CMAK REQUEST: Organised Withdrawal

[MAsta_KFC]

Currently in CMBB/CMBO we have an withdraw command that gets out little digital men out of danger zones quick fast, at the penalty or lower morale and the chance of getting shot in the back and suffering horrendous casaulties.

I was wondering if BFC will consider adding or splitting it into 'Organised Withdraw/Retreat' and 'Immediate Retreat'? I can think of several pros and cons with this command so it won't be abused.

ADVANTAGES:

* Troops do not suffer such high morale penalties. So rather than have several squads of panicked/broken troops, a successful fall back maneouvre ensures they don't get trampled. They can fall back to a whole new set of trenches/foxholes to wear down the attacker again or entice him to overextend. Gives rugged defence a whole new term.

* Troops fire back. This can be useful again for overzealous attackers. Also gives an aspect of 'covering fire' to other troops who may not have the opportunity at the moment to withdraw.

DISADVANTAGES

* Troops still suffer morale penalties. This command will still be useful when the outcome of the battle is uncertain. If your being chased by Tigers, being shot upon by overwhelming troops, your troops will still break, run and die.

* Reduced accuracy/ROF/suppression. Troops concentrating on withdrawing and firing do not perform as well as opposed to firing from a static position.

* Reduced speed of withdraw. As opposed to straight flight, troops may be exposed to enemy fire for greater periods of time because the need to keep some semblance of order. Commanders may wish to juggle this between ordering immediate withdrawal depending on the situation.

* Fatigue. It's tiring to not try to trip over your heels, looking backwards and firing forwards

* Reduced awareness. You're not going to spot much withdrawing. Then again, unless you've got eyes in the back of your head, at least if beats turning and running.

It may SEEM that the disadvantages greatly overwhelm the advantages. This can be used to prevent possible abuse of this command. HOWEVER, I believe that this command will be extremely useful if used at the right time and situation.

Imagine wearing down an attacker as he must fight every inch of the way to advance. As he thinks he has defeated your troops, to his dismay he must face a seemingly whole new line of defenders again. And again. He may be tempted to do something imprudent like rush his men/tanks forwadrd, only to be decimated and open to counterattack. Or it may work conversely. He may believe your troops are preparing another line of defence, hence interrupting his momentum. He takes time to scout and recon this new threat. In fact this gives you time rally routed troops and move into better defensive postions. It's all a mind game.

So what do you think BFC? Sorry for the long winded post, but knowing your attention to detail, I don't think you would pay attention to something that wasn't fully thought out.

[Joachim]

An organized withdrawal of scouts or single squads - yes. (IE keep running if shoot at as it is safer to run that stay behind). No reduction of command delay - this is organized, after all.

A company has to organize its withdrawal all by itself by using boundary overwatch in reverse. If troops keep running backwards, this is possible. Apart from the occassional panicked squad, of course. Organization under fire is very rare.

Gruß

Joachim

[SFJaykey]

The "Withdrawl" we have in CMBB v1.03 has worked pretty well for me, as long as the withdrawing troops are in command and moving through some cover, or at least not in LOS of powerful, nearby enemies.

For situations like you are describing I will often use "Advance" to withdraw. Kind of like the old story about "attacking in another direction."

[MAsta_KFC]

Originally posted by SFJaykey:

For situations like you are describing I will often use "Advance" to withdraw. Kind of like the old story about "attacking in another direction."

Yes, but if u 'Advance' backwards, u still present your back to the enemy, and in CMBB if you're getting shot at from the rear, you take major morale hit. In 'Organised withdraw' your troops are still facing the front.

And I meant 'Organised Withdraw' is only a command to be used for squads and HQ, not for tanks and support weaponry.

[SFJaykey]

You're right, though the morale hit seems less than when using "move" or "run." I haven't done any tests...maybe its no better than using "Withdraw."

Re: the name of your proposed command, how about "Fall Back?"

