From: Cologne, Germany
The CmdOps engine requires VL selection and point balancing, but I believe the scenario designers does this more to achieve proper point scoring than forcing behavior. (Of course, I could be totally wrong, since I never designed a scenario. However, I have read the manual and played lots of scenarios.)
I hate to disagree there, especially since your posts are often lengthy and really well-thought, and well-defined. But I think it's a bit different. If the AA (or CmdOps, if you will) engine wouldn't have the VL points, the game wouldn't just loose its base for the scoring system, but the (enemy) AI would end up "hanging" in the air as well, like we say, means it wouldn't have a base for its decision process where to commit troops or where to defend.
Given, except for Dave's rare snippets demonstrating one or another AI behaviour (i.e. unit facing and formation/disposition when commencing an attack), and maybe except for one or another posting where you tried to explain (or guess, sometimes) what might be going on under the hood, I don't know much about AA's innerds, code-wise. But I've tinkered with scenario design, so I'd say the AI wouldn't be able to defend a strip of let's say 30-40 km of "Hell's Highway" in Holland, like marching and counter-marching the road up and down, as it just happened 1944, if the designer doesn't arrange for this particular effect - by using a proper chain of Objective- and VictoryLocation-spots.
So, in my books, and despite the excellent AI in CmdOps, the VL points/objectives are forcing the behaviour very well, in many ways.
I saw this in my Cologne scenario (depicting an imaginary final fight for Cologne and the bridges across the Rhein in 1945) when the scenario's last 2 days ended up with creating a kind of anti-climax, as I set the VLs/Objectives in a way that the German AI would pull back to the right side of the river, consolidating its forces and waiting for the inevitable Allied para landings. Improper usage / allocation of these points will ultimately transform a good map into a scenario that won't put up a challenge in single player mode.
Writing this just created an idea in my head:
Is it possible to let the designer create Objectives/VLs as a line (he could draw) in the ScenarioMaker? Like a polygon line...pretty much like drawing a terrain layer: The designer could then draw jagged lines, elliptic lines, bulges, etc., he could even simulate troops that dug in (trenches), using staggered or jagged defensive positions / light fortifications. This would serve well for scenarios in Russia (i.e. where the Germans tried to cut the bulge at Kursk in 1943).
Dave, is that possible (maybe for the next installment? )?
If you had no order delays, then expert micro-management would always give superior results (note the emphasis on the word "expert"). You could force the cohesion of an attack by pacing units at the tip of the spear ... you could protect an exposed flank that was being hit by a spoiling attack ... But when you add order delays into the system, then you reach a point where the cost benefit analysis of no longer favors more micromanagement. .....  Effectively, the excessive micromanagement would totally cripple the player against either the AI or a human opponent.
So, in theory the expert micro-manager should always win and this would become an exploit which would totally negate the who hiearchical AI structure of the game. However, the brilliance of the system is that order delays prevents the pervesion of the game system.
I gotta tell you that micromanagement still is superior, and it still works even with max order delay, if you arrange for expected/unexpected events beforehand:
Let's say you detach a couple of companies (not neccessarily from the Bn or Rgt you want to commit to the planned attack), where you then issue the attack order to the Bn/Rgt scheduled to be the main body for the operation. Then, once you see that let's say the left flank of your main body ran into (unexpected) trouble (i.e. additional troops that couldn't be spotted beforehand, or troops that are stronger than expected), you can then send in these single companies (let's call em "special purpose"-companies) to stabilize the left flank. This all works even with max order delay, the player just has to know (an "amateur" might just not know it) that micro-management should not be applied to big/bigger units whose HQs have already planned and implemented the player's orders, as this would result in the HQ halting its troops and replanning the attack ... with an additional order delay, just as you described.
So, micro-management (with order-delay enabled) should only be applied to independent bodies: i.e. single companies or "bundled"/"grouped" companies (max. 2, to reduce command load) . That's the rule of thumb here.
The time frame for the decision/planning process may be the same, but the actual implementation for such units is faster - as less or even no sub-units units are involved.
What I just described is pretty much a man-made reserve pool, but, with the current limited AI capabilities regarding building/commitment of reserves, it would be desirable if the AI would be able to create (and commence) a dedicated reserve group.
Anyway, that said, micromanagement still works, and can be decisive in situations where either a human opponent or the AI may stick to initially issued orders, in order not to disturb / destroy the development of an attack.
