Combat Model Needs Tweaking

[Pocus]

I think the main thing to tweak is subject 'a'.

By the way, don't the cohesion rules come into play here? Why isn't a loss of cohesion stopping armies from continuing to attack and suffering massive casualties?

Cohesion for now is used to check for routing, but until routed the unit will continue to fight, unless the general order the retreat. But retreat for now is decided on the compared strength of opponents, not on the loss ratio.

This is where the problem lie.

[Pocus]

The way the AI re-adjusts yr tactical posture from defense to offense the minute you cross into enemy controlled territory pretty much renders the whole issue of whether you were "reasonably prepared" null and void.

this is exagerated. You auto switch to offensive only if you have NO control in the region. As soon as you have even the smallest hold (5%+) then you can enter and remains in defensive.

For us this is logical. And last but not least, this is to prevent the exploit of advancing against the enemy while benefiting from the defensive terrain matrix.

[Pocus]

This would help me to get the saved games, so that I can tweak the parameters until the results are more along an historical line.

[jackfox]

Pocus, regarding the 178k casualties, I'll send you a zipped file of my game including the backups. I would have sent it sooner, but was away for the weekend.

[Wilhammer]

To add to the discussion;

http://www.civilwarhome.com/foxschapter3.htm

The figures consider 63 regiments across 23 battles, showing a loss range in MIA, KIA, WIA from 50 to 82 %, averaging at 60%. Bloody indeed.

-----------------------------

He considers further comparisons with similar European Armies;

"It is well to pause here, and consider what these figures mean; to think of what such extraordinary percentages imply. Perhaps their significance will be better understood when compared with some extraordinary loss in foreign wars; some well known instance which may serve as a standard of measurement. Take the charge of the Light Brigade at Balaklava. Its extraordinary loss has been made a familiar feature of heroic verse and story in every land, until the whole world has heard of the gallant Six Hundred and their ride into the Valley of Death. Now, as the Light Brigade accomplished nothing in this action,-- merely executed an order which was a blunder,-- it must be that it was the danger and its attendant loss which inspired the interest in that historic ride. What was the loss? The Light Brigade took 673 officers and men into that charge; they lost 113 killed and 134 wounded; total, 247, or 36.2 per cent.
The heaviest loss in the German Army during the Franco-Prussian war occurred in the Sixteenth Infantry (Third Westphalian), at Mars La Tour. Like all German regiments of the line it numbered 3,006 men. As this battle was the first in which it was engaged,-- occurring within a few days of the opening of the campaign,-- it carried 3,000 men into action. It lost 509 killed and mortally wounded, 619 wounded, and 365 missing; total, 1484, or 49.4 per cent. The Garde-Schutzen Battalion, 1,000 strong, lost at Metz, August 18th, 162 killed and mortally wounded, 294 wounded, and 5 missing; total, 461, or 46.1 per cent.
A comparison of these percentages with those of the Union regiments in certain battles just cited will give some idea of the desperate character of the fighting during the American Civil War."

[caranorn]

Don't forget those are the extreme cases.

[Stonewall]

Cohesion for now is used to check for routing, but until routed the unit will continue to fight, unless the general order the retreat. But retreat for now is decided on the compared strength of opponents, not on the loss ratio.

This is where the problem lie.

The ai calculates the odds based on combat power, correct? Is that combat power number (used for calculation of odds) filtered by the command modifier penalties that the ai armies are usually suffereing from?

And I'd still like to see a cap on fortification levels for units not in a fortress. Or perhaps have an opportunity during a battle for the fortification line to be pierced or a breakthrough to occur and then have the defending army lose the bonus as it reorganizes and counterattacks.

[gbs]

I think Pocus and his guys will eventually sort all this out. This and issues around the weather patterns and duplicate units are the most glaring as far as I can tell. With all of these things fixed, what a game this will be. My question is how much work will this take and how long (roughly) until we can expect a patch for these things? I am at turn 65 and just keep playing as is because I don't want to start a new game as is. Any ideas? Sorry for interupting an intersesting thread.

[veji1]

I think several things are important here, as Stonewall said, entrenchment levels are an issue, they should progress throughout the war : in 1861, if your not in a fortress you can it is capped at 2, in 1862 3 til late june, 4 til late december, 5 til late june 1863 and 6 til late september, 7 til late december and 8 til late march 1864. that way you can get to max entrenchment only from early 1864 onwards... If you really want to fortify strongly before, build a fort... basicly entrenchments become somewhat efficient in mid-1862.

Second about the cohesion/power/numbers that are used to determin when a army breaks off the fight : You should definitly use cohesion value there : when a force or elements gets under 40 cohesion it cannot carry offensive action anymore, under 20 it runs away from the fight...