[zukkov]

cowards! you should be ashamed of yourselves! in my army, it's either advance or die! naturally, i don't win many battles, but that's besides the point. lol...

actually, most of my guys tend to end up groveling on the ground, heading to the rear, with their tails between their legs, all the while their faces are appropriately eating dirt. grrrrrrrr... cowards! the whole damn army is made up of cowards! :mad:

[Moon]

Somebody's gotta turn their backs on the enemy when withdrawing (walking backwards while firing is a little Hollywood...)

An organized withdrawal is something you would usually perform on the platoon (at least) level. One squad withdraws while the others are covering by suppressive fire. It works that way in CM with the orders at hand.

The existing withdraw command is intended to be an emergency exit, mainly geared towards allowing players to withdraw BEFORE impending doom (e.g. you see your guys run into something bad towards the end of an action phase and yell "fall back, back!").

Advance works quite like it should in this respect (in fact Assault might even be better!), if you ask me, allowing outgoing (covering) fire (the little that a squad has available when it's not covered by other platoon elements).

Withdrawing under fire is one of the most difficult tasks for any military unit, and no combination or addition of orders is going to make it much easier (and shouldn't).

Martin

[Sgt. Schultz]

The way it is now works fine with a little cover, like it is stated above.

I don't know how I would take seeing my little pixelated troops moon-walking to the rear while firing from the hip. I would be constantly zooming in looking for John Wayne.

[SirReal]

I was under the impression that withdrawing under fire is about the most difficult thing you can do with an infantry unit.

I think the game correctly reflects this. Firing effectively at the enemy while moving in the opposite direction... well, perhaps elite troops can do so, but certainly not regular infantry. You do bounding overwatch, as already stated. And I think this can be done in the game right now.

/SirReal

[Ligur]

Hmm. I think advance, move, sneak and the rest are enough to disengage when you have time to organize a withdrawal; when you don't you use "withdraw" command (hit panic button GET OUT GET OUT! IS-2 BEHIND THE HOUSE, 2 minutes to get the hell outta here!! GO!). And when the situation is too thick to use neither just admit you are under too much incoming fire to move without a total rout & a lot of casualties (i.e. when you use withdraw under intense fire and the whole platoon routs... but some may still live).

I see no pressing need. But if one more command for intanfry would be in the works I would gladly have it anyway

[YankeeDog]

I'm not sure I see the need for an additional "Fall Back" command, but I do have one minor issue with the way the current engine models movement under fire.

As near as I can tell, the current model does not take into account 'threat direction' when calculting the effects of incoming fire.

Here is what I mean: Currently, if you have an infantry squad pull back using ADVANCE orders, and it suddenly starts taking fire from its rear quarter from a previously unknown enemy unit, the squad will experience serious morale effects due to the fact that it is taking fire from behind.

However, if the exact same squad attempts to ADVANCE away from a known source of enemy fire, as near as I can tell it still experiences the exact same morale penalty as it would if the source of incoming fire was heretofore unknown.

This, IMHO, is a bit unrealistic. An infantry unit moving retrograde relative to a known and expected source of incoming fire should experience *less* of a morale penalty than a unit moving and suddenly taking unexpected fire from the rear. In fact, I would argue that, generally speaking unxpected sources of fire should be more rattling to a unit's morale than expected fire from know enemy units, no matter what bearing of the incoming fire. I do believe that, in general, fire from the rear should be more damaging than fire from the front. But whether or not such rear-quarter fire is from a known and expected source, or comes as a surprise, should also be a factor.

What is needed is some way of modeling 'expected threat direction' in the game. I don't think this would necessarily require a new command. for units not under fire and without any other orders, 'expected threat direction' should be the same as facing (the way it is for all units now). Once a unit starts to spot and/or engage the enemy, it should probably set its own 'expected threat direction' based on known enemy contacts or sources of fire. Incoming fire from the general vector of a units 'expected threat direction' would be less damaging in terms of morale and casualties than fire from other vectors.

Assuming a unit was already engaged and under fire from enemy units that were to some degree spotted, 'threat direction' would allow the game to model, for example, a unit moving retrograde from a known and expected source as less likely to get rattled than a unit moving retrograde and suddenly coming under fire from an unexpected angle.

