An MG study: Is there a problem here?

[The_Capt]

Maj Battaglia,

The test was designed as a "worst case" for the attacker. As to the sighting of the defender I think we are at a fundemental disagreement. MGs in real life are best sighted to fire from defilade (or from the side). This exploits two things which maximize the effectiveness of the weapon. First of all the "beaten zone" of the MG. It is an oval shaped imaginary area which will receive effective fire at a sustained rate from an MG in the SF role. Keep in mind we are not talking about LMGs on bi-pods but tripod, hardmounted HMGs. Defilade fire puts the oval (and trust me there isn't one in CM or much of one, I have run the test numerous times and I have yet to see an adjacent squad effected by non-targeted fire or take a cas) parallel to the frontage of the position. Two enfilade positions will provide "interlocking defilade fire" as well as mutual support for protection. The second thing it does is exploits the formation of troops in the assault. In order to bring the majority of weapon to bear during an assaut as well as prevent "friendly cas" troops spread out, again parallel to the frontage of the position. Both Soviet and NATO doctrine preach this. So what you get is the oval, covering the advancing troops.

Please if this is hard to visualize just let me know and I will try and post a picture.

Now in your example of a firing squad, if the SS trops were in front of an MG this makes for a very ineffective use of the weapon. Why, because you have to move the damn thing around to hit one soldier at a time. These had to be REMF troops doing the job. If they had lined up one MMG to fire from the side of the line of SS prisoners your cas rates would be much higher.

I have spoken to vetrans of Cyprus who describe "dropping dozens of Turks at a time) with a .50 cal MG sighted as I have described. I have personally seen an Coy cut to pieces by a couple of effectivly sighted MGs (thank God it wasn't one of ours). I think the factors are clear, morale, beaten zone and rate of fire. The question remains as to the degree.

As to the range, again I could set them further back but as CM models it, you lose firepower and effectiveness with range. I agree, you could set those Pillboxes back 500ms and still get a very effective fire pattern. The window of the KZ would remain the same however. If you sight your MGs in real life to fire forward at advancing troops even with greater range you will lose effectivness of the weapon, as well as expose it to return direct and indirect fire. Now I will try the test forward firing just to see the effect.

I understand quite well integration of weapon systems to complete the "chorus of destructive firepower" but this is an analysis of MGs and their combat power. I think it is the attack defence ratio which tells the tale. In CM you need a very low ratio advantage 1.5 to 1 to carry an attack when in reality 3 to 1 is the min. My question is why? Are MGs the problem or is it somthing else. I have no doubt there is a weakness in MGs but I do not believe it has been well-defined (well at least from what I know) and the solution to the problem has also not been demonstrated. I am hoping this is a case of "all will become clear in time" but it is a concern.

I would be curious to know who are the professional military advisors to the team. I know there are a lot on the forum.

Now as an aside take 6 MGs in Close Combat and try the same test. I would be interested to see the results.

[Spook]

<BLOCKQUOTE>quote:</font><HR>Originally posted by The_Capt:

Now as an aside take 6 MGs in Close Combat and try the same test. I would be interested to see the results.<HR></BLOCKQUOTE>

I only tried CC1 & CC2, but up through CC2, my experience with a heavy MG42 was that ONE such MG42 could cause a WORLD of hurt in a very short time at closer ranges. It was just not something to run around in front of if it was unsuppressed.

The question then begs to be asked: which game system gives a more realistic treatment of MG's? CC or CM? I couldn't say for now, but I can say that there does seem a measureable difference in MG effect between these two games. One might just as well argue that MG's are TOO powerful in the CC series.

Anyway, though, I do think that the Capt's points on MG usage & effect bear added discussion. I do think that CMBO DOES apply some "beaten zone" effect where one MG unit firing in the direction of multiple squads bunched together (while targeting one) will inflict casualties on multiple units. But it's something that I haven't measured or observed regularly so to "qualify" if the effect seems about right.

One thing, though, that I would firmly agree with the Capt on, is that if two or more MG's were set up to "interlock" in enfilade fire on an attacking force (firing onto both flanks of the attacker), then you would usually have a VERY effective means of suppression. This doesn't necessarily have to translate into added physical casualties, but it would be very hard to keep troops moving forward under such a crosslinked rain of lead.

Perhaps an added MG unit command could be devised beyond "targeting," like "zone fire" as one possible example. With such a command, an MG unit doesn't "latch" onto one unit tightly for targeting, but rather tries to target as many units within a given area specified by the player. (This could be analoguous to the "target wide" command for offboard artillery.) The idea again is to apply as much suppressive effect to multiple units moving/halted within a given area. Increased casualties, by ensuring a targeting spread to multiple units, may or may not be an offshoot from this.

One thing to note for reference, though. If the attacking troops get within 50 meters of the MG's, then the attackers, if in open ground, will keep moving to close the distance with the MG's or to make for some covering terrain (whichever is closer). This mechanic is understandable, but as a result, close-range suppression can thus be more difficult. Therefore, the hope of the MG's is to instead inflict greater physical losses, and to do so quickly.

So if you run some more MG tests, make a note of where that "50-meter line" is, in front of the MG positions.

[Maj. Battaglia]

Capt:

A couple of points about your last post:

As you say, the model for beaten zone may need work, and may indeed be modified in CMBB. I also agree that infantry may move too fast, and again, this is something BTS said it is working on. Finally, I will argee that more dynamic targeting would be realistic, although I'm not sure if enemy casualties would increase as a result, rather more units might end up suppressed.

My main issue was with your test parameters. Try the test with six wooden bunkers instead of two pillboxes. Use pavement if you want to simulate really open ground. Give the MGs more room. There will be even higher casualty rates than you originally recorded.