If I am not mistaken regarding cohesion, experienced units will have better cohesion and will be willing to fight longer, and this way a unit can stop fighting while actually having suffered few losses but a massive morale drop : it means that after withdrawing the force that lost will actually be able to fight just a couple of turns later, avoiding the snowball of effect of a big battle with major losses meaning the force is useless and the ennemy can roam around freely...

[James D Burns]

The figures consider 63 regiments across 23 battles, showing a loss range in MIA, KIA, WIA from 50 to 82 %, averaging at 60%. Bloody indeed.

My first point agrees with caranorn, he only lists the most extreme cases, perhaps to try and create a more dramatic feel to his writings.

But even given the fact he uses the most extreme cases, nowhere does he list a regiment being entirely destroyed outright in any of the examples given. Yet in the game we see 50-100 regiments completely destroyed in the larger battles on a routine basis.

On average most of his extreme examples show about 40%-60% overall casualties (killed, wounded and missing). If you account for the fact captured men (missing?) are handled separately from killed and wounded in the game, I think it’s safe to say individual regiments should never suffer more than a maximum of about 50% casualties in any given combat unless they go into that fight at less than 25% full strength.

But casualties as high as 50% should be a very rare occurrence in game. He is listing a small percentage of the total regiments engaged in the civil war and the average overall is probably about 10%-20% if you factor in the thousands of regiments he does not list that fought in the war and took casualties.

Regiments need to be a lot more survivable in the game.

Jim

[LAVA]

Personally...

I've never encountered such a disparity of losses from battles during the game.

I wonder how often such totally lopsided results actually occur.

Ray (aka LAVA)

[saintsup]

My first point agrees with caranorn, he only lists the most extreme cases, perhaps to try and create a more dramatic feel to his writings.

But even given the fact he uses the most extreme cases, nowhere does he list a regiment being entirely destroyed outright in any of the examples given. Yet in the game we see 50-100 regiments completely destroyed in the larger battles on a routine basis.

On average most of his extreme examples show about 40%-60% overall casualties (killed, wounded and missing). If you account for the fact captured men (missing?) are handled separately from killed and wounded in the game, I think it’s safe to say individual regiments should never suffer more than a maximum of about 50% casualties in any given combat unless they go into that fight at less than 25% full strength.

But casualties as high as 50% should be a very rare occurrence in game. He is listing a small percentage of the total regiments engaged in the civil war and the average overall is probably about 10%-20% if you factor in the thousands of regiments he does not list that fought in the war and took casualties.

Regiments need to be a lot more survivable in the game.

Jim

AFAIK and if the system is the same as in BOA, the casualties in the game are not actual KIA, MIA or badly wounded. A part of the casualties is recycled in the manpower pool. So you cannot directly compare with the RL statistics.

Regiments are not 'completly destroyed'. They are more 'reforming' in some rear area.

[Wilhammer]

My post on this was to serve the discussion - and I agree that the author was a bit lopsided on his presentation.

Further, I also have seen strangely wild results due to combat - though not the one as extreme in this post. I have wiped out huge formations with comparatively few losses, but I must admit I had those enemy units in a hell of a pickle - outnumbered, out supplied, and with zones around them controlled enough to block retreats. These are usually multi-round combats.

----------------

LAVA wrote;

Personally...

I've never encountered such a disparity of losses from battles during the game.

I wonder how often such totally lopsided results actually occur.

Ray (aka LAVA)


:::::

The last question is a good question: also, can it be duplicated? What is the formula to create a killer force like this?

When we play wargames, we 'fight' far more battles with far more vairables than the Real War generals had - which SHOULD result in an occasional wierdness - but this sounds a little buggy, what we have here.

In relatively straight forward combats, the forces are medium sized (no more than a corps of two divisions) wherein a disparity in casualties was seemingly way off - and the scale was large.

I can see small scale disparity results (say 5-10% of your force gets rendered ineffective while 1/10th damage applies to the opponent) but when 50% or more of the force in a pitched battle occurs, and the enemy suffers a tenth of that, then something needs tweaking.

Personally, off the top of my head, I'd say out of a couple of hundred combat resolutions, about 3-4 times was I left thinking something was wrong with the combat results indicative of a bug in the software.

So, Pocus, how many examples do you want in your email?

We all know that Ageod will perfect this near perfection.

[tagwyn]

I agree!! Pocus and Phillipe will remedy the problem. I'll keep playing; nothing seen that is ridiculous yet. Can"t get sailors to board ships?! Tag

[James D Burns]

Regiments are not 'completly destroyed'. They are more 'reforming' in some rear area.

