AGEOD Forums

Okay, I had Grant set up with the whole AotP - just over 100.000 men.
Good corps commanders, lots of high quality units, good mix of infantry and artillery. Not that much cavalry, though.
The CSA attacks. They outnumber me with about 25.000 men.
3 battles happen during the turn.
Grant wins the first one, though losses are about equal, 30.000 for me, 40.000 for the CSA. A bit over the top, I thought, for one battle, but hey, at least I won.
2nd battle that turn and my jaw hits the floor. I loose. Big time.
50.000 men lost and the CSA looses about half that number.
But Grant stays put and a third battle ensues which I loose horribly - another 25.000 men dead with neglible losses for the CSA.
The AotP wiped out in one turn.
Washington was taken the next turn.
War is lost. I assume.

So, I lost 100.000 men, the whole AotP in one turn.
Losses in general are too high in this game, but something else must be amiss here... Me, in other words.
Grant was at offensive posture - to make sure the CSA didn't slip by to hit Washington. Was this a mistake?
How can I avoid such catastrophic losses in the future? I.e. make Grant retreat in the second battle instead of staying and taking 50.000 (!!!) casualties and 25.000 more the next battle...
What attack/defend settings should I use?
Last question: If a loosing army has to retreat across a river - will this increase its losses or will it perhaps even prevent the withdrawal and force it to fight another battle that turn? - if so, can other terrain/weather conditions have this effect too?

Hope this isn't too many questions - just trying to get better at a hard game.

briny_norman wrote:Losses in general are too high in this game...

Yesterday, in a Kentucky scenario game, Buell's force of ~50,000 fought Hardee's force of ~13,000.

After one round of battle, reported losses were:

20,000 (out of 50,000), or 40% for Buell
13,000 (out of 13,000), or 100% for Hardee

In other words:

Buell's force suffered losses much greater than any historically (the bloodiest battles of the Civil War topped out at around 25%; see the earlier discussion about this elsewhere in this forum).

And Hardee's forces were wiped out entirely!

The punch line?

The game engine declared this a Confederate victory!

Game over, I couldn't bring myself to play any more after that point.

My memory is getting hazy (there having been so much patching and discussion about patching), but wasn't this whole excessive battle loss issue just dropped and forgotten?

There has been discusion on this all around several months. However, still no conclusion if this kind of Huge battle loses were POSIBLE or not. Some of the ideas mine and from others.

1.- ATTACKER seems to have TOO MUCH advantage (nearly) ALWAYS. Causes:
- Attacker always is ACTIVATED. Else, just he does not advance. Unactivated defensive leaders should have NO PENALTY at ALL, seems rules as they are designed give nearly no penalty.
-Attacker attacks with SUPERIORITY. Once 1,25 - 1 Ratio is achieved, seems nearly imposible to lose desastrously a battle, unless very rare situations (i.e. disembark)...

2.- Losses are too high, some reasons...
- Battles & Frontage seems to function nearly OK in a single UNIT vs UNIT combat. In this case, defensive power seems to function OK for me.

But if SEVERAL UNITS overwhelm a single unit, then the system penalizes too much defensive units, and the "hypothetical" advantage of the defense (even fully trenched & so) disappears completely.

- DAMAGE HIT RATIO versus COHESION HIT is too high. Seems damage Ratios should be lowered to give best results, mainly to allow for more chances to retreat.

- Delay into battles. There are option to increase DELAY before starting a battle. As default, seems to be too few delay, allowing for 3 battles in 2 weeks. Perhaps DEFAULT should be HIGH DELAY, and options should be done to reduce it.

Check CLOVIS MOD, he has tried to solve most of the questions in the game (production, conscription, losses), but is difficult to put into Vanilla scenarios

Coregonas wrote:There has been discusion on this all around several months. However, still no conclusion if this kind of Huge battle loses were POSIBLE or not...

Possible is one thing, probable is another. I see these sorts of excessively high (compared to history) battle losses regularly.

Except for a few cases of garrison surrender (Vicksburg, Port Hudson, Fort Donelson), arguably also Nashville (where Hood basically destroyed what was left of his army), the real war never saw the high loss rates we see (or at least I do) routinely in AACW open field battles.