Also, you say the test was meant to be worst case for the attacker, but it sounds to me like they did OK, and it is worst case for the defender before the test begins: an entire company has come within 200 m of the defenders' position unharmed before the test begins. Now each MG must take out or force to flee an average of 21 guys in the time it takes them to cover 200m.

Good MG crews can create, and use, beaten zones 1000m or more away. When you say you've seen a whole company decimated by two MGs, at what range did the MGs open up? How long did it take to finish the job? Was there any other supporting fire? What can you say about the quality of their training (and who the heck were they)? I'm not doubting you, just curious. Like I said, I could be convinced to change my mind.

I will say that MGs in WWII were often placed well over 100m behind the supporting infantry, which supports the idea that they were meant to be longer range (longer than 200m) support weapons (i.e. distance firing was more useful than close in). In CM scenarios, when used in conjunction with infantry and other assets, I think they perform quite well. Whether individual MGs perform historically, I am not 100% sure. But to me they seem to be a pretty good approximation.

[The_Capt]

OK a couple of good ideas from the last two post.

First, I will try targeting the ground with area fire and let the infantry run thru it maybe we will see just what kind of beaten zone the MGs have by doing that.

I will try the wooden MG bunkers and pavement. I will even back them off but I don't think that will help the MG case as firepower bleeds of over distance. Worst case for an attacker in CM will be when the MGs open up at close range in an ambush situation.

I will ensure the same exposure time 45-30 seconds, in fact I will narrow it down further to ensure a consistent exposure time. I picked 45-30 because that was how long the squads took to cover the 100m of open ground to the flag while being exposed to MG fire.

In the game if you start firing at 1000m you are going to be killed by arty without hitting anybody. In real life again, as the Maj said you can lay down a good beaten zone and get some "bang for the buck". My guess is that at 1000 (well I'll keep it to 500m) that the cas rates will be much lower and all the gain will be is exposing the position.

If anybody has Close Combat and is interested, please set up some similar trials so we can compare. My "gut feel" is 25% survival rate for this situation with no formed units effective in real life. Problem is there are no "staff tables" for this so we are going to be guessing. 6 MGs in interlocking defilade firing at about 100m distance from a coy.

[argie]

I found that MGs are quite effective firing from more than 500mts. If they are well concealed, they beat the Squads and tehy are no exposed. Also, if they are in wooden bunkers or pillboxes, are amost immune to Arty. I found that HMGs in Heavy buildings firing long distance are quite capable of making the enemy loss the timing, most of the time not being discovered until the enemy units are at less than 500 mts. Using a sort of talking guns, i.e., firing with alternates HMGs in each turn, hiding the other, helps to make it work. When the scouts try to get close for ID the fires, they fall in the ambushes made with infantry. I used HMGs with good effect so far as 1200 mts. This works well also with 20mm AA guns, except for the part on having them in Heavy Buildings... Woods are the choice then.

[Kanonier Reichmann]

<BLOCKQUOTE>quote:</font><HR>Originally posted by Hilltopper:

I have rushed lone MG positions and bunkers knowing that enough will get thru to take the position. Same with unsupported bunkers...just a quick BONZAI charge to the door and you've got it sometimes with very few casualities.<HR></BLOCKQUOTE>

How do your men kill the defenders in the bunker... by bashing them over the head with their specially tendered miniature potted trees?

Regards

Jim R.

[JasonC]

There is a very basic problem in this sort of analysis and the question it implicitly raises as a game design issue, that no one has noticed or mentioned yet. The real simulation problem is that the effective firepower of MGs in real life - or other small arms for that matter - are not constant, and do not add up linearly over time. But the game treats them as such, and therefore its hit rates must be average figures for a variety of situations, only one of which is actually occuring in this test attempt.

What happens in real life is a few shots may be directed at quite exposed men, but soon after those have their effect, small bits of cover, going to ground, and changes of route break up LOS between the few identified firing positions and each surviving member of the attacking infantry. Subsequent periods of firing are therefore not as effective, because they tend to be directed at men exposed only from time to time and briefly. And MG ROF varies over extremes between the two situations. The max ROF of MGs is high enough to burn their entire ammo load quite rapidly, but they cannot be fired that fast for any length of time, or they run out of ammo (or overheat, or both).

But in CM - or any game that attempts to treat firepower in a relatively constant way from shot to shot - each shot by the same firing unit at enemy in the same sort of terrain must result in approximately the same average effect. One cannot then design the proper casualties per small period of complete exposure, use that as the average rate of hits, and base CM firepower off of it. If you did, what you would see is far too much *cumulative* killing by MGs, not in 30 seconds, but in 20-30 minutes.

In fact, CM combat at present is more bloody than the real thing was. Losses of 33% for the winning, and 67% or more for the losing side, are common in short battles of less than half an hour. While in reality, infantry units fought each other for hours on end in close proximity, often taking only 10-20% losses in the process.

This was first and foremost because they did not mash themselves together as aggressively as CM players do in games. And CM players, worried about winning, do not mash their troops together as hard as in this test, either, because loss of half a company to move them 200 yards is not tactically acceptable in most situations.

It is still true, as others have noted, that green and regular troops play more realistically than veterans do. Saying "vets are commonly available" is a reflection on players, and perhaps on the fact that QB settings group "regular" with them rather than with "green", but not on MG firepower.

What the test is implicitly trying to determine is the peak killing power of MGs per unit time. The tester expects that number to be accurate, without making others - like killing power over 30 minutes or 60 times as long - being off. What the test actually showed, however, was each MG bunker accounting for ~20 men per minute. I say this because the detailed explanation said the exposure time was 45 seconds, and the average men hit was in the low 30s. Add 1/3rd again to get the amount per minute, and divide by the number of MG bunkers.