Even if the regiments dump some replacements into the pool, it is the wrong approach in my opinion. The fact is after Gettysburg Lee still had an army in the field large enough to pose a serious threat to the battered Union army and that allowed Lee to save his army.

In game terms he wouldn't have much of a standing army left because half or more of his regiments would have been destroyed outright by the combat engine and he would basically be easily pursued and destroyed in detail.

The civil war was a long bloody attrition war and armies need the ability to stand up to repeated punishing assaults if we hope to recreate it accurately.

I’m not saying regiments should be invincible, just that they need to become stunned in a battle where they suffer greater than 20% casualties in the interest of self preservation. As the regiments become smaller and smaller due to multiple engagements, soon they will begin to get destroyed outright when they start going into battle with just a few hundred men left.

But a full strength 700-1000 man regiment needs more staying power than just 1 or 2 battles. It would be interesting if someone could come up with a compiled list of all regiments actually destroyed in battle during the war.

Not those that surrendered, but those that actually ceased to exist due to battle casualties they suffered. My guess is only a very small percentage of regiments raised during the war actually were destroyed in combat.

Jim

[swang]

I don't see a problem with "destroying" elements. Because right now, it is way to cheap to refill an element, even if the element is down to 10%.

IRL, many of these elements would be amalgamated (Confederate side) or left as a small element (Union side). In fact, "refilling" elements seemed to not happen at all from what I've read. The problem with no "refilling" is that it makes the game a lot more micromanaging, I don't know that the mechanics would work at all.

[Heldenkaiser]

IRL, many of these elements would be amalgamated (Confederate side) or left as a small element (Union side). In fact, "refilling" elements seemed to not happen at all from what I've read.

Very rarely with the Federal army. Raising new regiments gave the home states more prestige (and officer commissions) than just sending replacements to old ones. Hence by mid-war, the early regiments were down to 300-400 men from the original 1,000.

Wisconsin apparantely sent new recruits as replacements into existing units. Hence in late war, Wisconsin regiments were often unusually large compared with their other state counterparts. Rumour has it that in the Overland Campaign (I think) one day a large body of men was marching by, someone inquired, "what brigade is this", and the answer was "the ..th Wisconsin infantry".

[caranorn]

As far as I recall the individual regiments had to recruit their replacements themselves. That means they had to send a captain or two with non-coms back to their homestate to try and raise a fresh company (which would not necessarily serve as such once it reached the regiment). But it was indeed much harder to recruit those replacements rather then create a new regiment (with regimental elections, bounties etc.), unless of course the regiment had gained some fame.

Note that I don't recall whether that system was applied on both sides or even in most states. But regiments could definitely get filled up, but rarely would one reach full strength or near full strength after the original mustering.

[Montbrun]

As far as I recall the individual regiments had to recruit their replacements themselves. That means they had to send a captain or two with non-coms back to their homestate to try and raise a fresh company (which would not necessarily serve as such once it reached the regiment). But it was indeed much harder to recruit those replacements rather then create a new regiment (with regimental elections, bounties etc.), unless of course the regiment had gained some fame.

Note that I don't recall whether that system was applied on both sides or even in most states. But regiments could definitely get filled up, but rarely would one reach full strength or near full strength after the original mustering.

This system was applied in the South. The Union had no such replacement system. The North preferred to de-mobilize regiments that were severely understrength, and then raise new ones. The South preferred to "recruit" on a regimental basis, and then consolidate regiments that were severely understrength. The "Stonewall" Brigade is an excellent example of this. By the time of Appomattox, it's constituent regiments had been consolidated, and the whole brigade was the size of a small regiment.

[denisonh]

i think the main issue is what Jim is referring to, that even a defeated Army had combat power enough to defend itself to prevent destruction.

Most battles (not skirmishes or small engagements) did not involve a complete destruction of an Army.

No reason for the combat model to really allow that outcome of an Army "disapperaing". I think the removal of non activated units being allowed to "attack" by enetering enemy territory would help tremendously prevent this imbalance in the larger battles, as that really seems to skew some the results.

[James D Burns]

Because right now, it is way to cheap to refill an element, even if the element is down to 10%.

Sorry but I have to disagree strongly on this point. In my AAR I raised about 2 million in taxes and spent about 1 million on my replacement pools and only had about 20 or so line infantry replacements after that due to the high cost of all the other types of replacements I had to purchase.

The reason it seems cheap to you right now is because you don’t have to refill those 50 or 100 destroyed regiments the engine creates in big battles. Were those regiments to stick around and require refilling, I guarantee you we will all be running out of replacements post haste.

Right now we all basically buy enough replacements to refill our starting armies and then we basically ignore the pools until a half year or so goes by and we need to buy a few more. Let the currently destroyed regiments stick around instead and you’ll see lots of under strength regiments going into battles because players will not be able to afford to refill all of them.

