI choose to swap around the companys.

What yourself and DanO said reminded me of what i read recently in Gen Horrock's XXX Corp history "Corp Commander". Here he states that:

"The two armd div (11th & Guards) operated quite diffferently. The 11th was the more flexible of the two, 'pip' roberts, one of the most experienced Tank Commanders in the Army, would switch his troops rapidly, attacking with tanks only, tanks and infantry, or infantry supported by tanks, according to the situation. The Guards, on the other hand, always worked in regimental groups, Coldstream Tanks and Coldstream Infantry, Grenadier Guards with Grenadier Tanks, etc. This system was not calculated to get the full value from the mobility of an Amrd Div, but it ensured the closest possible cooperation between the two arms, as the infantry and tank crews all came from the same Regiment and had more or less grown up together. Neverthless, i preferred the standard Armd Div practice of the 11th. Fortunatley, however, by now I had gained sufficient battle experience not to interfere with the way a Div Cmdr handles his command."

Perhaps, being British and having read, for the most part, British military history only up to the end of WW2, from British military historians my gaming is a product of that "cap badge snobbery" and conservatism that went some way to retarding tactical flexibility and the acceptance of new technology. Certainly, US "Combat Commands" and Jerry "Kampfgruppe" always look terribly messy to me. They certainly sound like a storeman's worst enemy.

Surely all this mechanisation nonsense is just the result of a drop in musketry skill, abysmal New Amry "junior officers, NCOs and ORs", or the forgetting of Good Old British family values...

[;)]

[RayWolfe]

ORIGINAL: Chief Rudiger
Certainly, US "Combat Commands" and Jerry "Kampfgruppe" always look terribly messy to me. They certainly sound like a storeman's worst enemy.

Agreed.
My OCD reacts very badly to all this swaping about .... AND they don't all face the same way. I may have to go back to hexes. [;)]
Cheers
Ray