Now, if that rate were sustained as an average over a whole fight, it would put down 600 men per MG bunker. Recall that the total casualties on Omaha beach over more than four hours, by nearly a division's worth of defenders, came to less than 10 times that (around 5000 actually). The number is quite sufficient to make it tactical suicide to leave large bodies of infantry perpetually in the LOS of MG bunkers while in open ground.

Notice that 20 men per bunker per minute, if only *one* bunker were firing at a time, would be enough to generate the Omaha causalty toll in a little over four hours. Not counting the mortars, the artillery, the field guns, the infantry and their LMGs, or allowing for lots more than one firing much of the time. So peak rate and sustained rate are obviously two quite different things.

As a game design matter, what can force them farther apart, and thus increase rush losses without unduly increasing sustained ones, which are on the high side as it is? Lower morale can, by causing units to break after short exposures to fire and avoid it afterward. Ammo limits can, by allowing many hits for relatively brief periods, then low rates of fire thereafter. These are possible now by using regulars and greens, and longer tests.

More variable ROF can, with rapid firing rare but possible, slower firing normal. A problem remains in that case to match up the more rapid fire periods with the proper situations, which incidentally are not obviously linked to range. And the way the way cover works, the "% exposed" numbers, can force the two apart, if occasionally they can be very high for brief periods, but usually are moderate or low.

I have suggested in the past my primary recommendation of this subject. Which is that % exposed numbers be tied to movement states. Thus a running unit would get little benefit from the lighter forms of cover, and have 90s exposure in open ground. While stationary units would not take unrealistic casualties over longer periods, because their exposure numbers would be lower.

I have also suggested that "hot MG", variable ROF behavior, be tied more to target cover state (% exposure), rather than to range. Instead of making MGs more like normal infantry, with the same sort of steeply increasing fp envelope (as rapid fire in close would tend to do), this would leave them longer ranged weapons. But I think it would accurately model their greater pinning power, rather than killing power, at range.

Because, combined with the above exposure-speed connection, it would raise losses against those who moved "recklessly" in MG LOS, while more deliberate movement might be possible (at least for modest time periods - staying too long means more shots, obviously).

At the risk of repeating myself, I return from my CM2 tweak suggestions to my main, basic point. All CM firepowers are averages over most situations the units will face when used realistically. They will give reasonably accurate overall results when infantry spends some time rushing through open, some times out of LOS, and other periods of time in decent cover.

You would not achieve that proper long-time average if you simply increased the amount done in the most exposed short periods, because the higher rate would "multiply through" to a higher overall average, unrealistically. It took more than two MG bunkers to shoot down a division in a few hours. The losses would add linearly, because CM uses discrete shots over and over and resolves each in the same way. If the game conditions are the same, the average results will be. And that already gives results that are on the high side, not the low side.

CM firepower tends to do less in a maximum minute than real combat sometimes can (though its high end is still quite high). It tends to do more in 30 minutes than real combat does. It strikes a design average, and is "off" in my opinion by relatively small proportions at each end (with randomness, incidentally, better able to handle the variation at the short-period end). The reason real life is more "skewed" (higher in 1 minute, lower in 30), is because in real life losses do not add up linearly over number of shots. Because men avoid LOS, retain better cover, do not mash against each other so forcefully - all making the periods of high hit rates, brief and exceptional.

For what it is worth...

[Michael Dorosh]

Once again, Jason has made some interesting comments that seem to be correct, but has failed to properly support them.

Where do the 10 and 20 percent casualty figures come from?

Looking at some real life examples:

At Buron on 7 July 1944, the HLI of Canada suffered 262 casualties during its assault. The defenders' losses don't come immediately to mind, but definitely exceeded 20 percent.

262 men, the majority of whom came from the rifle companies, who at full strength would have numbered 480, accounts for in the neighbourhood of 50 percent.

See my scenario "Bloody Buron II" for a simulation of this action - unfortunately, without an accurate appreciation of German forces down to the last man, we cannot see if this historical result is likely when played in CM.

This is obviously only one battle; perhaps Franko or another scenario designer, with access to detailed OOBs for both sides of a particular action, can speak with more authority on how well CM simulates casualty rates? While I suspect Jason may be correct in the over-bloodiness of CM, I am not convinced beyond any kind of gut level that this is so and would prefer to see some hard evidence.

Franko?

[Username]

I think the covered arcs, increased rates of fire, adjusted response to fire, decreased infantry speed using assaults, etc., Will all add up to a completely different infantry game .

Proper overwatch and cover fire will be needed and other tactics brought closer to reality.

It is hardly worth discussing unless someone with inside play information can relate some test results.

Lewis

[ 08-04-2001: Message edited by: Username ]

[JasonC]

Average losses were obviously lower than the anecdotal worst case of a particular formation in a particularly expensive attack. Units did not regularly take 50% losses in a single day's action (let alone 30 minutes), for the obvious reason that they would not exist after doing that a few times - they'd be gone inside a week.

Detailed statistics are available for some longer time scales, that give an idea of overall loss rates. The average losses per day in combat for a US infantry division in the ETO was around 50 men. There are a few outliers - a handful of divisions that didn't reach the front until 45 have averages lower than 25, and one, the 106th, was badly smashed in the Ardennes offensive shortly after going into the line, and has a higher average, 130 per day. Most are between 25 and 70. Some of that may reflect spells out of the line, and of course two up, on back formations were also standard, within the divisions, so that may reflect the losses of four battalions, 3/4ths of the time. Around 15 men per up battalion per day.

That reflects plenty of quiet periods and exploitation moves, however. In heavy action, the same figure is a typical average for a unit one step lower on the organizational scale, and in the heaviest fighting it might be twice that. That is, 15-30 men per front line company per day.

