Combat Model Needs Tweaking

[caranorn]

Looking at the screenshot once more I notice both sides are shown as in offensive stance. So I'm starting to feel really confused.

[caranorn]

Looking over a battle in my current game I think moving into enemy controlled territory automatically places your forces into offensive stance, at least for a day or two. I wonder whether this is intended as it seems to be contrary to activation...

In my case it was a dumb move involving just a single division on a "reconnaissance en force". Good old Kirby Smith was not supposed to attack those 4 divisions of the Union Army of the Tennessee entrenched at Nashville. I'm happy he even got two regiments and his HQs (114 men total) out of that mess.

[denisonh]

I just experienced a pretty costly battle. Just thought this might add to the discussion.

Truthfully, I was expecting something closer to yours than what I experienced. The CSA actually took casualties in your battle.

Edit: Upon looking further, you lost a $h*tload more than I did. I think it does cause some concern that once the battle is joined, that you can not only lose, but dissappear as an organization. Unless you are surrounded, that is just not a likely outcome IMO.

[type7]

I just experienced a pretty costly battle. Just thought this might add to the discussion.

178,000 in a single day's battle?!!??

[Chris0827]

What were the casualties on both sides during Pickett's Charge at Gettysburg? Could they really have hit any Union forces from their position?

I've seen estimates of between 1500 and 2300 union casualties during Pickett's charge

[jackfox]

178,000 in a single day's battle?!!??

That was just McDowell. Two other Union generals attacked days later in the same turn suffering 28,000 and 32,000. Lee took 1,200-3,000 in each of these engagements.

The 29,000 Lee took in the first battle were what I expected. What I didn't expect was that McDowell would continue those costly attacks resulting in 178,000 casualties in a single day.

[johnnycai]

One thing to keep in mind is that the entire Corps "disappeared".

An general forced to attack with a posture set to defense should most certainly not prosecute that vigorous of an attack.

Dieing to the last man is a very unusual case (particualry in an attack on enemy held territory), particualy for an entire Corps. Even Pickett had more left the Hunter did after this battle.

I recently had this happen in my PBEM game, Oct '61, I had the AofP minus Hunters Corps on defensive posture around Manassas facing off against Jackson's corps, Hunter's Corps also on defensive arrives on Day14 and triggers a battle against Jackson's Corps well dug in but am not sure what the entrenchment was. I also heavily outnumbered his artillery and was hoping to do better than the 24K to 10K in losses I suffered considering I didnt set offensive but Hunter's arrival triggered the Union attack on day14, the rest of the AofP also joined the attack.
I also agree an artillery advantage as presented here should aid in the suppression of the fortified artillery especially in open terrain.

[James D Burns]

I've seen estimates of between 1500 and 2300 union casualties during Pickett's charge

Here's a breakdown for the South:

http://gburginfo.brinkster.net/ChargeCasualties.htm

5,675 killed and wounded, 792 taken prisoner. About 65% of the 10,500 attacking.

Jim

Edit: This site gives a higher estimate of troops who made the attack (12,500-15,000), but total casualty figures are about the same. Also has 2 awesome photo's of the field from both the north and south's prespective.

http://www.brotherswar.com/Gettysburg-3c.htm

[Pocus]

Looking over a battle in my current game I think moving into enemy controlled territory automatically places your forces into offensive stance, at least for a day or two. I wonder whether this is intended as it seems to be contrary to activation...

In my case it was a dumb move involving just a single division on a "reconnaissance en force". Good old Kirby Smith was not supposed to attack those 4 divisions of the Union Army of the Tennessee entrenched at Nashville. I'm happy he even got two regiments and his HQs (114 men total) out of that mess.

yes, if you move into a region which has less than 5% control, you are auto switched to offensive, even if not activated. But if this case, you are received a full -35% combat efficiency from fighting in offensive and unactivated, so don't expect prowess, even with numbers on your side.

[Dan]

178,000 in a single day's battle?!!??

Looks like there were a lot of POW's for the Union. Did the Union forces have no place to retreat? The Union was attacking an equal force with great generals in a well defended position. What caused the AI to not break off at some point before 10,000 troops were lost that day?

The even bigger question: What is Lee's army doing in Washington in '63 with almost level 4 entrenchments?

[jackfox]

Looks like there were a lot of POW's for the Union. Did the Union forces have no place to retreat? The Union was attacking an equal force with great generals in a well defended position. What caused the AI to not break off at some point before 10,000 troops were lost that day?