Check CLOVIS MOD, he has tried to solve most of the questions in the game (production, conscription, losses), but is difficult to put into Vanilla scenarios

Yes, now I remember: Hearing that his mod had solved this problem, I was toying with the idea of applying it, but the install instructions were so complicated and daunting that I chickened out.

What if some of Clovis' best ideas along these lines were to go "official"?

(Why is it that many times I feel I'm more participating in an extended beta test than playing an actual game? )

Gray is trying to pick & use some of his ideas. Seems he uses ideas from some of the masters of the athena here...

Just the problem is Clovis uses some kind of "FAST" mod programming mode, that is GOOD, allows for FAST results...

but not useful for quickly turning into Vanilla.

AFAIK, Gray is starting to review production figures

Here the results of 2 early battles:





To lower losses, I made only a few changes:

- lowered the cbtHitCoef to 105 in the Combat setting file, so units will have less chances to hit enemy. I suspect default value to be based on the wrong belief mass introduction of rifled firearms to have substancially raised the firepower effect of small arms. Some recent studies are pointing out both low training level and unwillingness to adapt tactics to new realities kept the usual range of fire very close to smoothbore period. So the AACW combat model is certainly too much lenient with long range hit chance.

- lowered the cohesion value of each unit. A low value gives more rout chances. Inexperienced units will leave battles more quickly, shorter battles will give less losses for both sides

- lowered the maximum entrenchment levels at start.

On the contrary, an experienced unit in the last years will get more chesion and protection and, like vanilla settings, more firepower. So the losses should be higher.

As I mentioned in another thread - I wiped out AoP by surrounding it and allowing it nowhere to retreat to. I lost as CSA almost 70,000 but I wiped AoP for over 120,000 in 3 battles. I could not turn it to my advantage though I did not have a single division, brigade or element that had not suffered casualties. All of my Generals went inactive after those battles.

berto wrote:What if some of Clovis' best ideas along these lines were to go "official"?

I urged this a little while ago in another thread and think its really would be a sensible way to go. Seems that the community can get all fired up about how fast a corp or army can move over a particular period but be so unconcerned with totally unrealistic battle casualties.

Mind you its still an extremely enjoyable game, even with daft battle losses. apy:

I have a little theory about this...
From my experience, it seem the high casualty results were rarer and of less severity with patches 1.09 and before. They happened a lot on the early patches but latter there was some tweaking done and the results got much better. But since 1.10 patch it seem we are again seeing this kind of results frequently.

My theory about this problem is that this can be caused by the movement tweaks introduced on the 1.10 patch.
Besides being faster, movement now is MUCH LESS cohesion costly and for what i see on my games units and armies are most of the times at 100% cohesion even when moving around (except on very prolonged marches, bad weather/terrain, forced march...)
When two of this high cohesion armies engage in combat the losses are much higher as they have little desire to withdraw or rout (high cohesion).

Before the movement changes, even moving one or two zones on good terrain could lower the units cohesion under 50% max and as result the armies stopped killing each other sooner by lack of cohesion in one or both armies.

Now, i like a lot the new movement rates and thing its OK that normal movement does not cause too much cohesion loss. Seems more historic to me.
But i think it would be advisable to tweak the combat system down to adjust it to the new situation: increasing retreat chances, cohesion losses or doing some of the adjustments Clovis proposes.
Whatever is easier to do.
IMHO the important thing is that the intended effect is achieved: battles considerably less bloody.

Regards!

deleted

deleted

A little dissenting voice here...
I am currently playing a vanilla CSA April campaign started in 1.10c and patched to 1.10d ; nominal options, with the exceptions of +2 FOW advantage for AI and historical attrition for both sides. It is now just December 1861, but I have had several battles in Virginia, and also some engagements in Kentucky and Missouri. Particularly in Virginia, the engagements have been several and between large numbers of men, almost allways with me in the defensive (last one, on top of my head, was between around 60.000 entrenched CSA troops, with less than 200 guns, led by Beauregard, Jackson & co vs more than 100.000 AI USA soldiers with 300+ guns, led by McDowell &co). I won, with around 10-15% losses for the attacker and slightly less than 10% in my army. I have had NOT ONE battle with more than 20% losses for either side, IIRC, with the exception of one or two small scale engagements in which Athena's cavalry raiders were cornered and finally (after 2 or 3 battles) were wiped out/surrendered.
I am absolutely happy with current ratio of battle losses, at least in the game that I am playing.
Saves available if needed, of course. ( Of course I presume that berto already sent his saves to Pocus for anaysis... )

Regards

deleted

For the record my game is with a freshed (re) install patched to 1.10c and then to 1.10d.