If you want examples, the 30th Division in the St Lo fighting lost 3934 battle losses in 15 days, or 262 per day. Over that period it cycled its regiments and companies, obviously, and also took replacements regularly. It was fighting with 4-6 battalions up regularly, sometimes 8 with reserves committed, for brief periods. That is 44-66 men per front battalion per day, or 15-22 per company. That would be 11% at TOE, but they were regularly somewhat below it for the second half of the period, but above 50% (they rotated out for replacements, etc). The 35th division in the same period averaged 162 per day, somewhat lower.

If you look at the German defenders, you find similar results in men per day. After losing 1200 men on D-Day, the 352nd division averaged about 250 per day for the next 2 1/2 weeks, and 150 per day for the 2 1/2 weeks after that, already down to about half strength. Rear area troops were being cycled forward to replace infantry losses by then. The 3rd FJ averaged 135 men lost per day in its first month in combat, with lighter fighting in the first half of that (after a few heavy days), heavy fighting in the second half of it.

You see anecdotal reports of higher loss rates for particular smaller formations, used by staff to illustrate losses by their "outliers". Thus, the U.S. staff history of the 2 week push to St. Lo records two seperate occasions on which a unit took 2/3rds losses in one affair, in both cases by being hit with mortars. These are described as "disasters". But both were platoons. Or you read about a German battalion that took 50% losses in 3 days - which is 20% per day - mentioned as an outlier, the worst off unit in the division. This is on a sector of the front where the US was expending 20,000 rounds of arty per day, on a corps wide front.

Of course, these average losses are made up of some days of light action, some days in which particular sub-units get mauled while others are relatively unscathed, some days without any serious opposition, others with multiple combats longer than 30 minutes, or under sustained shelling outside the scale of CM battles. But it gives an idea of the resiliency of WW II infantry units, and of their gradual reduction by attrition if not receiving regular replacements in quantity, if in heavy action.

Even 5% losses repeated every day for a month, if no replacements are received, will reduce a unit to 1/5th its original strength. The same will happen if a unit takes 15% losses in more intense fights twice a week for five weeks, with local inaction, or other units rotated forward to bear the brunt at other times. If the loss rate per day for the whole formation (rear area troops included) gets much above 2.5%, the unit cannot sustain the pace for more than about one month, without regular infusions of new men. The front line units are taking more than there share of that kind of average, but they will be well above that even moderately sustainable rate if they average more than 20% losses in every local fight.

Of course, that did happen to some units. They were reinforced, relieved, or destroyed if they could not get either sort of help. But you do not find 3 US infantry divisions running into 2 German ones (one understrength, say, to give more than 2:1 odds), and wiping all of them out 3 days - let alone in three 30 minute actions, as you might easily imagine from CM-scale fights.

[ 08-04-2001: Message edited by: JasonC ]

[Spook]

All good to relate, Jason, but there's little connectivity here between your examples of attrition rates and what one or more MG's COULD do at a given range in a given timeframe.

Under select conditions, MG's could be devastating; under another set of conditions, they could have only minimal effect. But an MG's ability to "apply" attrition wasn't directly defined by a unit's willingness to absorb it.

Your discussions would be useful as a seperate topic, in that casualty percentage levels for a CM force might be adjustable by scenario design so to "call off" a scenario sooner.

[JasonC]

Well spook, if you will go back and look, I think you will find that subject was the exact matter discussed in my prior post. The game design issue is that what an MG would generally do in 30 minutes was not 30 times what it might do in 1 minute. It was much less.

But that means you need two things in combination - one, some averaging or trade-off balance between getting the fp per shot right, and getting the fp over a whole game right. And two, some factors that drive apart the amount that a "best minute" can do, compared to an average minute, thus less good minutes of fire.

I gave examples of the second sort of factor, some already available, some possibilities for CM2 tweaks. Changes in % exposed numbers, relating those to movement state, varying ROF by % exposed number, lower morale so units break and avoid LOS sooner, etc.

But if people recognize the general issue, they can also be realistic in their expectations, and see that some of the first - averaging - is going to be involved too. In general, if the game conditions are about the same, 10 MG shots are going to do around 10 times what one shot does. Just from game mechanics.

To keep that from being too high on the many-shot end, and thus from causing excessive losses in other, less ideal situations for MG shots, that is going to mean something a bit low on the one-shot end, if you are thinking of best case, most effective shots and running men in the open, close by.

[Spook]

I know what you were talking about in your initial post, Jason, and my prior feedback was directed rather at your more immediate post. But even with your initial post, your extrapolations can be problematic. Borrowing from above:

<BLOCKQUOTE>quote:</font><HR>Originally posted by JasonC:

What the test is implicitly trying to determine is the peak killing power of MGs per unit time. The tester expects that number to be accurate, without making others - like killing power over 30 minutes or 60 times as long - being off. What the test actually showed, however, was each MG bunker accounting for ~20 men per minute. I say this because the detailed explanation said the exposure time was 45 seconds, and the average men hit was in the low 30s. Add 1/3rd again to get the amount per minute, and divide by the number of MG bunkers.

Now, if that rate were sustained as an average over a whole fight, it would put down 600 men per MG bunker. Recall that the total casualties on Omaha beach over more than four hours, by nearly a division's worth of defenders, came to less than 10 times that (around 5000 actually). The number is quite sufficient to make it tactical suicide to leave large bodies of infantry perpetually in the LOS of MG bunkers while in open ground.

Notice that 20 men per bunker per minute, if only *one* bunker were firing at a time, would be enough to generate the Omaha causalty toll in a little over four hours. Not counting the mortars, the artillery, the field guns, the infantry and their LMGs, or allowing for lots more than one firing much of the time. So peak rate and sustained rate are obviously two quite different things.