The even bigger question: What is Lee's army doing in Washington in '63 with almost level 4 entrenchments?

McDowell attacked from Annapolis, which was still under his control at the end of the turn. So he did have some retreat options. I had just taken Washington, but was playing the game out a little bit further.

[el_Gato]

I think the actual confederate losses are OK, assuming the Union broke before reaching melee... but since the union took so many casualties and kept advancing, by the sheer force of inertia it should have reached the trenches and then losses in the 2-3000 for the rebs would make more sense...

That said I think the main issue here is frontage : An entrench army is a way deeper body than a field army, this means that far less troops are on the first line... the same applies for an army attacking trenches... So I think the main issue here is the fact that the Jackson army being entrenched its frontage should have been shortened a lot, and therefore the Union's as well... Which means that 6 divisions attacking means actually 3 attacking in the first wave and 3 in the second, which means that under these conditions the second wave should never have actually made it to the fight, because of the stream of survivers rooting from the first 3 Divs...

So I say the devs should try to reduce the frontage of an army the more it is entrenched...

Exactly what I was thinking.

AACW's combat resolution routines seem to be suffering from a common problem in war-gaming: It doesn't take human nature into account. Soldiers aren't robots, who just keep attacking until wiped out to the last man. They break and run long before reaching the 100% casualty mark. Actions like Pickett's Charge, or the Charge of the Light Brigade, at 60% casualty rates, are outliers. A 20 - 30% casualty rate is usually enough to get most attacks called off --- assuming the units involved haven't already routed on their own.

Jackfox's example of 178,000 casualties in a single battle is just plain ridiculous. That's an entire army wiped out in a few days! As the attacker in a seige, with plenty of terrain to retreat to, there's no way those kind of losses would be incurred. Or tolerated.

[rasnell]

Here's a breakdown for the South:

http://gburginfo.brinkster.net/ChargeCasualties.htm

5,675 killed and wounded, 792 taken prisoner. About 65% of the 10,500 attacking.

Jim

Edit: This site gives a higher estimate of troops who made the attack (12,500-15,000), but total casualty figures are about the same. Also has 2 awesome photo's of the field from both the north and south's prespective.

http://www.brotherswar.com/Gettysburg-3c.htm

One of the most dramatic lines of the movie, "Gettysburg," derived from the book, "Killer Angels"...

General Pickett, devastated by what he had just witnessed sullenly replied to his commander, "General Lee, I have no Division now."

[tagwyn]

I hesitate to point out that Gen. Grant lost almost the entire beginning strength of the AoP in the Wilderness Campaign of 1864 in driving Lee's ANV back to its defenses at Petersburg. This was the "I will continue on this front for the entire summer" or some such quote from Grant. Agreed, this was not a single battle but was a period of only a few months. Rebel losses were not negligible and were irreplaceable. Pocus please keep us briefed on any proposed changes to the combat results tables. Tagwyn

[jackfox]

I hesitate to point out that Gen. Grant lost almost the entire beginning strength of the AoP in the Wilderness Campaign of 1864 in driving Lee's ANV back to its defenses at Petersburg. This was the "I will continue on this front for the entire summer" or some such quote from Grant. Agreed, this was not a single battle but was a period of only a few months. Rebel losses were not negligible and were irreplaceable. Pocus please keep us briefed on any proposed changes to the combat results tables. Tagwyn

I'm afraid that's not entirely true. Grant started with somewhere around 105,000 and received reinforcements throughout the campaign. The total number of casualties from Wilderness, Spotsylvania, North Anna, Bethesda Church, and Cold Harbor came to about 50,000. Even after Cold Harbor, Grant still had in the neighborhood of 90,000 men.

Though considered by many to be a butcher during this campaign, Grant never came close to destroying his entire force the way we're seeing in some of these examples from the game.

[Childress]

Soldiers aren't robots, who just keep attacking until wiped out to the last man. They break and run long before reaching the 100% casualty mark.

A bit OT: notice that a force carelessly left out of supply will eventually dissolve to the last man. But a real army, unless besieged or surrounded a la Vicksburg, doesn't remain in situ and starve to death. This is where a 'TacAI', if one existed, would awaken and march the force back into a supplied region. But now without player input they die like, well, robots.