Hi

From my experience, this kind of very heavy battles don't happen always, just form time to time. Most of the time results are OK likes the ones posted by Franciscus.
I think some factors that intervene on the huge casualty rates are:
- Good leaders: Grant, Lee... and corps leaders with bonus from them can turn a battle into a meat grinder. Historic to a degree, but with results much too extreme IMHO.
- Red or even orange Rules of Engagement.
- Late war unit models (with too high cohesion maybe?).
- Repeated battles on the same turn. Not sure how this work but they sometimes happen after a previous battle so heavy that no army should be fit to fight again. Check the first post example. His second battle is excessive. The third... Grant should have retreated after the second one (he had 80% casualties!!)

The problem is we don't exactly know the details of Birny norman and berto examples: which ROE they used, did they have somewhere to retreat (blocked rivers behind them, no Military control on surrounding areas...)? were the battles on open field or inside structures (that would explain the 100% losses)...
Regards

I also have some theories...

Although from time to time we are bound to have loopsided results, most of the time I feel that that may occur due to player's choices (or lack of ) - namely:
- the ROE: the default - orange might be sometimes excessive; red is extremely dangerous; maybe these options, although to me very welcome are sometimes not "remembered" by most players.
- The terrain: nowhere to retreat (rivers, military control, etc); factors that generals in "real" ACW usually took in account but players might not...

Regards

(PS to berto:

berto wrote:(Why is it that many times I feel I'm more participating in an extended beta test than playing an actual game? )

Well, do not forget that your "official" title is "AACW tester" )

Hi there, I have been playing a 1861 full campaign Pbem game as the Confederates, we are in 1863 and the game is still about 50/50 as either of us could win.

All of the battles we have had have been normal except for the one Ive described below.....


I had taken St Louis, the Union player obviously didn't like that very much and hastily attacked my Army of the west which was occupying the city, the CSA army had about 80 canon of various size and type and the Union army had no artillery at all, my troops opened fire at a range of 7 and the union army opened fire at a range of 4 as they had muskets only, now the loses where about 700 on my side and 16,000 on the Union side, on the battle report most of the Union army had routed I think before they got in musket range, I would think probably because of my massive artillery advantage.

Now when the latest patch came out I uninstalled and did a totally fresh install of the game.....

I dont know if this helps but I thought I would tell you anyway

Steven

In my Kentucky game:

--base reinstall (1.05) directly patched to 1.10d (sorry for not mentioning this; I've advocated doing this in the past)
--latest Kentucky scenario
--Buell (and all subordinate corps commanders): passive posture, retreat if engaged; I can't say how Bragg was disposed (my mistake: it was Bragg, not Hardee as reported earlier)
--Buell's Army of the Ohio attempting to retreat northward from Coffee, TN to Rutherford, TN (to return to Nashville to fix a bad supply situation)
--Buell & co.'s forces all low cohesion (all no more than ~50%)
--fluid situation, open field battle, no entrenchments
--retreat possible everywhere (IIRC)

Battle loss results:

~20,000 (out of ~50,000), or 40% for Buell
13,xyz (out of 13,xyz), or 100% (down to the very last man, horse & gun) for Bragg

Declared Southern victory!

BTW, in my case, if there were extenuating circumstances that might explain high battle losses (e.g., assault posture, hold at all cost; or besieged; etc.), I would know to factor those in, account for the high battle losses thereby, and qualify my statements here.

I don't know if this has been brought up, It seems to me that the casualties in full 1861 campaign are OK but the smaller scenario battle casualties are a bit severe.... has anybody else got any examples for battles in the full campaign?