<HR></BLOCKQUOTE>

And that's just it, Jason. What the Capt was trying to grapple with was a situation that would focus far more on a "peak" (close-range, short time interval, with targeted troops moving in a highly exposed state) than on "averages," 30 minutes or otherwise. It's within bounds to infer that we CM gamers should consider how it "averages out," but again, are you or I or anyone else here "on target" in regards to "historical averages"?

What the Capt has posed is interesting, but I myself am not ready to venture an opinion as to whether or not his test results were "right" or "wrong" in regards to historical trends.

I will iterate from an earlier view that I would rather that a comparable "test" be run with CM2 beta, with the results related in comparison with the Capt's here. Then let each of us draw our own opinions as to how we think that MG effect has been improved (or at least changed) from CM1.

[Michael Dorosh]

<BLOCKQUOTE>quote:</font><HR>Originally posted by JasonC:

Average losses were obviously lower than the anecdotal worst case of a particular formation in a particularly expensive attack. Units did not regularly take 50% losses in a single day's action (let alone 30 minutes), for the obvious reason that they would not exist after doing that a few times - they'd be gone inside a week.<HR></BLOCKQUOTE>

That depends on how many times they fought a CM type battle. The Canadian Black Watch was wiped out on four seperate occasions; and much of the rest of the time, the Calgary Highlanders had to carry the 5th brigade. They were the worst affected of Canadian inf bns in NWE, but other battalions were not far behind. Most of the actions fought by infantry battalions in NWE were patrol actions - not CM type battles. So I think your average losses per CM type battle are a little low, barring evidence to the contrary.

<BLOCKQUOTE>quote:</font><HR>Detailed statistics are available for some longer time scales, that give an idea of overall loss rates. The average losses per day in combat for a US infantry division in the ETO was around 50 men.<HR></BLOCKQUOTE>

This seems irrelevant.

<BLOCKQUOTE>quote:</font><HR> There are a few outliers - a handful of divisions that didn't reach the front until 45 have averages lower than 25, and one, the 106th, was badly smashed in the Ardennes offensive shortly after going into the line, and has a higher average, 130 per day. Most are between 25 and 70. Some of that may reflect spells out of the line, and of course two up, on back formations were also standard, within the divisions, so that may reflect the losses of four battalions, 3/4ths of the time. Around 15 men per up battalion per day.

<HR></BLOCKQUOTE>

Figures per battalion would be of more use.

<BLOCKQUOTE>quote:</font><HR>That reflects plenty of quiet periods and exploitation moves, however. In heavy action, the same figure is a typical average for a unit one step lower on the organizational scale, and in the heaviest fighting it might be twice that. That is, 15-30 men per front line company per day.

If you want examples, the 30th Division in the St Lo fighting lost 3934 battle losses in 15 days, or 262 per day. Over that period it cycled its regiments and companies, obviously, and also took replacements regularly. It was fighting with 4-6 battalions up regularly, sometimes 8 with reserves committed, for brief periods. That is 44-66 men per front battalion per day, or 15-22 per company. That would be 11% at TOE, but they were regularly somewhat below it for the second half of the period, but above 50% (they rotated out for replacements, etc). The 35th division in the same period averaged 162 per day, somewhat lower.

<HR></BLOCKQUOTE>

That's a little better, but how many CM type battles were fought? What was the largest loss per battalion in a single action?

[Michael Dorosh]

Double Post.

[ 08-05-2001: Message edited by: Michael Dorosh ]

[argie]

Go to http://salts.britwar.co.uk/salt5.htm and look for "WO 291/1169 An Analysis of Infantry Rates of Advance in Battle"

[JasonC]

"this is all averaged out over a variety of types of combat."

The fight forward to St Lo was slogging frontal attacks through thick hedgerow country against determined defenders. The divisions involved attacked every day for two weeks with 4-6 battalions "up", sometimes more, as I have said. The attacked daily and measured their gains in hundreds of yards. The attackers had an important edge in weight of artillery support. The result was division wide losses between 135 and 265 per day, for both attackers and defenders. It is mathematically impossible to reconcile those facts with the idea that a *typical* battalion attack involved 50% losses in half an hour.

"The numbers are interesting but essentially meaningless"

They are not meaningless, and they are also not even outliers, unlike the anecdotal evidence you prefer to cite. With dozens of battalions fighting for hundreds of days, you take the worst you can find and pretend it was typical. That is the shoddiest type of statistical ignorance. The averages are much lower, and some occasions - such as less intense fighting - have to be well lower than the average for that to be the case. The average is not 250 either - that is a high figure for heavy offensive fighting. The average for all kinds of warfare - your patrolling and static included, etc - is only 50 per division per day. Including plenty of days "25 and under" to balance the heavier 150-250 range losses for heavy fighting.

"suffered 325 casualties among their rifle companies at Verrieres Ridge - the worst day for the Canadian Army since Dieppe and admittedly atypical"

So, it is so extremely relevant because? Because it is an outlier, the worst you can think of out of dozens of battalions over hundreds of days. You might as well call US 29th Infantry's losses on Omaha beach typical. Which would be silly, not to mention insulting. The reason historians cite the outliers is to give a sense of the worst things sometimes got, to impress upon readers what the participants faced.

But there were participants. For numerous battles in a row. The war was not over in 5 weeks with almost everyone involved a casualty. All you have to do to see how impossible the idea that such losses were typical is, is to assume that really is what usually happened in a day's attack, and deduce the consequences. No one would be left, or attacks could not occur daily, but only once in a blue moon.

"many battalion level battles inflicted casualties on this order"

"Many" is a wonderful word. It means more than three. In that sense, the statement is certainly accurate. But the average or typical battalion level fight did not inflict losses of that order. On the whole front, there were hundreds of battalions engaged for hundreds of days, and spent a significant portion of them attacking or defending, actively, sometimes for more than 30 minutes, too. "More than three" out of thousands or tens of thousands, does not rise to "typical".

