Discussing the limitations of the current CM engine for simulating WW2 engagements

[Runyan99]

I just want to throw out the following quote from danielh, because I think it is a fair summation of the limitations of the current engine, as they compare to operations as they were actually carried out, more often than not.

I hope that by simply understanding these limitations, we may see some or all of them removed in the next generation of the CM engine.

6. CM centers around "stupid" frontal annihilation engagments normally avoided, unless one uses very big maps, but then again the relatively low movement speeds and short timeframes hinder extraordinarily. The average battle didn't last under half a day including movement to contact other then a classical storming out of the trench straight ahead as in WWI. Moreover a battle seldom was undertaken with smaller than Battalion sized forces, because small unit engagments are normally to be avoided because of their average high casulaty rates and therefore uneconomic.

[Michael Dorosh]

danielh seems to have no idea about actual battle doctrine in World War Two; certainly the battalion "frontal assault" was commonplace among Commonwealth units. In Canadian battalions, for example, the classic approach was with two companies up two in reserve. Two companies makes for a nice size CM game.

Where CM differes from reality is in the artillery model - CW doctrine was soundly artillery based.

As for battles lasting over a half a day as a matter of course, I can see Cory is championing his pet peeve again. It seems that no matter how many examples are provided, Cory still wants to see all CM scenarios drawn out to 350 turns. Wasn't so.

The "classic" company assault described above could be, and was, accomplished in under an hour.

http://members.shaw.ca/calgaryhighlanders/quescamb.htm

Has an example of a battalion advance. Bear in mind that only the contact made by the leading companies need be represented by a CM scenario. Hell, you could focus on a single platoon with a map small enough.

I would like to know what danielh's bona fides are - I wasn't aware that infantry battalions in WW II avoided contact as a matter of course. Certainly the history of the Canadian Army in that conflict, be it in Italy or NW Europe, was one of deliberate battalion attacks on firmly defended ground.

[ July 15, 2004, 04:12 PM: Message edited by: Michael Dorosh ]

[Kingfish]

6. CM centers around "stupid" frontal annihilation engagments normally avoided, unless one uses very big maps, but then again the relatively low movement speeds and short timeframes hinder extraordinarily.

This sounds to me like a problem with scenario design rather than limitations with the game engine.

[Sergei]

Originally posted by Michael Dorosh:

I would like to know what danielh's bona fides are - I wasn't aware that infantry battalions in WW II avoided contact as a matter of course.

Maybe it is in reference to Italians?

...sorry, I know it's wrong...

[Runyan99]

Originally posted by Michael Dorosh:

As for battles lasting over a half a day as a matter of course, I can see Cory is championing his pet peeve again.

Guilty.

It seems that no matter how many examples are provided, Cory still wants to see all CM scenarios drawn out to 350 turns.

Not all. Just some. Right now CM tends to represent the short end of the spectrum. CM Operations are good, but need some improvement.

You can quote instances of 30 minute battles, and I can find examples of 6 hour battles, and we can go back and forth forever. In the end, the fact is that both happened, so ideally we need the power to simulate both equally well.

[ July 15, 2004, 04:49 PM: Message edited by: Runyan99 ]

[Runyan99]

dp

[Sergei]

Originally posted by Runyan99:

Not all. Just some. Right now all CM scenarios represent only the short end of the spectrum.

How many people would actually want to play a 6 hour battle if it was possible? If that is only two or three, would that justify all the extra work required to implement better systems of resupply and reinforcement? I like to play something from 30 to 60 minutes long (unless it's an operation), below 30 for TCP/IP. I want to have my PBEM games finished some day, AI's stupidity accumulates by time, and I want to finish TCP/IP games in an hour or two.

Of course, maybe with game design it would be possible to have 60 minute games played in the same time and with the same amount of work as now a 20 minute game, in which case that would become the standard for quickies and the average length of a battle would be two or three hours.

[Runyan99]

Originally posted by Sergei:

Of course, maybe with game design it would be possible to have 60 minute games played in the same time and with the same amount of work as now a 20 minute game, in which case that would become the standard for quickies and the average length of a battle would be two or three hours.

That's kind of my idea. To open up the restriction of 1 minute turns, at least as an option. Maybe 2 minute turns, maybe three minute turns, maybe 5 minute turns, maybe turns where the computer decides the length of each turn, maybe variable length turns, whatever. Let's think outside the box of convention we have built for ourselves.

Only the We-Go convention of paused plotting and simulanteous execution should be sacred in the next game. One minute length turns is an arbitrary design decision.