For units not under fire, threat direction could be set by the player with the covered arc commands. Basically, assuming no known threats or sources of fire, 'expected threat direction' should be through the center line of any set Covered Arc. 90% of the time, the threat direction you want for a unit in a given situation is more or less the same as the center point of a covered arc that you would set in that same situation. So if you wanted to move a squad retrograde, but strongly suspected that it would come under fire while withdrawing, you could set a covered arc in the direction of the most likely source of fire, and so somewhat mitigate the effects of any new source of fire from that direction.

Setting the 'expected threat direction' through the Covered Arc command like this would be the game equivalent of issuing an order along the lines of, "Lieutenant, pull your platoon back to that tree line, but be careful, we suspect there's enemy units on that hill to your North, so watch out for fire from that direction as you cross the open ground." This strikes me as entirely reasonable and realisitic. A platoon receiving such an order would be much more careful to avoid exposing itself to fire from the hill in question than a unit simply given a nonspecific 'fall back' order.

Note that 'threat direction' modeling would also help in situations in situations other than retrograde movement. For example, if you were trying to flank an MG, you could set covered arcs to let the flanking squads know that they should expect fire from the direction of the MG while moving. Of course, if a squad on such a flanking mission suddenly started taking fire from a previously unknown enemy unit along directly in front of it, this new source of fire would actually be more damaging than normal I think this is entirely realitstic, though.

I don't know if this kind of refinement would be possible for CMAK, but it would be nice. I would really like to see some kind of modeling of 'expected threat direction' for CMX2, though.

Cheers,

YD

[Lt Bull]

I have argued before how an "organised withdrawl" type of command would be a good inclusion into CM (please read HERE and HERE)

I beleive that in CM, infantry units wanting to disenage from the enemy are perhaps too unrealistaiclly disadvantaged given the current game orders/commands/morale/unit facing constraints.

Not convinced? I have set up a little test to show that it is SAFER to issue infantry units an ADVANCE command that sends them straight towards incoming enemy fire than an ADAVCNE command that sends them directly in the opposite direction from the enemy fire (ie. disengage).

A platoon of German infantry is facing a platoon of Russian infantry:

1) In THIS file, I have given them orders to ADVANCE across open ground to a rubble tile in a direction that leads them directly towards the enemy (decreasing the range).

2) In THIS other file, I have given them the ADVANCE command but they are heading in the opposite direction, directly AWAY from the enemy, across open ground to another rubble tile (reducing the range).

Draw your own conclusions.

Lt Bull

!!!!!!!!!!!!!!

IMPORTANT NOTE ABOUT SAVED GAME FILES: Becasue of my ISP constraints, I can not host any files but JPG and GIFs. Right click the link, select "Save Target As.." , rename the file extension from .jpg to .cme. Put them in the CMBB SAVED directory. When you open the files, just press OK at the password screen.

!!!!!!!!!!!!!

[ July 19, 2003, 02:32 AM: Message edited by: Lt Bull ]

[Sailor Malan]

I think this may be one of those seriously academic issues in a CMBB timescale/game scale.

1. If you need to withdraw a significant number of forces literally under fire, you are probably in trouble anyway. To matter, they have to withdraw, re-organise, possibly move, and attack again. All in one scenario. If you haven't time to do this, all that matters is the total casualty count (for VP). So it may seem like an issue, but is it really?

2. In RL, withdrawl has a negative effect most of the time anyway (except with high quality troops)/ (the 'oh @@@@, we've crossed that open area to attack once, now we are being pulled back, must be serious, and we get to cross it again', right up to 'all is lost..' - which is covered by the withdraw command)

3. Exactly what are the squad doing with this new command. The only possblilty I can think of is fire and movement by sections (i.e. advance/assault), but with a known threat from the rear, i.e. the 'direction of fire' moral loss might be lessened. On the other hand, not all incoming fire is equal, and I could easily think of anomalous situations where the command caused as many problems as it solves. Imaginine a failed ambush...squad opens fire at enemy, or is spotted prematurely. Enemy player retreats (or what ever). How are you going to cover the diference in threat caused by the jittery squad (seen) vs the remainder of the company (not seen). And how are you going to deal with 'retrograde retreats', i.e. attacks using the retreat command, and gaining a bonus (or loss of a hit) caused by new threats.

