Casualties to break a unit

[67th Tigers]

I'm currently researching when units become unreliable on the basis of casualties. There are two major breakpoints:

I: the unit ceases to be able to attack, but is still capable of defence
II: the unit ceases to be able to defend itself

Point I seems to be around 20-25% casualties for experienced units, which is where a lot of "hard fought" units often end up in battle. That makes sense because they'll attack to their breakpoint, and then cease to be able to attack. At around 40% casualties units are incapable even of defending themselves. This is visible in official effectiveness charts, thus:

I then realised that the probability of a unit break I was the same as the observed probability of attacks "stalemating" due to casualties, thus:

We know that units slowly recover morale when rested, and that it takes about 3 months for a broken unit to fully regain combat effectiveness.

My interest was due to a recent argument over the effectiveness of the 1st, 2nd, 9th and 12th Corps on 18th Septemebr 1862. I found that all of the divisions had likely reached breakpoint I (using South Mountain and Antietam casualties, as the recovery wouldn't have happened), and some had likely reached breakpoint II. Estimating combat value from the remaining troops and a linear model of effectiveness loss (value in attack declining to 0 in first 20% and then in defence linearly in the second 20%) and Carman's troop strengths

		Strength	Casualties			Value
Corps	Division	(Carman)	(SM+A)	% Loss	Category	Attack	Defence
1st Corps	Doubleday	2,975	1361​	46%​	II	-	-
	Ricketts	3,037	1222​	40%​	II	-	-
	Meade	2,607	968​	37%​	close to II	-	246
2nd Corps	Richardson	4,029	1136​	28%​	I	-	1,736
	Sedgwick	5,437	2255​	41%​	II	-	-
	French	5,740	1818​	32%​	I	-	1,569
9th Corps	Willcox	3,002	687​	23%​	I	-	1,968
	Sturgis	3,013	827​	27%​	I	-	1,421
	Rodman	2,791	1083​	39%​	close to II	-	85
	Kanawha	2,908	553​	19%​	close to I	118	2,355
12th Corps	Williams	4,725	1076​	23%​	I	-	3,102
	Greene	2,504	650​	26%​	I	-	1,298
Total		42,768	13,636	32%​		118	13,779

Using the 22nd September return for the rebels, it looks like Lee's army had completely pass breakpoint I, and hence would be incapable of offensive actions, but could effectively defend itself:

Division	Str 22nd	Cas	Tot	%	Category	Attack	Defend
McLaws	3,928	2,081	6,009	35%​	I	-	520
Jones	3,810	1,435	5,245	27%​	I	-	2,477
Anderson	5,324	1,465	6,789	22%​	I	-	4,792
Walker	3,428	1,103	4,531	24%​	I	-	2,742
Evans	556	290	846	34%​	I	-	167
Hood	2,847	1,016	3,863	26%​	I	-	1,993
DH Hill	5,071	3,241	8,312	39%​	close to II	-	254
AP Hill	4,777	702	5,479	13%​	0​	1,672	4,777
Ewell	3,442	1,344	4,786	28%​	I	-	2,065
Jackson	2,553	700	3,253	22%​	I	-	2,298
Sum	35,736	13,377	49,113	27%​		1,672	22,084

 

[JD Mayo]

My cousin and I went to Malvrin Hill this year back in February we figured that from every 100 miles they marched from the beginning of the campaign that they had to had several 100 casualties just in that battle alone they had 3,000 for the Confederates but for wounded I am sure they had like 10,000.

 

[Saphroneth]

So the implication of the model above is that essentially 100% of the fighting for the Union in the event of a resumed assault would have to fall upon 6th Corps (which had suffered casualties in Irwin's bde) and 5th Corps (which had suffered casualties in Sykes' division).

 

[67th Tigers]

So the implication of the model above is that essentially 100% of the fighting for the Union in the event of a resumed assault would have to fall upon 6th Corps (which had suffered casualties in Irwin's bde) and 5th Corps (which had suffered casualties in Sykes' division).

Indeed. That was my contention, and it was Harsh's contention.

Then I started looking at the casualty history of 5th and 6th Corps. Both 5th Corps divisions had passed breakpoint I less than 3 weeks earlier at 2nd Bull Run, and it's likely that Morell at least was a bit shaky. Morell had 3,630 infantry PFD on 1st September, and Sykes 3,202 (of whom 523 are Warren), making them 25% and 22% casualties.