[Mike]

Here's a web site with battle losses for both sides for comparrisons apy: apy:

http://www.americancivilwar.com/cwstats.html

[Primasprit]

A bit OT: notice that a force carelessly left out of supply will eventually dissolve to the last man. But a real army, unless besieged or surrounded a la Vicksburg, doesn't remain in situ and starve to death. This is where a 'TacAI', if one existed, would awaken and march the force back into a supplied region. But now without player input they die like, well, robots.

The force will dissolve, right, but the soldiers don't die but rather desert. Their general order them to stay in that region without any supply and the soldiers will do it for a while... but not forever...
... sooner or later one of the soldiers will spread the rumour that their general is is nutty as a fruitcake...
... so they decide to go home for a while... waiting for the next call for volunteers...

[Hidde]

This feels more and more like a big issue. As I've written somewhere else, stuff like this have a big negativ impact on the immersion factor. If it goes bad enough it also affect the outcome of games.
I haven't played that much, but my impression is that often there is a string of battles during a turn. Do this loopsided results occur as often in the first battle or is it with battle two or three it happens? If that's the case maybe there should be some mechanism to prevent a big battle to be followed by a new one a couple of days later. Lets say that loss of cohession gave a severe impact on the will of the generals involved, to fight again for the same turn, at least.
Casualtie numbers of 178,000(+40,000 prisoners) shoulden't be possible, no matter what. That's my firm opinion.

[von Beanie]

There was a situation just like this in BOA that I complained about a few weeks ago in the Matrixgames forum. IT DOES NOT MATTER THAT YOU SET YOUR FORCE ON DEFEND if you advanced into an enemy held area. The game engine automatically resets it to full attack, and then proceeds to bash itself to pieces.

After a lengthy exchange in that forum, I was basically told that there is no such thing as a cautious advance into enemy territory (and I'm pretty sure that is what you intended by setting your force on defend, because that's what I was intending to do).

This is an artifact of the game system, and it takes a while to learn about it. Once you are aware that this is what the game engine does, you can plan your moves accordingly. The critical thing to notice is who controls the area.

Personally, I'd much prefer a cautious advance option, or a subroutine in the combat system that breaks the units when they suffer a set percentage of casualties, but don't hold your breath waiting for it.

[pasternakski]

There was a situation just like this in BOA that I complained about a few weeks ago in the Matrixgames forum.

Well, I do not see this as being "just like" the other, but I have no wish to fire all that up again here. I can only say that I have agreed several times with posters who see a problem with numbers of the loser's casualties in AACW combat. That said, I also believe that most of us are still neophytes prone to doing very silly things that can lead to horrendous losses (and the AI is currently in the same boat with us, but at least "she" has Pocus on her side). Just because McDowell, Banks, McClellan, or one of the other "questionably competent" guys didn't lose a whole army through silliness doesn't mean WE aren't capable of doing so ... it kind of acts as a "cautionary" tale for me in not judging the hesitancy of commanders in this war too harshly.

At least to some IT DOES NOT MATTER THAT YOU SET YOUR FORCE ON DEFEND if you advanced into an enemy held area. The game engine automatically resets it to full attack, and then proceeds to bash itself to pieces.

I tend to agree with your implicit criticism of how the game allows your forces to be demolished in these circumstances (others have stated much the same complaint), but I also look at the intended dynamic built into the game mechanics, and see some sense to it.

If you form up your boys on the road and march into territory you know full well to be controlled by your enemy with forces in significant strength, you are going to get your head handed to you on a dinner plate. Just because your troops can panic, throw down their weapons, and rout does not mean that you are not going to suffer a most-deserved whuppin'.

I guess I don't see how you can "cautiously advance" in these circumstances. "We sees yuh, we's a-waitin' on yuh, and we' ankshus fer y'all to get heah, 'cause we has pree-pared a pow'ful warm welcome."

After a lengthy exchange in that forum, I was basically told that there is no such thing as a cautious advance into enemy territory (and I'm pretty sure that is what you intended by setting your force on defend, because that's what I was intending to do).

Well, it all got a little pedantic and overwrought, but that's the nature of the place whenever anyone wants to discuss anything seriously. I am much happier here and have no plans to return to "the other place." I got a little tired of "being told," too.