IMHO, not an issue that needs fixing for CM size games.

[Andreas]

Originally posted by Lt Bull:

Not convinced? I have set up a little test to show that it is SAFER to issue infantry units an ADVANCE command that sends them straight towards incoming enemy fire than an ADAVCNE command that sends them directly in the opposite direction from the enemy fire (ie. disengage).

Looked at them, still not convinced. In your test you are not using bounding overwatch. Bad tactics. To punish you for it your guys trying to withdraw should lose cohesion, and the morale effects are correct, IMO.

[zukkov]

it's discussions like these that makes me love this forum! it seems for every grog arguing some new feature needs to be added, there are an equal(or greater) number that disagree. would a new "organized withdrawal" command be realistic? for that matter, would any new commands that have been suggested and not yet incorporated be needed? who knows(except the shadow of course)? but if there were no way to improve the game, then i doubt seriously that bfc would be working on their next generation engine as we speak(you are working on it as we speak, right? hey, sleep time's over! *cracks whip* mush you dogies!). i doubt they're doing this just to make the game prettier(though that alone would be great. well, that and FULL MOVIE PLAYBACK!!!). there will be new features and i'll bet a few new commands as well. perhaps even an alien spacecraft will join the fray. then again, perhaps not....

[Lt Bull]

Originally posted by Andreas:

Looked at them, still not convinced. In your test you are not using bounding overwatch. Bad tactics. To punish you for it your guys trying to withdraw should lose cohesion, and the morale effects are correct, IMO.

Andreas, the point of the files is to demonstrate that in CMBB, it is "safer" for an infantry unit to move towards enemy fire than it is to move away from it. Try perhaps only having 2 sqds move while the other 2 stay an give cover fire. You will see that regardless of what you do, once your squads turn to face away from the enemy so that they can move away from the enemy, that is enough to basically doom that unit in a way I think disadvantages ANY infantry unit wanting to move AWAY from enemy fire under ANY circumstances.

By "safer" I mean the following:

-more likely it will sucessfully complete it's movement order by moving from Point A to Point B without getting pinned/panicked/broken/routed

-more likely to remain a cohesive functional unit at the end of the turn

With CM infantry, once they are engaged, it is like they are doomed to a fight to the death. There is no turning back, LITERALLY. Once you are engeged, the safest direction to move is TOWARDS the enemy. Disengaging by moving back across the same terrain you advanced across to get you where you are is WORSE than advancing further towards the enemy.

I also believe that this "no turning back" with infantry in CM contributes to casualty levels that far exceed what "realistically" you would expect for enagagements at the CM scale.

Lt Bull

[Andreas]

I think the main point here is that without suppression of the enemy, your units should suffer a lot if they withdraw, more so then when they attack, since they do not even have their own fire to suppress the enemy (I am assuming that shooting from the hip while walking backwards is pure Hollywood). Your test does not test for that, it tests for what is the equivalent of 'let's get out of here' with no support. In that case, I find it acceptable that my men would be happier to attack forward than to retire while being shot at.

The more fundamental point though maybe that borg spotting makes it unrealistically difficult to just melt away from the enemy, by dropping out of sight of just one unit, and thereby reducing the volume of incoming fire. That is an underlying problem, and I seriously doubt that a 'controlled withdrawal' command (whatever that is supposed to look like or do) could negate this effect.

That soldiers are less happy to be shot in the back then they are about attacking is IMO just as it should be, and gels with what I have read about combat. Disorganisation is the price to pay for trying to get out of a tight spot in a hurry.

