Background

Some time ago, I wrote a series of articles about the potential of computer moderated wargames. In the articles, readers were asked to comment on aspects they agreed or disagreed with. There was a large, wellinformed response and I would now like to discuss one of the main, in-depth issues that arose.

In general, readers seemed to be quite interested in the historical factors upon which the games are based. The issue that received the most attention was not technical, but historical. The issue centers on the view one takes of losses and casualties and their relationship to the combat capacity of a tactical unit in battle.

What is Meant by "Losses" and "Casualties?"

An important aspect of the computer-moderated systems I make is that they use "losses," not just "casualties" to depict the current state of a unit's combat effectiveness. In this context, the term "casualties" is used to refer only to those troops who are injured or killed. The term "losses," on the other hand, includes not only casualties, but also those troops who become ineffective for other reasons. Such troops may have their combat effectiveness negated by a number of factors, such as becoming frozen by fear, helping wounded to the rear, or simply becoming disoriented, lost or separated from their unit. Naturally, losses are much higher than casualties.

What is the central issue for Wargamers?

There seem to be two major schools of thought. One says that only actual casualties matter in determining whether a unit breaks and that reported effects of battle on unit strengths should relate to this. The other school says that casualties actually have only the most tenuous relationship to whether a unit breaks and that a host of other factors are more important. To represent the effect of these factors on the battlefield effectiveness of a unit, the best approach is to use losses.

In other words, for the gamer the central issue is: Which is the better way to represent this manpower drain on a unit's combat effectiveness -- by reflecting the level of losses or of casualties?

On the battlefield, the central issue seems to be (apart from qualitative factors such as fatigue and morale): What is the more important factor we should take into account when determining whether a unit breaks? Is it simply casualties? Or is it the total loss of eff ective manpower for a variety of other reasons? Not only wargamers discuss this issue. Many military writers go further and consider that the level of casualties to be the deciding factor in victory or defeat. I will now try to narrow down the views held by both groups more precisely.

The view taken by the popular, "Pro-Casualty Group" is:

• The level of casualties a unit takes determines its survival (the breakpoint seems to be approximately 30% casualties, according to this view).
• The level of casualties was fairly low in Ancient times, but grew rapidly as weapons became more lethal. Thus, it became an increasingly reliable indicator of a unit's battlefield effectiveness.

The opposing "Pro-Losses" view is that:

• Casualties are only one of a number of determinants of a unit's loss of effective manpower and eventual collapse. In most situations, it is the least important. Instead, the level of losses are a better indicator of combat strength.
• The level of losses is high compared to the level of casualties -- and getting higher. The level of casualties is in fact steadily reducing as time goes on, thus making it a poor indicator of a unit's battlefield effectiveness. Since the level of losses is growing it continues to be the better indicator over all gaming periods.

Why Is this Important for Computer Moderated Wargames?

This topic is of general interest for wargamers, but is particularly relevant in the context of computer moderation. This is mainly because computers make it possible to calculate things to a finer level of detail. In non-computer based games, the preferred approach seems to be to ignore the actual number of casualties and/or losses as they would be too tedious to calculate. The method usually adopted is to remove whole elements representing several hundred troops to represent a diminution of fighting strength, without making any precise reference to numbers of losses and casualties.

However, with computer moderation, such detailed calculations can be handled by the computer. Consequently, the issue of how many losses and/or casualties are suffered, and their relation to the point at which a unit breaks has to be confronted fairly and squarely. While the option does exist for the computer game developer to adopt the same broad approach as a normal paper-based set of rules, this would waste the power of the computer. It would also intrude a computer into the game process with little extra benefit and so is not really a worthwhile direction for computer moderation to take. The power of the computer has to be exploited, so the issue must be tackled.

Major Information Sources

There is a lot of information about the topic, but it is rare to find serious writers who bring it together and subject it to analysis - particularly quantitative analysis. So, I would like to refer you to the key sources I have located. This could help you search through the literature yourself if you wish to pursue the topic further. The main source I have found of use is the work of retired U.S. Colonel Dupuy. Dupuy has done more work on the quantitative basis of warfare than anyone else I have been able to locate.

Many writers and researchers describe battlefield situations using terms such as "heavy losses," but never give (even approximate) numbers. Others refer to actual numbers as an aid in giving a graphic description of the battlefield situation, but without subjecting them to solid analysis to determine their actual effect on unit effectiveness. When developing computer systems one has to be precise and use solid quantitative information where possible. Consequently, I have relied heavily on Dupuy's work to supplement the other books.