Anyway. Reconnaissance-in-force is one thing. Assault is another. I just don't see the sense in pushing a wad of foot troops into enemy territory and expecting to get out with only scrapes and bruises if "what they brung weren't 'nuff to empty the weasels out the henhouse." It's one thing if what's there is just the "Caspar Milquetoast Volunteer Militia," but what if you are facing an entire army, well situated and fortified, skirmishers and cavalry patrols active, in short, in full military possession of the place?

This is an artifact of the game system, and it takes a while to learn about it. Once you are aware that this is what the game engine does, you can plan your moves accordingly. The critical thing to notice is who controls the area.

Well said.

Personally, I'd much prefer a cautious advance option, or a subroutine in the combat system that breaks the units when they suffer a set percentage of casualties,

But, see, why would you send entire divisions, corps, or even armies into situations where you are firmly convinced they will be trounced? I guess I am not understanding what you think the game should do to help you out when you do that.

Now, if you have some presence in the area already that prevents the auto-shift to attack, you already get what you want - don't you?

but don't hold your breath waiting for it.

Sorry, I don't think I want it, because I don't think any of that is necessary. The huge losses may prompt some change, but, in light of the circumstances I have seen described in which they have occurred (and I have not seen any so egregious, although I have raised a Spock-like eyebrow a couple of times), I would hate to see anything too extreme done too soon, requiring another readjustment at a later time after we have become better players and the results appear to be skewed too far in the other direction.

[dinsdale]

If you form up your boys on the road and march into territory you know full well to be controlled by your enemy with forces in significant strength, you are going to get your head handed to you on a dinner plate. Just because your troops can panic, throw down their weapons, and rout does not mean that you are not going to suffer a most-deserved whuppin'.

I somewhat disagree with this. While it wasn't impossible to wander ass-backwards into a battle without having a clue what the enemy disposition was (that's right Davout, I'm speaking to you. Hold that baton straight when you're being addressed by an omnipotent commander!) or to believe that you're pursuing a beaten enemy only to have them rope-a-dope you and take your flank in 40 minutes (there there Marmont, it's all right, Wellington kicked everyone's ass) such instances should be rare.

They should, IMHO, depend on relative cavalry strength, a commander's disposition and quite a healthy dose of dumb luck.

However, I think it's a difficult problem for WEGO with two week turns to solve. If posture is overriden by a combination of factors out of a player's control, there will be a flood of hand wringing when Ney's 40,000 men halt, wonder if there's a couple of regiments or army in front of them and decide to dig into the fromage instead of the 2000 men baring the way to Brussels.

The other alternative is a 400 member matrix of posture orders, though I'd say that would become rather unwieldy: "Did I remember to set the cavalry-should-take-the-heights before we watch 30,000 Russians dig in, or am I still in suicidal Ney mode on odd numbered days when the weather is poor?"

Personally, I'd like to see the likelihood of battle, and the posture entering a battle be dependent on leadership, scouting, cavalry and luck. But I'd also like to see the ability for posture to change. After all, a good General might realise that he's not going to be foraging boots today, and decides not send Picket on Operation Certain Death......or not.

[pasternakski]

Personally, I'd like to see the likelihood of battle, and the posture entering a battle be dependent on leadership, scouting, cavalry and luck. But I'd also like to see the ability for posture to change. After all, a good General might realise that he's not going to be foraging boots today, and decides not send Picket on Operation Certain Death......or not.

Well, and I think you are right in all particulars (and I would have brought along Soult no matter how much of a shambles Paris might have been when I got back).

The only "kicker" for me is the business of dispatching large forces (particularly under generals of limited experience and questionable ability) into circumstances where the outcome is highly likely to be disaster (you do, after all, get pretty good intelligence on immediately adjoining areas, and you always know your percentage of control; woe to the general who arrogates to himself - without thorough reconnaissance - sufficient knowledge of enemy dispositions in areas farther away to commit his entire command to marching down that road to uncertainty ("Please, Mr. Custer, I don't wanna go..."). Once you have set a sizable number of Civil War-era troops in motion, it's the devil's own task trying to get them to do something else. Now, smaller forces, particularly those composed of cavalry or light, irregular troops are a different story, and I think the game pretty accurately allows you to use them for reconnaissance and the other ad hoc purposes to which they were put historically...

As a side note, I have seen players in several AARs depending heavily on - and succeeding with - the offensive generalship of some guys who weenied out completely in real life. Maybe it's too easy, and the leadership penalties and restrictions ought to be re-examined ... maybe ... just a thought.