"If a 10,000 man division loses 250 men in a day, that may not seem like much"

That is because you haven't paid much attention to the math I already presented. It sure does "seem like much", and in fact it is on the high end of results for divisions attacking, in heavy, sustained combat. It is five times the average for US infantry divisions for the whole war. Between that, and half of that, was common for heavy fighting periods, rather than "patroling, static warfare, unopposed advances", etc. And as I already showed, losses at that rate will reduce a division quite rapidly, over a period of a month or two, unless it receives a large stream of replacements continually. 250 times 30 is 7500, which is high losses indeed, for one division in one month. In fact, it would turn over the men in the front line companies more than once. Recall that infantry divisions only turned over their entire personnel twice in the entire campaign in the west.

"if all 250 men come from a single engaged battalion"

But they don't. They come from the division, not one battalion. Which has 4-6 battalions fighting in the front line, in addition to lighter losses scattered through its rear area units from artillery fire, modest losses to other arms besides the infantry, etc. If every battalion in the front line were taking those losses in every attack, the divisional total would be five times as high. But it flat isn't.

The losses in the real deal were lower than they typically are in CM because the real participants did not mash into one another as recklessly as players push their cyber warriors together, first and foremost. One can also deduce from this, however, that it is not the case that CM firepowers are ridiculously low by historical standards. As the entire thread has dicussed, there may be extremes of exposure that CM depicts as less bloody than they would be, but over whole fights CM combat is certainly not less bloody than the real deal.

[JasonC]

Ariel's data fit very well with the figures already presented. Losses in 217 actions, with advances of around 600 yards per hour (plus or minus 200) against "heavy" opposition rather than light, averaged 17 per company in 1 hour fights and 24 per company in 3 hour fights. The highest figures out of 217 were 32 per company in 1 hour and 57 per company in 3 hours. Some involved single company advances, some full battalions. Figures for slight opposition have one digit.

One would therefore expect heavy opposition to generate losses on the order of 60 per battalion facing it, light opposition around 10-20. 150 to 250 for an attacking division in a day fits, with 2-4 out of 4-6 up battalions hitting heavy oppposition, 0-4 hitting light. Checks with the division level figures for heavy action. The overall war-long average then includes some static periods, lots of light opposition, and a moderate amount of heavy opposition (say about half the front, half the time, etc).

CM battalion level fights of one hour do not typically involve only 50 casualties on a side, but by the data given that was fairly typical for advances against serious opposition.

[Username]

Jason is right that MOST CM battles are at the FAR end of the scale. Most CM battles are not conducted anywhere near reality either.

In reality, an attacker does not have the coordination the game allows. In reality, most moving units, unless part of some organized assault or human wave, would just stop and become pinned instead of running willy nilly.

I would like to see orders limitations at some stage of CM evolution. This would be when you have depleted squads in a platoon without a HQ not having many options. Perhaps defend/withhdraw being the only remaining orders they will recieve. This would crimp an attack when you can't give attacking type orders. The player would become mindful of concentrated casualties within units like platoons and companies.

In the meantime, it would be nice if there was a formation elan level. Certain units could and would take atrocious levels of casualties. This elan number would represent that. Many US units wouldnt and the player would have to use arty/tanks/smoke better than most.

Lewis

[Maj. Battaglia]

I think a lot of intersting points have been raised by Jason C and the subsequent posts. These issues are related to the original question posed here, the general question of MG effectiveness, and the CM engine in general. Not to squelch debate, but it seems that we are getting off the point. The good ideas and debate topics, especially made by Lewis in his last post, deserve a thread of their own (although Lewis may have already addressed these before, I seem to remember so; that doesn't mean they can't be again).

The_Capt originally brought up the topic of MG beaten zones and provided the methodology and results of a test he ran. I'd like to return to that again if I may.

Capt, have you had a chance to run new tests? To clarify my point on range, I mentioned that as a trigonometric argument: the further the distance from MG to target, the longer the enemy is in a true enfiladed target area. True, increasing distance decreases numerical effectiveness, but I was suggesting maybe an extra 100m. This also allows a slightly wider beaten zone (or should). Also, the longer the "gauntlet," the more time the MGs have to work. I thought that was important because there were more individual units than MGs.

I seemed to remember BTS saying, and have since confirmed with tests, that there is a "beaten zone," or at least a MG penetration model in CM. Squads have suffered casualties when a nearby squad was the target. Other nearby squads suffered suppression effects. I will agree that the model is probably not true to life, and I think this is being modified for CMBB.

I found a site hosted by a guy who collects, of all things, slide rules. BUT, he does have a slide rule made for Vickers MG crews in WWII that includes a way to determine the beaten zone over range. Up to 500 yds, the length is infinte, but width is less than 2 meters. If you haven't seen one of these, it is worth checking out: Vickers MG slide rule

At short ranges you will do severe damage to whatever is within the 2m wide zone and along your axis of fire. But if the men are spread out somewhat and moving quickly, you have to follow them because your fire covers fewer of them. The closer you are, the larger the angle to be changed and the faster targets pass through your field of fire.

I would argue therefore that CM models things pretty well (with room for future enhancements) in terms of MG effectiveness. 1 MG vs. 1 squad, whether the MG has them in enfilade or the squad is making a frontal assault, and the squad is going to suffer heavily. Once you start adding more units, the MG can't cope with the numbers. In the former case, men will get by and in the latter the MG is doomed (but the infantry will still suffer some stiff casualties). From what I can tell, this is realistic.

