IV. Comparative Firearm and Calibre Information

1.0 Introduction

In Miami, FL in 1986, an FBI agent named Dove, while trying to apprehend a bank robbery suspect named Platt, shot Platt with a 9x19mm 115gr Winchester Silvertip. The bullet penetrated through one of the suspect's arms and into his chest.[1] It stopped in Platt's right lung, lacking sufficient penetration to reach the heart or any of the major blood vessels of the mid-chest. After being hit, the suspect killed two FBI agents, including Dove, and wounded five other agents. [2]

This infamous incident has been a catalyst for research into the performance of handgun ammunition. Unfortunately, much of this research has failed to reach the participants of rec.guns. For example, one reads of the continued recommendation of light, fast 9x19mm rounds that certainly, on average, do not penetrate enough to reach the vital structures of the chest nor of the spine, especially if one considers how the bullet might have to reach these structures after penetrating an arm or light cover.

Usually, the ones recommending such cartridges cite the problematic statistics of Marshall and Sanow, whose work seems, in many cases, to contradict the intuitions of other wound ballisticians and the recommendations of other wound ballistics models. Hopefully, the people who place so much faith in Marshall and Sanow will reconsider their positions when confronted with information from other sources.

Others in the group are fond of crowing that their particular cartridge is the best. They would have one think that no lesser cartridge than the one they choose is worthy of consideration. One might conclude that an aggressor would take his own life in abject terror when confronted with a .45ACP or a 10mm. While a larger, more powerful cartridge is desirable, there is a level of adequacy at which one is perfectly justified in feeling comfortable. It is much more desirable to hit one's target with an adequate bullet than to miss with a super bullet.

Dr. Martin L. Fackler and other wound ballistics researchers following his lead address all of these issues. Duncan MacPherson, with the help of Dr. Fackler, has developed a model that predicts bullet penetration and tissue destruction. The research of Fackler, MacPherson, and others answers many of the questions of wound ballistics. These include how handgun bullets incapacitate aggressors, how one might measure relative wounding capability, why Glasers, et al., suck, and what some of the problems are with combat data and statistics drawn from it. Even Evan Marshall seems to agree with some of their basic premises.

2.0 How Handgun Bullets Incapacitate

Handgun bullets incapacitate reliably via two methods: through damage to the central nervous system or through rapid blood loss. Breaking bones or shattering pelves do not necessarily take people out of a fight. Other mechanisms, such as "pressure effects," are unreliable at best and nonexistent, or at the very least not sufficiently proven, at worst.

As a bullet enters tissue, it creates a permanent, "crush," cavity and a temporary cavity, the latter of which is produced by cavitation. The bullet certainly destroys tissue in the permanent cavity. It damages tissue affected by the temporary cavity only if such tissue is stretched beyond its elastic limit. Unfortunately, many tissues of the body are very tolerant of such stretching. In addition, many of the vital structures that must be damaged by the bullet are deep within the body. A bullet, to be effective, must penetrate enough to reach these vital structures. The problem is compounded by the fact that many times a bullet enters the torso at an odd angle or must pass through an intermediate target, such as an arm, before reaching the torso at all. Drawing on his experience as a combat surgeon and his research into wound ballistics, Fackler observes the following:

Cavitation can be dramatic in tissue simulants, but many body tissues are flexible and elastic. In the living animal, it has been shown that much of the energy spent in forming the temporary cavity can be absorbed with little residual damage. The more elastic tissues--bowel wall, lung--and muscle suffer minimal permanent disruption compared to the relatively inelastic liver. Large blood vessels are also quite elastic and tolerate temporary cavity displacement well. Removal of a part of their wall by direct bullet impact however, is not well tolerated.

Rapid blood loss from bullet holes is the proven incapacitation mechanism from small arms. Other than disruption of the brain or cervical spinal cord, it is the only reliable mechanism available to the handgun user.

Insufficient penetration depth to reach and disrupt the body's major blood vessels is the greatest deficiency of the light--under 120 grains (7.8g) in .38 calibre--handgun bullet. The large blood vessels--aorta and vena cava--are located in the back of the abdominal cavity. Their average distance from the front of the abdomen is about 15cm. [3]

Finally, one should note that there is an important difference between handgun cartridges and powerful rifle cartridges. While the temporary cavity of a handgun bullet is not usually large enough to stretch tissues beyond their elastic limits, the temporary cavity of a rifle bullet is usually large enough to do so. [4] This is one reason why a rifle is vastly preferable to a pistol in a gunfight. Each round can be much, much more destructive.

3.0 Measuring Relative Wounding Capability

Bullet wounds fall into four categories. They are CNS wounds, vital wounds, nonvital bone fractures, and nonvital wounds. With CNS wounds, incapacitation is immediate. With vital wounds, where one hits the heart or a major blood vessel, incapacitation may take up to several minutes, but will eventually occur due to rapid blood loss despite an aggressor's psychological state. [5]

Nonvital bone fractures, while they may impair the capability of the attacker to continue, do not necessarily incapacitate. [6] For example, one might still be quite conscious and capable of returning fire even with a broken pelvis.