Finally - unrealistic casualty levels have many reasons, but I doubt that withdrawal problems feature highly. If I see that the game is up during an attack, I ask my opponent for a cease-fire. If he accepts, the game ends with quite low casualty ratios. If I wanted to withdraw to fight elsewhere on the map, then I think it would be up to me to execute a competent withdrawal, using bounding overwatch, suppressive fire etc. I would not expect the game to do it for me anymore than I expect the game to organise my assault.

[YankeeDog]

Interesting point about borg spotting making withdrawal under fire more difficult and casualties in general higher. Makes sense to me.

I still think some sort of 'expected threat direction' modeling would add to the realism of the game and probably make withdrawing under fire *a little bit* easier.

In addition to these issues, I also think part of the problem is that the TacAI isn't very good at recognizing when a unit is 'almost out of danger'. While a unit trying to withdraw across open ground will usually try to make it to the next decent patch of cover if it goes 'pinned' or 'panicked' due to incoming fire, the TacAI doesn't appear to be able to recognize when a units is about to reach relative safety due to, for example, making it over the crest of a rise, or behind a row of buildings, or whatever. The result is that units will sometimes hit the dirt in very bad cover, or even reroute themselves to marginal cover like a nearby patch of brush, rather than moving just a few more meters into the greater safety of full defilade.

Especially for Green and Conscript units, I actually find this behaviour pretty believable. Low quality units shouldn't be too savvy about figuring out how to get out of danger, and having bullets nicking the dirt around you probably isn't too conducive to logical thought processes. However, more experienced units should be able to recognize that making it over a rise and out of view of the enemy will largely put them out of immediate danger. Therefore, Veterans and above should probably be more likely to push a few more meters on the withdrawal if the relative safety of defilade is nearby.

I imagine that this is a rather difficult AI behavior to program, though. For one thing, it requires the TacAI to make some pretty advanced analyses of terrain and incoming fire vectors in order to evaluate where dead ground lies. It may be asking too much of the AI engine to make such evaluations, especially 'on the fly' for multiple units multiple times during a turn.

Still, it would be nice. . .

Cheers,

YD

[Lt Bull]

I conducted a series of tests to get some data on this. I used the files I posted above. I probably could have designed the test a bit better to keep it consistent (eg. remove the HQ unit, as units withdrawing got within CnC range and perhaps benefited).

Anyway, here are the results. Basically I divided it up into

1) casualties

2) whether the units made it to the target rubble tile

3) the state of the unit

The total results seem to clearly show that although the casulaty rate is lower advancing away from the enemy, there is a 10% greater chance that your infantry squad will not get to the rubble tile and 7% greater chance your unit will break/rout, as opposed to advancing forward.

The results for the withdraw command make it look like an auto deathwish command for your infantry. However, it should be mentioned that at the begining of each turn, each unit was unspotted and in an OK state, except when the wthdraw command is given, which can instantly panic or pin a unit at the begining of the turn without an enemy or shot ever being fired.

Lt Bull

[ July 23, 2003, 09:05 AM: Message edited by: Lt Bull ]

[Andreas]

So you are not of the opinion that suppressive fire has a role to play in achieving a decent break from the enemy?

[Joachim]

Originally posted by Andreas:

So you are not of the opinion that suppressive fire has a role to play in achieving a decent break from the enemy?

Maybe he has the opinion that suppressive fire should work for advance and retreat exactly the same, so the result of his test that retreating causes more causalties (and suppression) than advancing will hold with suppressive fire alike. But the firing Soviet plt represents just the unsuppressed element of the enemy.

For the retreat a viable test would see 3 Soviet plts firing on that 1 moving GE plt while a single other GE plt and maybe a HMG42 provide covering fire. I'd bet that there is at least one plt of the Soviets firing unsuppressed. Thus I would expect more or less the same results as in MASTA_KSF's (sorry, just had to change those letters) test (Force ratio about 1:1.5).

On the advance, I'd put 2-3 MG34 and a mortar in support, have null unsuppressed enemies and expect less casualties (and suppression). (Force ratio about 1.5:1).