Steven

Correction!! Thomas and Schofield destroyed Hood's army (Battle of Franklin). I have a friend who lives on that battle ground!! How can he stand it? Johnston should have remained in command to continue to stab at Sherman. IMHO>

tagwyn wrote:Correction!! Thomas and Schofield destroyed Hood's army (Battle of Franklin).

Although he decimated his force at the battle, Hood still had a functioning army after Franklin. He went on to effectively destroy what was left of his army at the Battle of Nashville, about two weeks later.

The Battle of Nashville was one of the most stunning victories achieved by the Union Army in the war. The formidable Army of Tennessee, the second largest Confederate force, was essentially destroyed and would never fight again.

http://en.wikipedia.org/wiki/Battle_of_nashville

Referring to some earlier researches I did, and to set a context for the remarks following:

berto wrote:In the bloodiest large-scale battles of the Civil War (Murfreesboro, Gettysburg, Chickamauga, Shiloh, Wilderness, Spottsylvania, Antietam, and others), the highest casualty rates were ~30% (3/10, or less than 1/3) at Murfreesboro, followed by ~27% (Gettysburg, Chickamauga), followed by the rest at about 20% or below. These were battles of one day (Antietam) to three or more. Note: These casualty figures account for killed, wounded, captured, and missing, not just KIAs. (Source: Wikipedia.)

There's nothing like a little controlled, empirical testing to clarify some issues.

I've conducted a series of tests of the "Thunder at the Crossroads," Gettysburg 1863 battle scenario, chosen because everyone is familiar with it, and because it is likely to result quickly in substantial battles. In all tests, I played the South with the intent to (a) engage the enemy and (b) to keep moving when necessary so as not to invite auto-entrenchment. After each test, I exited and restarted the game for the next test.

Game version: original install patched directly to 1.10d.

Here are the optional settings:

--FOW: yes
--Activation Rule: 2nd of 3 (may move with large movement & combat penalty)
--Delayed Commitment: 2nd of 4 (Small Delay)
--Historical Attrition: 3rd of 3 (both sides historical attrition)
--AI Detect Bonus: 3rd of 5
--Difficulty: 2nd of 4 ("normal")
--AI Aggressiveness: 2nd of 3 ("normal")
--Use All Behaviors: yes
--Give AI More Time: yes
--Activation Bonus: 2nd of 4 (+1 to AI)

In all cases, I set every (Southern) command to: Defensive Posture, Defend & Retreat--i.e., for both settings, level 2 aggressiveness, on a scale 1 (lowest) to 4 (highest).

In the test summaries that follow, I will try to be brief, rounding the numbers, highlighting the results in support of my assertion that too often, battle losses are excessively high. (It would become too tedious to describe each test in detail. Also, yes, some of the minor battles had lower, "historical" loss rates--but far from all of them. The battle results shown represent maybe 40% of the total battles fought in my tests.)

TEST GAME 1
Turn 3, Days 5 & 6
Northern forces 160K, losses 65K, loss rate 40%
Southern forces 106K, losses 56K, loss rate 53%

TEST GAME 2
Turn 3, Days 5,6,7
Northern forces 203K, losses 66K, loss rate 32%
Southern forces 97K, losses 88K, loss rate 91%

Note: Meade's army 203K?! By end of turn 3, ANV was essentially destroyed.

TEST GAME 3
Turn 3, Days 1 & 2
Northern forces 148K, losses 48K, loss rate 32%
Southern forces 78K, losses 35K, loss rate 45%

TEST GAME 4
Turn 2, Day 5
Northern forces 119K, losses 38K, loss rate 32%
Southern forces 106K, losses 52K, loss rate 49%

Note: This was a one-day battle. Even in the real war's bloodiest battles, single-day losses were usually ~10-15%, 25+% tops (25+% Murfreesboro day 1; ~10% per day across Gettysburg's three days; Antietam's single-day losses were no more than 20%).

(Note over all tests: AoP forces often 150K+, 200K even? When at Gettysburg, the AoP was 90K+, and the largest AoP ever deployed was 140K at Chancellorsville. )

Try it yourselves. Play around with the Gettysburg battle scenario and see what results you get. Will your mileage vary?