6th Corps went into the campaign in good shape. Taylor's brigade (Slocum) likely hit breakpoint I at Bull Run bridge 3 weeks earlier. Slocum's division took 513 at Crampton's Gap and another 65 on the 17th. Smith's division took 20 at CG and 373 on the 17th (mostly Irwin).

 

[Pat Answer]

Interesting. My tactical game project uses 10, 25, and 40% thresholds for die roll modifiers affecting unit "stance" (charge, attack, defend, retreat, or rout). I rather stumbled into this with bits and pieces of information over the years; it is nice to see it here laid out formally. Is the Army document you cited available online?

(Ultimate General can be fun but needing upwards of 75% casualties to cause a unit to quit never sits right.)

 

[67th Tigers]

Interesting. My tactical game project uses 10, 25, and 40% thresholds for die roll modifiers affecting unit "stance" (charge, attack, defend, retreat, or rout). I rather stumbled into this with bits and pieces of information over the years; it is nice to see it here laid out formally. Is the Army document you cited available online?

(Ultimate General can be fun but needing upwards of 75% casualties to cause a unit to quit never sits right.)

I took the table from a thesis.

My reading for wargaming is that there should really be the following thresholds:

0-5% cas: no effects
5-20/25%: decreasing ability to attack linear to casualties, once the threshold is reached a unit can't attack
-40%: unit can defend but gets shakier with casualties
>40%: unit is completely ineffective, depending on the game you might remove them.

 

[Saphroneth]

(Ultimate General can be fun but needing upwards of 75% casualties to cause a unit to quit never sits right.)

I think for your sanity it's best to treat the depleting strength of a unit as remaining men in the line, and that damage incurred results from both casualties and "battle shock". So a unit that was 1000 strong and is down to 300 has actually only suffered say 200-300 men KIA/WIA (and is thus about to disintegrate per the numbers 67th has been discussing) but has had the rest run off.

My reading for wargaming is that there should really be the following thresholds:

It's interesting to think about that in terms of the GCACW series, because they have a "disorganized" state which instantly drops a unit to about 2/3 total combat strength but there's no point at which a unit becomes completely combat ineffective. A 10,000 man division (20 manpower) can suffer 2,500 casualties in one badly lost defensive battle and end up Disorganized (10 combat strength instead of 15) and Demoralize-2 (which is to say, completely incapable of attack) but it still defends almost as well as a fresh unit of 5,000 men (10 combat strength).

In fact, in GCACW terms, Morell's division (13 combat strength) advancing in broken terrain against an initially-10,000-man division after just about the worst lost battle the game can permit (Disorg, Demoralize-2, lost 7 strength points) is at +1 for the enemy being demoralize-2.
Morell's division advancing against a fresh unit of the same strength AP Hill is then (6,500 men, 13 strength points) in broken terrain is at +0. So... almost no difference.

 

[Pat Answer]

I took the table from a thesis.

My reading for wargaming is that there should really be the following thresholds:

0-5% cas: no effects
5-20/25%: decreasing ability to attack linear to casualties, once the threshold is reached a unit can't attack
-40%: unit can defend but gets shakier with casualties
>40%: unit is completely ineffective, depending on the game you might remove them.

Thank you for the link.

That sounds close to what I ended up with, each threshold reached shifts the possibilities of the unit stance until finally a high roll merely keeps a unit together while it rests away from any combat.

I think for your sanity it's best to treat the depleting strength of a unit as remaining men in the line, and that damage incurred results from both casualties and "battle shock". So a unit that was 1000 strong and is down to 300 has actually only suffered say 200-300 men KIA/WIA (and is thus about to disintegrate per the numbers 67th has been discussing) but has had the rest run off.

That's always an important point to remember: "casualties" or "losses" usually exceed the number of men actually "hit".
I found adding the 10% threshold tended to give me consistently historical results with many attacks, especially against prepared enemy positions, falling apart on morale rather than hits. Instead of (game) hours of continuous slugging, I started to see a lot more combat pulses - firing, wavering, falling back to rally, etc. - and the majority of units reaching critical stages much later in the day.

 

[Saphroneth]

That's always an important point to remember: "casualties" or "losses" usually exceed the number of men actually "hit".

The best example of this IMO is that one of the 1st Corps divisions at Antietam (3rd division) was estimated as down to 300 men in line on the afternoon of the 17th. This wasn't even the corps or division-within-that-corps that got worst hit, so functionally speaking a lot of the divisions mentioned above were probably completely combat incapable - as in could be outfought by a fresh regiment, let alone a fresh brigade.