[eichenbaum]

I find the limitations quite acceptable.

[Runyan99]

Originally posted by eichenbaum:

I find the limitations quite acceptable.

Of course it is. We've all accepted it. It's a great game.

On the other hand, we considered sail-powered ships acceptable for hundreds of years too, until we thought of something better.

[Runyan99]

Originally posted by Runyan99:

That's kind of my idea. To open up the restriction of 1 minute turns, at least as an option. Maybe 2 minute turns, maybe three minute turns, maybe 5 minute turns, maybe turns where the computer decides the length of each turn, maybe variable length turns, whatever. Let's think outside the box of convention we have built for ourselves.

Let me expand on that a little.

Consider the added FOW that unknown turn lengths would bring.

Imagine that the next turn might run anywhere from 30 seconds to 3 minutes, but YOU don't know exactly how long it will last ahead of time.

How would that affect your order plotting? How would it affect the realism of the simulation? Does this help to simulate the unpredicability of real warfare? Does it help to mitigate the omnipotent information sharing (between units on the battlefield A.K.A. Borg spotting) and micromanagement that players are currently capable of? Would this help to simulate the friction of dissemenating orders and getting men to move where you want them to? And of course, would this help us to simulate longer periods of engagement better than we can now?

[MikeyD]

For what it does it does it better than any other game. Most of the 'limitations' tend to wander into what it 'doesn't' do.

- No on-map out-of-LOS artillery fire (I'm mainly thinking of 75mm pack howitzers firing from reverse slopes),

-No forward air control ability,

-No laying of smoke screens (Churchills and late T-34s carried smoke generating equipment),

-No means to resupply during battle.

But most of that stuff is ancillary to the basic job of company-size tactical engagements.

It would be nice if the AI could be 'turned up or down' like some chess programs so the same scenario could yield a quick-and-dirty 15 turn slug-fest, or a 60 turn methodical game with the AI cautiously scouting, feinting, retreating, concentrating forces, etc. I guess that's why there's human vs human mode

[Berlichtingen]

Originally quoted by Runyan99:

Moreover a battle seldom was undertaken with smaller than Battalion sized forces, because small unit engagments are normally to be avoided because of their average high casulaty rates and therefore uneconomic.

So, why is it that I have at least 6 books that detail company or lower actions in WW2? Sounds like that bit of knowledge got pulled straight outta someone's ass

[Speedy]

I like sail powered ships.

[Andrew H.]

Variable (random) turn lengths are an intersting idea, especially if it were combined with C&C and the different turn length applies to different units.

For example - you have three platoons. On turn 1, you give each platoon orders. On turn 2, you are permitted to give the first platoon orders again, but the other two platoons are still carrying out their other orders and can't receive new orders. Maybe you can give platoon 2 orders on turn 3 or 4 (or maybe not), but you might have to wait longer to give new orders to platoon if, say, it's out of C&C. I'm sure this would have to be tweaked for AI issues, but it could be kind of cool.

I don't think it would help much with game length, though. I wouldn't be surprised if the average length of small unit actions were 4-6 hours - but if you read the AARs, there's lots and lots of waiting around. I'm not sure how many people would want to play a game like that.

[Nelson 1812]

Think it would be fun if move one ended with some sort of contact........

[Wisbech_lad]

Quite. Think of it as time compression. Rather than your infantry sitting around waiting for the tankers who are 30 minutes late, CM gives you the reinforcements after 5 minutes.

If you read the Omars battle AAR

http://www.lonesentry.com/battleoftheomars/

looks to me that the CMAK "battle" was short, intense and bloody, with a start line 300 yards from the objective. Of course, the getting into position, then regrouping took a long time...

[Joachim]

Cory,

personally I don't like the computer deciding when a turn ends. Gameplay is faster, approach march is in the game - but at which price?

The only exception being that moves without contact might take some more time. (IE turn 1 lasts unitl contact is made). But what happens if my plan is bounded overwatch? I'd need pauses triggered by events. I don't want my overwatch to move out at minute 15 - I want them to go once the advance party is in position.

Maybe the ability to have the player issue commands that upon certain events trigger the end of the turn.

Anyway... South of Kharkov is in turn 40+. It is a leisurely stroll across the countryside. Seems the StuKas got one plt one each side. Losses caused by enemy fire are less than 5 until now. 2 MGs are captured, 2 are killed, 1 is panicked and should have taken losses during its continueing dance of death. 1 FO team, 1 Plt HQ, 1 gun killed. (All on the right flank). Given that we play since last year this is quite an impressive score! Andreas and me are grimly determined to continue that fight. Right now there has been no action in the last few turns (except for that panicked MG in the center). A 5-minute turn would speed gameplay. But that MG... it is good to control the MGs firing on it now and then. Just keep it panicked. No need to waste ammo. Would need some SOP to ensure that if the turns last longer.