Amongst the "qualitative" authors, I have found John Keegan's works have been the single most useful source. Keegan is particularly useful in presenting the deeper cultural basis for warfare, something that is not always covered in the quantitative approach. His books The History of Warfare, The Mask of Command and The Face of Battle are particularly useful in placing a human interpretation on the quantitative model.

Are Casualties a Good Guide to Unit Strength?

When first designing games, I was a member of the "pro-casualty" school of thought, and so did a lot of reading to try to establish the level of casualties a unit suffers before it breaks. While precise quantitative data is not easy to come by, a large body of well substantiated historical data does exist. (I guess people don't stand in the middle of battle fields with calculator and note book, which is a difficulty for us wargamers.)

General Observations

Before we get onto the figures, I would like to make some general observations. I will confine these to the gunpowder period, so as to be more succinct, although the same types of observations can be made for pregunpowder period as well.

The first thing I noticed was that a number units broke in rout after receiving first fire, and not a fire that inflicted many casualties at that. Some units even broke only in anticipation of being fired upon -- without even receiving a single casualty. Clearly the number of troops hit by bullets (i.e., casualties) was only one factor. Subjective factors, such as the anticipation and perception of extreme danger, were very important, even more important than actual casualties inflicted.

It was at this point that I first started to think that perhaps the use of "losses" rather than "casualties" may in fact have something to recommend it. The number of troops who ran, froze, skulked away to the rear, or became otherwise ineffective on the battlef ield was much higher than those who were hit, and may in fact have been a much more important indicator of the battlefield effectiveness of a unit.

Reading of experiments by the British Army during the Napoleonic period reinforced this view. A six foot high sheet of white linen with the same frontage as a battalion in line of battle was placed at a certain range in front of an infantry unit. The bullet holes were counted after a volley was fired and a percentage hit rate was calculated. The results showed a hit rate of only 2% could be achieved as close as 50 yards. And that was in good, calm conditions, when firing at a harmless target, rather than a threatening enemy unit.

The actual hit rate in battle would be much lower than 2%, due to factors such as extreme stress, fatigue, dense smoke and the fouling of gun barrels after first fire. The presence of an enemy about to attack also tended to make the troops fire wildly and to cause them to use their weapons unreliably. On the other hand, the actual hit rate can be increased if the rather unpleasant facts I found in an interview with John Keegan are true. In the interview, the historian mentioned that army surgeons often removed not bullet fragments but pieces of musket stock, metal buttons and even bone and teeth rather than musket balls. This ghastly shrapnel came from comrades hit by enemy bullets. Pieces of musket and equipment as well as bone and tooth would be literally blown away at considerable velocity and would hit those nearby, apparently inflicting significant injury. (I thought you'd be interested in thatl)

A further note is that many of the casualties reported in the postbattle "Butcher's Bill" were inflicted after a unit broke. In other words, the final casualty levels tended to reflect the eff ect rather than the cause of defeat. One final observation is that many troops who had apparently been lost in the heat of battle still returned to camp afterwards, assuming there was no pursuit that captured or scattered them entirely.

Consequently I soon began to doubt that casualties were anywhere near the heart of why units break. Psychological factors seemed dominant, and casualties levels seemed to have no clear correlation with whether units stayed and fought.

Military Writers and Modelers

But what about all those military writers and modelers who use casualty levels (such as the 30% level) to determine whether a unit breaks? What is the basis for their view? Dupuy reviews this at some length, and it will suffice here to say that there is some basis. Research shows that, on average, units tend to break at about a level of 30% casualties. However, on closer investigation, the 30% average shows a very large standard deviation. That is to say, the range of casualty levels sustained by the individual units before they broke is actually very wide -- varying from over 70% casualties to almost none. In other words, there is no significant correlation between casualty levels and unit collapse points.

What Does Cause a Unit to Collapse?

In his studies on the causes of defeat, Dupuy has established the major factors that cause units to retire. This is summarized in the following table.

Maneuver by Enemy

Envelopment, Encirclement or Penetration 33%
Adjacent Friendly Unit Withdraw 13%
Enemy Occupied Key terrain 6%
Enemy Achieved Surprise 8%
Enemy Reinforced 4%
(Total Maunuever by Enemy Reasons 64%)

Firepower by Enemy

Casualties or Equipment Losses 10%
Heavy Enemy Artillery and Air Attacks 2%
(Total Firepower reasons = 12%)

Shortage of Resources

No Reserves Left 12%
Supply Shortages 2%
(Total Resource Shortage reasons = 14%)

Conditions External to the Battle

Truce or Surrender 6%
Change to Weather 2%
Orders to Withd raw 2%
(Total External Conditions reasons = 10%)

(Information drawn from Understanding Defeat by Col. T N. Dupuy.)

Clearly the factors bear little relation to casualty levels.

Part 2