[el_Gato]

A bit OT: notice that a force carelessly left out of supply will eventually dissolve to the last man. But a real army, unless besieged or surrounded a la Vicksburg, doesn't remain in situ and starve to death. This is where a 'TacAI', if one existed, would awaken and march the force back into a supplied region. But now without player input they die like, well, robots.

Yah, tell me about it.

I just got this message: "Sioux Indians has been destroyed due to lack of supply in Region Great Plains". Uh-huh. Sioux out of supply in the Great Plains. Sure.

Funny thing is, they'd been locked in place from the moment they arrived, yet were besieging the US fort there. They wiped out the US garrison the same turn they starved to death!

[Pocus]

we will be tweaking the numbers, but I would like to get the saved game (with the backups so that I can reenact the battle) of both reports, the one with the 20.000 to 1000 casualties and the one with the 178k.

Things that can change are:

a) willingness to stop battle when ratio of losses is too high
b) frontage when entrenched
c) propensity of stopping a fight for a regiment when taking losses
d) impact of trenches on defense
e) impact of trenches as an artillery bonus
f) routing formula

and some others

[pasternakski]

Yah, tell me about it.

I just got this message: "Sioux Indians has been destroyed due to lack of supply in Region Great Plains". Uh-huh. Sioux out of supply in the Great Plains. Sure.

Funny thing is, they'd been locked in place from the moment they arrived, yet were besieging the US fort there. They wiped out the US garrison the same turn they starved to death!

I know, I know, but ...

this only means that this war party or raiding band has ceased to exist as a fighting force due to logistical problems attendant on their having sat outside Fort Courage for so long waiting for the series to be renewed, or maybe Chief Wild Eagle and Crazy Cat decided it was time for a Turkish bath, some canasta, and a spritz.

Besides, it took Kevin Costner doing his "hey, looky me, I got horns" dance to tell the stupid injuns that buffalo were nearby...

Anyway. Sustaining a couple of hundred braves for extended war operations took some doing, eventually beyond the capabilities of their villages of origin. I'd just chalk it up to that...

[caranorn]

Well, and I think you are right in all particulars (and I would have brought along Soult no matter how much of a shambles Paris might have been when I got back).

You obviously meant give Soult the left and taken Davout along regardless of...

[caranorn]

Pocus, I think the trenches themselves are okay considering how long it takes to achieve the level of entrenchment in the first example. The often mentioned case of Pickett's charge involves breastworks thrown up over night or possibly over two days (seems I just forgot on which day the charge took place, I believe late on the second, so just overnight and part of the day), so non comparable with our example here.

Though one thing you might want to look at is how you can add units to an already entrenched force without risk to lower said entrenchments. this is particularly flagrant when you have a Militia entrench and later add an entire corps to it and have it adopt the entire trenches... I fear the AI might not know how to do those things.

[James D Burns]

While surfing the net I came across this wonderful site. Talk about a wealth of information, some of the individual regimental links list the name of every man that served in the unit.

Ok AGEOD now we want extreme detail, you must list the name of every soldier in the unit and add/remove them according to their individual historical timelines. apy:

Here’s the link:

http://www.geocities.com/Area51/Lair/3680/cw/cw.html

Jim

[PDF]

Pocus, I think the trenches themselves are okay considering how long it takes to achieve the level of entrenchment in the first example. The often mentioned case of Pickett's charge involves breastworks thrown up over night or possibly over two days (seems I just forgot on which day the charge took place, I believe late on the second, so just overnight and part of the day), so non comparable with our example here.

Though one thing you might want to look at is how you can add units to an already entrenched force without risk to lower said entrenchments. this is particularly flagrant when you have a Militia entrench and later add an entire corps to it and have it adopt the entire trenches... I fear the AI might not know how to do those things.

I don't think so... In the CW time trenches were not WW1 trenches all along the front, so benefited the defenders only if the attacker attacked head-on. Strategically it was stupid to build big field fortifications as the enemy could just pass by ! AFAIK except when defending cities or forts the entrenchments were quite "light".
I would rather have a lower limit on entrenchment levels (or on their effects) to prevent abuse. Plus level should be unit-based and not force-based as you point out.
It seems also that the other problem is that any advancing force is "offensive" and break combat much too late if casualties mount. Granted, Pickett's charge cost 60+% of the division, but it was a last-chance attack and nowhere did generals send whole corps in such charges !