If there is a favorable local force ratio for the defenders, suppressive fire would reduce the effects on the moving plt, too. But I'd guess the overall tendency would still hold - the effect is proportional to the amount of unsuppressed fire.

Gruß

Joachim

[Joachim]

Originally posted by Scarhead:

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

So you are not of the opinion that suppressive fire has a role to play in achieving a decent break from the enemy?

Maybe he has the opinion that suppressive fire should work for advance and retreat exactly the same, so the result of his test that retreating causes more causalties (and suppression) than advancing will hold with suppressive fire alike. But the firing Soviet plt represents just the unsuppressed element of the enemy.

For the retreat a viable test would see 3 Soviet plts firing on that 1 moving GE plt while a single other GE plt and maybe a HMG42 provide covering fire. I'd bet that there is at least one plt of the Soviets firing unsuppressed. Thus I would expect more or less the same results as in MASTA_KSF's (sorry, just had to change those letters) test (Force ratio about 1:1.5).

On the advance, I'd put 2-3 MG34 and a mortar in support, have null unsuppressed enemies and expect less casualties (and suppression). (Force ratio about 1.5:1).

If there is a favorable local force ratio for the defenders, suppressive fire would reduce the effects on the moving plt, too. But I'd guess the overall tendency would still hold - the effect increases with the amount of unsuppressed fire until a threshold, when there is so much incoming that every squad gets annihilated or pinned fast and there is no difference between attack and retreat

Gruß

Joachim </font>

[Dook]

Those test results do not seem unreasonable to me. As other posters have said, you should suffer a morale penalty for retreating under fire even though casualties will be less.

Withdrawing is not a death wish - it causes lower casualties than advancing and your troops stand a better chance of making it to the cover-providing rubble behind you than with other commands headed away from the enemy. The price is a greater possibility of panic/rout.

[Andreas]

Originally posted by Scarhead:

If there is a favorable local force ratio for the defenders, suppressive fire would reduce the effects on the moving plt, too. But I'd guess the overall tendency would still hold - the effect is proportional to the amount of unsuppressed fire.

Gruß

Joachim

Joachim, I guess my opinion with this is quite simply summed up in saying 'So what?'. Is it unrealistic that a unit that gets shot in the rear while bailing from a position is more likely to rout than a unit attacking? In my (admittedly naive) opinion, that is consistent with the little that I have read (and can remember) about such occasions.

It appears to me that true 'grit' (groans from the audience) is easier to generate in soldiers when you are moving forward trying to get to grips with the enemy than when you leave the field to him. I would be interested in other opinions on this though, in particular based on Real Life experiences/tales.

[Joachim]

Originally posted by Andreas:

Joachim, I guess my opinion with this is quite simply summed up in saying 'So what?'. Is it unrealistic that a unit that gets shot in the rear while bailing from a position is more likely to rout than a unit attacking? In my (admittedly naive) opinion, that is consistent with the little that I have read (and can remember) about such occasions.

It appears to me that true 'grit' (groans from the audience) is easier to generate in soldiers when you are moving forward trying to get to grips with the enemy than when you leave the field to him. I would be interested in other opinions on this though, in particular based on Real Life experiences/tales.

Andreas,

now this is a totaly different question...

While I think that it helps a lot if you have suppressive fire in all case, there is an abundance of factors to compare the moral effects for attacking or retrating:

Attackers have some moral benefit, as they usually have local superiority. If they attack blind with bad leadership, I don't know if that effect holds.

OTOH if you just pin, the others will do the dirty work. Routing and running is only bad (in terms of incoming) if the others pin.

Standard troops retreating to a 2nd line should have the added benefit of knowing that they get probably pulled out of the line once they reach the 2nd line. Pinning will keep them in danger. Routing but running to safety would be ok.

Plus all the other factors - like penal btn.

But I have no experience, no abundance of data and thus all is just a theory.

My point is just that suppressive fire is not a solution to the problem - though it definitely helps.

Gruß

Joachim