But this is where I wish that there would be a variable order-delay. Once we go up in the command chain, it's surely realistic if an attack order, developed on the Korps level, will take it's time to make its way down the ranks, as historically
- 1) an objective had to be discussed,
- 2) possible outcomes, benefits and disadvantages had to be envisioned,
- 3) after the order (to develop a plan) had been issued, the Supreme HQ had to take all parameters into account:
Terrain, weather conditions, enemy strength, strength and supply situation of friendly troops, available resources (manpower, supplies - ammunition / fuel / food / amount of equipment/vehicles, transport), axis of attack, recon/intel data, level of confidentiality, plans for inclusion of occupation forces and military administration), general logistics, protection of extended supply lines,
- 4) once the plan had been approved, the Supreme HQ had to reserve/allocate resources and pass the plan down the ranks, where lower ranks often just knew the details affecting their particular units (i.e. company), to keep a certain level of secrecy,
- 5) then, prior to the attack, while the Rgt HQs coordinated the supply efforts, the Bn and Coy HQs then had to evaluate and report their status, and raise their hands in case equipment/replacements turned out to be insufficient for the scheduled attack.
That said, order delay makes sense, and even with just a few hours game time it's still realistic: Rommel had to develop and implement an attack plan (I think after Kasserine Pass?) within 2 days, for example, a plan that incorporated 3 units (almost 3 divisions, Italian Division, German Panzer Division, and Infantry), where he aimed at throwing out and pursuing the American units. Rommel couldn't convince his German superior to approve his initial plan, so he tried to bypass him by contacting the Italian High Command, which agreed to a revised plan that envisaged a different axis of attack, though. He then had to replan the whole operation within 2 days, as speed was essential, giving the US units time to regroup and dig in, though.
But in general, the German doctrine gave German officers quite some room for improvisation, as, in most cases, they were relatively free to pick the axis of attack, and type of approach, if they received an order to take an objective. This applied even to company commanders. So, if a company commander figured that the opposing unit decided to leg it, he would have diverted a part of his force (or even the entire unit) to pursue the enemy, in order to avoid that the unit got a chance to regroup. Taking them out of the battle, either by destroying the particular enemy unit or by capturing its members, was often essential. On a bigger scale, the numerous battles of encirclement and annihilation (Russia, France) aimed for the very same thing described above.
Ok that's where my idea of the variable order delay kicks in (which i expressed in the COTA forum around 2 or 3 years ago, i think): A company should be able to
- a) step out of the line without disturbing/interrupting the current attack carried out by the superior unit
- b) pursue a given enemy unit without any order delay, as "hunting"/pursuing an enemy won't take a sophisticated plan, as it's just running after an enemy who's routing,
- c) to withdraw way faster than currently (COTA) possible, as - once the player figures that a given unit won't be able to withstand another attack - the unit in question won't disengage, even if it's set to the lowest aggro/fire rate settings... usually, such a unit will retreat in no time, with the player being unable to save/control this particular unit. Somewhat frustrating. It often feels pretty much like the unit would take 10 minutes to "prepare" the run, with the enemy making it route or retreat within the next 8 or 9 mins. In real life, once a commander issued a "drop everything and move"-order, you bet that the soldiers actually did it without even thinking about it, maybe with the commander issuing an order to keep up a tiny group of rear-guards, covering the withdrawal. The CmdOps engine kinda won't allow for an orderly withdrawal, currently.
In reality, the penalty of excessive micromanagement is far greater than just simply the order delays as that is the most obvious and immediate issue. Excessive micromanagement will also result in an endless string of replans.
In WW2, micro-management performed by able commanders would have changed the outcome of those battles where local commanders acted faint-hearted or indetermined. In turn, micro-management in some battles, i.e. the Ardennes offensive, where the German High Command drew routes that had to be followed, even sometimes down to company level, hampered improvisation and timely local successes. For instance, there were empty villages (means possible bypass-routes) undetected by the Germans, which were only taken into consideration as alternative route (after 2 days) once patrols (on foot) had been sent out to recon the area, after the scheduled route appeared to be defended by US units, and all this was caused by a Bn commander who didn't dare or consider to deviate from the original plan.
It's not just black + white when it comes to micro-management, imho.
Take care Mark.
< Message edited by GoodGuy -- 5/12/2009 12:54:43 AM >
General Anthony McAuliffe
December 22nd, 1944
"I've always felt that the AA (Alied Assault engine) had the potential to be [....] big."
8th of August, 2006