[Michael Dorosh]

Lewis et al; I am not saying JasonC is wrong, but he hasn't presented anything in the way of proof. Extrapolating numbers and using vague generalities like "heavy fighting" etc. does not provide decent numbers - I still haven't looked at the other reference but hope to do so soon.

It's quite possible for a division to advance and 'fight' without engaging in anything like a CM battle - ie two sides roughly equal, weighted for attack/defence, and capable of defeating the other. A division generally tries to avoid fair fights such as that.

At my web site at

http://members.home.net/calgaryhighlanders/

I've detailed all the fatal casualties for one infantry battalion in NWE from July 1944 to May 1945. 400+, with a handful of non battle casualties (interestingly, including one suicide shortly before D Day). Great, but so what? You'd have to look at the battalion's history to know how many "fair fights" it got into.

Interestingly, Jason's statement is backed up by looking at Walcheren Causeway - this one has been simulated for CM and is at my Canuck website. In the real battle, the Highlanders suffered 64 casualties all told, killed and wounded. You would be hard pressed to duplicate that feat in my CM treatment.

I would be interested in other relevant examples of this rather than irrelevant number crunching. Canadian infantry battalions in NWE suffered on average 400 battle deaths during their service and on the order of 1600 wounded in action; If you exclude the winter stalemate from November to the 1st week of February, you have seven or eight months in which they saw intense front line action. For the Calgary Highlanders, that works out to about 300 killed and wounded every month, or 10 a day.

But if you look at the table of fatal casualties, you will see that they are not evenly spread out; in fact, you can tell which days involved the toughest fighting (ie CM style battle) - Walcheren, Hill 67, Hoogerheide, Wyler. The overall numbers become misleading.

My point, and perhaps Lewis is saying this as well, is that CM does indeed this "far end of the scale", but that these kinds of battles did occur - Buron, Verrieres Ridge, Point 195.

Perhaps the problem isn't with machineguns, but rather in the way BTS handles casualties? For every person considered "wounded" in CM, perhaps one might consider them as simply hors de combat - in the way Squad Leader used to abstractly model those who simply stopped fighting during a certain action.

Incidentally, I agree that the modelling of machine guns in CM needs work - but since they are doing this in CM2 anyway, I don't understand the need to discuss it until we've seen the fixes. I just don't see that extrapolating divisional casualty figures and using them as grounds for demands that changes be made makes any kind of sense.

[The_Capt]

OK I see this has generated some good discussion.

Maj B,

Yes I ran some preliminary tests and I have found that the "beaten zone" of the MG Pill boxes is razor thin at all ranges. It is there though but a squad has to be standing pretty much in line with the one next too it. Cas rates at 400m+ dropped off dratically as did the NE status. Your point is valid though, MGs are best used at range where the beaten zone widens up. The Vickers data is strange because the ol C-5 "Gpig" which was basically the MG1919 had about a 50m wide beaten zone at the same range. I would have to check that figure but it was in the area. Does CM model each MG types "beaten zone" differently? Good question. Beaten zones are also dependent on sighting and terrain contour. But from what I can see the effectivness of MGs in CM degrade with distance, when in fact I think they should increase due to the greater area the weapon can cover.

Jason,

As usual you are a fountain of hisorical facts and figures. But unfortunately they really do not apply here. I am trying to isolate the "mad minute" here which is the max effect of MGs under optimum conditions. Maj B has brought up some good point as to what those conditions are and I am working on them.

Your points on protracted conflict over time are very valid. I have no doubt CM i smore bloody than reality but I think it has less to do with weapons effectivness and more to do with several other factors 1) Gameplay, we do here much we wouldn't have the walnuts to try in realife 2) Troop morale model, in real life troops are much more cautious than in CM becaus they aren't backed up on the hard drive 3) Too much CCC and sighting ability, things move much slower when you not only know where the enemy is but where some of our own guys are. So trying to extrapolate without taking into account these other factors has programed error into the equation early.

What I have set up is the first 45 seconds of the "The Somme". Totally exposed troops, in a dead run, ambushed by two pillboxes at close range (100m) and the question is: "Is what we see in CM realistic"? Now we can check (and have) what CM says about it but we have to draw on history and eyewitness accounts to figure the same thing in real life. There will be no conclusive evidence but my "fee" from both experience and personal observation is that the MGs in CM are weak. The only hard evidence I can find is the low attack ratios needed in the attack. And this is circumstantial at best. My original statement stands "There is no way in real life that a Coy running, over open ground, between two Pillbox MGs with 3 MGs apiece is going to get 60-70 percent of their force to the flag." Now whether that is because of casualties or a morale/self preservation failure, again it remains to be seen. I guess my point is that if CM isn't getting MGs right than a major weapons system isn't "getting it right" which in turn effects the whole game.

Of course if you don't care about realism than it isn't a problem, you simply adapt.

I think that steps have been taken to address the issue and I will be running the same test in CM2 to see the difference. As has been stated, CM2 could be a very different game with very small changes to the engine. CM is now very infantry heavy IMO.

[Username]

I am rereading Steel Inferno. Its apparent that casualties are higher when elite troops and units with high elan meet. The SS fought like rabid wolverines. The canadians, being mostly volunteers, had a very large capacity for not backing down (even though they had crappy officers and brit equipment).

But the numbers in the book rarely match the typical CM battle. Jason it right in that the game doesnt represent typical WWII combat casualties EFFECTS. My point is that the player just isnt hamstringed by casualties. In reality, battle usually meant losses to the units officers and NCOs. This would shut down attacking as coordination is lost.

There isnt enough rallying in the game. Units are capable of doing that themselves usually.

Another consideration is that line units, if they take the typical 25-50 percent losses seen in alot of games, would be combat ineffective thereafter even if resupplied with men. The green troops being untried and the vets unsure of them and of the leaders who are getting them killed.