Right now it’s pretty easy to keep your regiments filled up to full strength because battle casualties are such an all or nothing affair.

My main problem with the current system is that a much smaller army like Lee’s was in late 1863 to the end of the war will disappear in one or two fights due to the outright destruction of the elements rather than just causing heavy casualties.

Let’s try and reflect the war and allow regiments to stay on the map even if they’re down to just 20% strength or so (how many regiments fought at Gettysburg with less than 200 men?). If the replacements turn out to be too plentiful, it’s a much easier task to increase the price of each replacement to get a good historical balance than it is to try and fight a 4 year war with no on map armies.

Jim

[el_Gato]

Right now we all basically buy enough replacements to refill our starting armies and then we basically ignore the pools until a half year or so goes by and we need to buy a few more. Let the currently destroyed regiments stick around instead and you’ll see lots of under strength regiments going into battles because players will not be able to afford to refill all of them.

Which, IMHO, is how it should be. If not nescessarily for the reason that there isn't enough manpower to refill them (for the CSA), than for the logistical difficulty of doing so (USA).

As you point out, there were several regiments at Gettysburg the size of companies --- and these were in fresh army formations. The replacement situation didn't get any better from there.

[James D Burns]

I think the removal of non activated units being allowed to "attack" by enetering enemy territory would help tremendously prevent this imbalance in the larger battles, as that really seems to skew some the results.

I wonder what ever happened to the idea of not allowing armies to enter enemy controlled territory at all unless lead by an activated leader? To me this would go a long way to recreating the inaction of McClellan throughout 1862.

As things stand now it's easy to get an early offensive going even if you fail every single strategic roll for an entire year. Granted you take a hit to your combat abilities, but in my opinion offensive moves into enemy controlled territory simply shouldn't be allowed unless a stacks leader makes his roll first.

And stacks without leaders should never be allowed to set an offensive or assault posture period, unless in friendly controlled territory (this would allow cavalry to chase raiders).

Jim

[veji1]

agree with Jim there. I would like to see the following changes :
- hardly any destruction of elements in battle : they are knocked off the fight when their losses are too high, that's it. then we need to make sure that the model handles the amalgamation of regiments say under 25% strength properly. That way we should find ourselves with plenty of regiments at 40-50% strenght further along the war.
- No invading ennemy territory when you are inactivated... this is a major rule, frustrating yes, but otherwise the whole activation/Strat stat is just a combat penalty and not a major feature anymore...
- No offensive posture for non-led stacks except in own territory ( say a 75% controlled or more ).
- More leaders dying.

[Pocus]

there will be (when time allows) a new set of optional rules that you can choose if you want to, that will apply more stringents conditions on inactive leaders and leaderless units.

But don't rule out too fast the combat penalty a general get if he enters an hostile territory while inactive. This is a tremendous penalty, which can lead to a complete slaughter.

I react also on the army leaders not impacting enough the corps commanders for some. Don't forget that a bad army commander have a tendency to give malus to corps commanders, to all ratings. Thus Mc Clellan will reduce by 1-2 the strat rating of all subordinates very often.

[el_Gato]

So, then, I'm attacking with a Corps (of two Divisions), across a river, into Lexington, KY. The unit defending the city retreats before I get there, but I run into a Union Independant Command in open terrain, outside the city:



The Union force retreats back to the Ohio, and Bonham's Corps ends up in control of the region, and the city. CSA casualties: 875 / Union casualties: 2104.

Why is this a Confederate defeat?

[veji1]

I think we need to slooooooooooowww down the game : Like all games it tends to go way faster than history :
- tighter activation rules, notably for leaderless units would slow the game consistently.
- Stronger losses of cohesion for many operations ( movements, change of command, etc...) would force the player to slow down its operations to regroup :you see a couple of good divs go from 350 to 200 just for marching around, makes you pause...

[Hidde]

veji1, you are so right!
Every wargame I've come across suffer from this. More planning,fewer battles,
every action to have consequences is the way to go.
Not everyones cup of tea, but that's how I would like things to be.

[caranorn]

I'd just like to note that these leaderless units were the ones to conduct offensive operations out West during the early phases of the Civil war. That is to say that many offensives were launched with single brigades, without a superior leader around.

In short, I would not remove those abilities in the game.

[Jagger]

One option to reduce casualties could involve increasing the terrain impact on combat.

I finally looked at the terrain spreadsheet and was surprised at the minimal impact of most terrain on combat. By increasing defensive aspects of terrain, overall casualties would be reduced and tactically, defensive positions outside of towns/forts would make more sense.