Gruß

Joachim

[Sirocco]

I'm not sure whether the references to "battalion sized attacks" isn't confusing what CM does. A company attack in CM doesn't necessarily mean that only a single company attacked. What it might well mean is that we are focussing on a single part of an overall plan.

I'm personally not so concerned with the handling of armour, but the handling of infantry, which is much more difficult to simulate. There is a lot of scope for improvement there, in particular the modelling of suppressive fire on the advance, dealing with cover seeking issues, and better modelling of battlefield communication.

[Jack Carr]

Six hour games? You may have to add some type of resupply element if the engine were changed to handle that length. Most infantry units engaged for any extended period run low on ammo pretty quick.

Not sure where this comes into the discussion but if you have a few squads that get caught up in a firefight for 5+ turns, which is not uncommon, they're out of ammo in no time at all. In a game that lasts so long what would these squads do for the remainder of the scenario? They would have to tactically withdraw, taking the inevitable casualties associated with such a maneuver and find a safe spot on the map to wait it out.

Resupply would be handy. I think this gets a little out of the scope of the current CM engine. As some have already posted perhaps Battlefront is already taking this into account for CMX2.

[junk2drive]

i like operations. some are very well done. ops = large maps in smaller pieces, fow unseen part of map, chance to pull back troops and equipment, change of time of day - weather, resupply of both sides. imho there is a lot more to think about in an op. a lot more planning ahead.

[Michael Dorosh]

Originally posted by Sirocco:

I'm not sure whether the references to "battalion sized attacks" isn't confusing what CM does. A company attack in CM doesn't necessarily mean that only a single company attacked. What it might well mean is that we are focussing on a single part of an overall plan.

Exactly. You can have a CM Battle depict only the assault by a single company. In fact, you have to draw a line somewhere. Even a battalion attack is simply - in many cases - part of an overall attack by a brigade or regiment, which may be only one facet of the attack by that Division, which is attacking astride another division under the direction of Corps headquarters...

[John D Salt]

Originally posted by Joachim:

Maybe the ability to have the player issue commands that upon certain events trigger the end of the turn.

[snips]

What defines a "turn"? It need not be defined by any particular length of time. The thing that matters is that it is a time for players to issue orders. To my mind, players should not be allowed to issue orders unless and until they have new information available to them on which to revise their plan. If the next CM were to cast the player in a single role as force commander (instead of making the player be all the commanders from battalion to section at once), then once the flow of information about the battlefield had been modelled, turn legths would be defined naturally. Every time a commander changes plan, that's a turn. All the time in between is a video under the control of the AI. Opposing commanders need not take turns at the same time, as they need not change their plans at the same time.

Under this system, a stupid commander with poor comms might have only one turn in the entire game. A commander who had put an appropriate number of decision points and conplans into his orders and was served by competent signallers would expect to do it much more often.

This is the way a computer game could really demonstrate what it means to get inside an opponent's OODA loop. In game terms, it would mean having more turns than the opponent.

All the best,

John.

[Ace Pilot]

Originally posted by Runyan99:

On the other hand, we considered sail-powered ships acceptable for hundreds of years too, until we thought of something better.

Personally, I prefer a sailing vessel to a "stink pot." But I see your point. Your idea on variable turn lengths is interesting.

[Pzman]

Originally posted by John D Salt:

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

Maybe the ability to have the player issue commands that upon certain events trigger the end of the turn.

[snips]

What defines a "turn"? It need not be defined by any particular length of time. The thing that matters is that it is a time for players to issue orders. To my mind, players should not be allowed to issue orders unless and until they have new information available to them on which to revise their plan. If the next CM were to cast the player in a single role as force commander (instead of making the player be all the commanders from battalion to section at once), then once the flow of information about the battlefield had been modelled, turn legths would be defined naturally. Every time a commander changes plan, that's a turn. All the time in between is a video under the control of the AI. Opposing commanders need not take turns at the same time, as they need not change their plans at the same time.

Under this system, a stupid commander with poor comms might have only one turn in the entire game. A commander who had put an appropriate number of decision points and conplans into his orders and was served by competent signallers would expect to do it much more often.

This is the way a computer game could really demonstrate what it means to get inside an opponent's OODA loop. In game terms, it would mean having more turns than the opponent.

All the best,

John. </font>