Elite units like armor and airborne realize heavier losses are part of the job because they are not going to do it everyday. They are like fire brigades that place winning over losses.

Lewis

[ASL Veteran]

Did somebody mention the Somme? How about this for the extreme end of the scale?

<BLOCKQUOTE>quote:</font><HR>The 30th Division had lost a total of 3011 casualties, while the 18th Division lost 3115. It was, on the whole, a successful day for the 13th Corps <HR></BLOCKQUOTE>

This Corps did the best on the first day of the Somme. Here are some others

<BLOCKQUOTE>quote:</font><HR>General Horne, commander of 15 Corps, upon receiving reports of success to both his right and left, ordered the 22nd, 62nd, and 50th Brigades to assault Fricourt at 1430. Although the 1st Royal Welsh Fusilers of the 22nd Brigade actually reached the outskirts of the village, 50th Brigade was decimated and made no progress; 7th Green Howards lost 15 officers and 336 men in 3 minutes. <HR></BLOCKQUOTE>

I would like to point out that there were more MGs per unit in WW2 than in WW1, and that the lethality of the squad was much greater in WW2.

<BLOCKQUOTE>quote:</font><HR>For this, 7th Division lost 3380 men, 17th Division 1155, and 21st Division 4256; a heavy price for such gains. <HR></BLOCKQUOTE>

Moving on down to the 34th Division we see this

<BLOCKQUOTE>quote:</font><HR>The assaulting battalions left their trenches at 0730, within ten minutes 80% of their troops had fallen. <HR></BLOCKQUOTE>

Now the 8th Division

<BLOCKQUOTE>quote:</font><HR>The three battalions of the 25th Brigade lost over 50% of their strength and the 1st Royal Irish Rifles succeeded in getting only ten men across no-man's land.<HR></BLOCKQUOTE>

finally an overall look at 3 Corps

<BLOCKQUOTE>quote:</font><HR>Out of approximately 17000 troops of 3 Corps who went into action that morning, 11500 were casualties, mostly by 0800, and the German line remained unbroken. <HR></BLOCKQUOTE>

the start time was 0730 - that's only 30 minutes BTW. All these casualties are for the first day of the Somme - about 12 hours. Overall the British army lost 60000 men on the first day, some 19000 in the first 30 minutes. The spots where the British did the best was where they beat the Germans to their trenches after the bombardment lifted.

I've never had any issues with the casualties generated by firepower in CM. Overall the firepower in CM feels about right. I do think that the MGs are undermodelled for one specific reason, and that there are other issues such as cover states and squad behavior which I think Jason has alluded to. I am also sure that MGs will be more effective in CMBB and that the increase in effectiveness will not be sufficient for me since I doubt that true grazing fire will be included (do to its difficulty to code).

The Major (my goodness, we have a Major and a Captain in this thread :eek: ) has stated that:

<BLOCKQUOTE>quote:</font><HR>I would argue therefore that CM models things pretty well (with room for future enhancements) in terms of MG effectiveness. 1 MG vs. 1 squad, whether the MG has them in enfilade or the squad is making a frontal assault, and the squad is going to suffer heavily. Once you start adding more units, the MG can't cope with the numbers. In the former case, men will get by and in the latter the MG is doomed (but the infantry will still suffer some stiff casualties). From what I can tell, this is realistic.<HR></BLOCKQUOTE>

Of course this is just flat out incorrect. The problem is how Grazing Fire is being defined. You are referring to a Beaten Zone as if it is the only thing involved. Grazing Fire is fire where the line of fire extends no higher than 1 meter off the ground. A Beaten Zone is the spot where the 'cone of fire' lands on the ground, but if you are using Grazing Fire you aren't necessarily engaging the enemy with the beaten zone, you are engaging the enemy with the danger space created by the line of fire. In other words, the Machine Gun is engaging all troops along the entire distance between the muzzle of the MG and where the beaten zone is hitting the ground. This line of fire is going to be a few meters across and will correspond to the width of the beaten zone at the far end of your line of fire. This is a fundamental method of employing MGs, and I'm sorry to say that it will not be included any time soon (perhaps the rewrite - although with the number of features put off for the rewrite they can't possibly put all of them in ) The reason it won't be included in CMBB is that the current engine does not recognize squads or tanks etc. between the weapon and the target point. This is obviously not realistic, and it is the primary means by which an MG will put effective fire on multiple targets all at once. If you remain unconvinced Major, then I can repost a paragraph or two I made (with my own commentary between quotes) of a current US Mechanized Infantry FM (well, 80's, not current anymore I guess) which defines these terms and explains the proper employment of the Machine Gun.

[Andreas]

I think BTS have done themselves no favour by the end-screen design, and look forward to changes there. Clearly the casualty figures on the screen and reality don't stack up. This does not mean that they also don't stack up in terms of casualties in the game and in reality, at least not for the same reasons.

Casualties in the game include 'battle exhaustion', and soldiers dropping by the wayside without being wounded (went to ground, bit late in getting up, squad went on, runs back to the field kitchen to wait for the return of his comrades - Mellenthin I think mentions this as a big difference between the Germans and the Soviets, who had no field kitchens around which to gather). This category is not accounted for in the end-screen. There would also be the lightly wounded, who are for the moment out of action, but back in the line after some minor attention.

Lewis has a point with the high casualties of officers/NCOs. I am not sure whether this could be helped by playing green troops instead of the regular/vet mixes we now see. I would have thought that green troops with vet HQs would be an interesting combo.

Finally, the high casualties are of course caused by the lack of 'tomorrow' in the game. Even untenable situations are being drawn out to The End, because of a believe by some players that if you surrender/withdraw you are a sissy. Shrug. Don't think the programmers can be blamed for that.