For me, these latest results confirm what I've been seeing all along:

berto wrote:Battle losses are (sometimes) too high, even way too high.

In fact, I'll go so far as to say: more than sometimes, fairly often even.

Not possibly high, but probably high.

berto wrote: AoP forces often 150K+, 200K even? When at Gettysburg, the AoP was 90K+, and the largest AoP ever deployed was 140K at Chancellorsville. )

Hi
That one is a problem caused by the 1000 men per regiment numbers which, even with hardened attrition option, rarely go down to less than 800-900 men except after heavy combat.
Probably the Gettysburg scenario OOB is very precise, with all the historical regiments represented, but at much higher men per unit that was historically common (around 500 men per regiment during campaigns).

Thats why i think this battle loses tests should be done looking at the % of losses (as you do) and not looking too much to the actual numbers of men lost which is just for flavor (no used by the game engine).

But by your tests if fact there seems to be some problem. The % losses seems to high
Of course on a long campaign game it only happens from time to time, but on a short "big battle scenario" like Gettysburg with huge armies posed for battle, with late war units and plenty of very good leaders around the results are much more common. Is the perfect setup to experience this problem.

On a side note: the number of men per regiment is less of a problem than the too bloody combat as is just a flavor thing. But i think it would be a very good idea to tweak the numbers down (maybe to 600 men per full strength infantry regiments).
With this we would get historical numbers on armies, divisions (5-7.000 men instead of 14.000) and battle losses and the game would gain a lot in immersion

Regards

arsan wrote:Of course on a long campaign game it only happens from time to time, but on a short "big battle scenario" like Gettysburg with huge armies posed for battle, with late war units and plenty of very good leaders around the results are much more common. Is the perfect setup to experience this problem.

See my earlier post in re the 1862 Kentucky scenario, where the setup was not "perfect", and I saw similar excessive battle losses. (I have seen this problem, and commented on it, before in other, non-Gettysburg scenarios.)

I would conduct further tests using the Kentucky and other scenarios, but they don't conveniently lead to significant battles so quickly (i.e., I'm not willing to spend days and days to conduct unassailably "proper" tests across the board and definitively prove my thesis).

Hi!
Berto, don't misunderstand me.
I'm not saying your test are not valid. It's the opposite.
Using the Gettysburg scenario seems a very good idea to test this problem and possible future tweaks as the setup of the scenario is perfect for it. Fast and easy to check results.

I justs wanted to point that on this scenario the "too much bloodiness" happens on 40% of the battles as you said and in the long campaign you only see it form time to time as most battle are not massive two huge armies engagements.
Regards!

arsan wrote:I justs wanted to point that on this scenario the "too much bloodiness" happens on 40% of the battles as you said and in the long campaign you only see it form time to time as most battle are not massive two huge armies engagements.

If I see lopsided or excessive battle losses in smaller engagements, I just shrug it off. I won't assert it to be true, but it's possible that, as percentages, losses in smaller battles and skirmishes are excessive, too. I just don't make special note of it.

As CSA I try to manouvre my Armies in such a way as to catch Feds where I can apply the maximum force concentration against them - and thus destroy Fed Armies. So high casulaties dont come as a total surprise. One I catch Yank army it is a case of Jackson mania - kill them all.

Here's an example of a 3-days battle in late September 61:

Clovis, those results look encouraging!
But I think the real test is the late war battles, where the armies are experienced and powerful and led by excellent generals - that's where I have had the most absurd results.
Have you been able to test in a late war environment?

By the way, I have downloaded your mod and just begun playing it - looks good!

briny_norman wrote:Clovis, those results look encouraging!
But I think the real test is the late war battles, where the armies are experienced and powerful and led by excellent generals - that's where I have had the most absurd results.
Have you been able to test in a late war environment?

By the way, I have downloaded your mod and just begun playing it - looks good!

Much less than for the first year, but nothing was strange in the late years. My mod is based on cohesion but too on some permanent variant like a lesser chance to hit. Units should have too harder time to get experience and then it's another limitation to the raise of losses, whch will occur like in reality howewer.

After all, the Overland campaign in 1864led to a total of around 80,000 to 100,000 losses cumulated for both sides in one month...