Obviously, a nonvital wound also does not necessarily incapacitate. However, it is reasonable to assume that, while probably not provable, the larger the mass of tissue destroyed, the more likely incapacitation is. This mass is the amount of tissue destroyed by the permanent wound cavity. [7]

The first two types of wounds, CNS wounds and vital wounds, are, given adequate penetration, a function of bullet placement. [8] MacPherson observes, "[T]he trauma resulting from opening up a major artery is relatively insensitive to the bullet that causes the breach and the wound track produced in other tissue."[9] Given good bullet placement, a .25ACP FMJ can work as well as a .45ACP Black Talon.

The third type of wound, nonvital bone fractures, is not really desirable and happens, says MacPherson, "in less than 5% of all shootings."[10] So, how well a particular round produces this type of wound should not affect one's cartridge selection.

Therefore, when selecting handgun ammunition, one is most concerned with nonvital wounds. By assumption, the more tissue a bullet destroys, the more likely it is, with a nonvital wound, to incapacitate. Given sufficient penetration to reach vital structures, one can compare bullets on the basis of how much tissue they destroy. If one bullet destroys more tissue than another and has adequate penetration, it is more desirable. Assuming enough penetration, the mass of tissue destroyed by a bullet is its measure of effectiveness. [11]

Adequate penetration is, according to Fackler, et al., about 12". [12] Since this is a minimum, one should choose cartridges capable of penetrating, say, 13" or 14" or 15" to ensure that, on average, the minimum is met. Beyond about 15", additional penetration, since it will probably be beyond vital structures, is unlikely to be useful.[13] To cause more destruction where it is useful, a bullet should expand to a larger diameter rather than penetrate beyond about 15".

Given a nonvital wound, however, it is not clear how much tissue destruction one might expect to cause incapacitation. One might wonder what level of performance one might consider adequate. For example, a .50AE may very well incapacitate more effectively than a .45ACP, but most would not want to carry a weapon capable of firing the cartridge every day.

Attempting to get a rough answer to this question, Duncan MacPherson looks at the old Thompson-LaGarde studies in which steers are deliberately shot in nonvital areas. What he finds is that, on average with popular self-defense cartridges like 9x19mm, it takes 220 grams of tissue destruction for the cattle to fall. Doing a little napkin mathematics and scaling this down from 1000lb. by a factor of 5 or 6, he speculates that it takes about 40g of tissue destruction to incapacitate a human. Note that this reasoning requires many assumptions and does not take into account the psychological factors of human aggressors. [14] This is simply a rough figure that one might use to decide whether one is carrying enough gun, to decide if one's cartridge is adequate. Since the Thompson-LaGarde Board actually recommended a .45ACP round nose cartridge, which only produces 26.8g of destruction but is putatively adequate, MacPherson posits that one should, given the level of precision of his ponderings, consider 30-40g of tissue destruction a "'level of adequacy'". [15] A table of tissue destruction for common bullet diameters is as follows:

Tissue Destruction of Common Bullet Diameters & Configurations (g) [16]

                                                9mm             10mm            .45
Configuration
-------------
JHP Maximum(13"-15" pen.)                       34              42              55
Cylinder(first 15" pen.)                        24              30              39
All Others(first 15" pen.)                      16              20              26

The level of adequacy in wound trauma does not imply immediate and certain [incapacitation] as a result on nonvital (or even vital) wounds. If wound trauma at this level of adequacy is generated without creating [incapacitation], it is likely that [incapacitation] will not occur without one of the following:

1) An added CNS wound

2) An added vital wound and the time for it to be effective

3) A large increase in [the mass of destroyed tissue] from added wounds

4) Elapse of time (possibly substantial)

There are various anecdotal stories of more than ten wounds being required to achieve [incapacitation]; this corresponds to over 200 grams of total [mass destroyed]. It is not at all practical to create this level of wound trauma with a single handgun bullet (the recoil would be at least 4 times the .45ACP). The 55 grams . . . [which is] the maximum available from the .45ACP obviously represents more trauma than the 30 to 40 grams of the level of adequacy, but this difference is not significant relative to 200 grams. This conclusion argues that all of the calibers listed . . . can have a satisfactory [ability to incapacitate] with well designed JHP bullets because they are at the level of adequacy.

. . . [S]tatistical variation in impact location effects and psychological effects can completely explain the performance variations observed in actual shootings. The observed variation may not actually be entirely due to variations in impact location and psychological effects, but it seems obvious that other factors are too small to demonstrate empirically. The present controversy over cartridge efficiency would not exist if the practical differences were not too small to determine empirically. [17]

As a final note, when considering a cartridge, one must consider how it will perform in its likely uses. For example, a highway patrolman might want to select a bullet that performs well after penetrating auto glass. A police officer in Alaska might want to make sure that the hollowpoint of his favorite bullet is not subject to clogging when shot through heavy clothing. One should test rounds in relevant scenarios and weight the various scenarios according to the importance one assigns to them. Also, one should ensure that a round achieves adequate penetration after defeating expected obstacles.

4.0 Why Glasers, et Al., Suck

The problem with Glasers and other frangible rounds is their inability to defeat intermediate obstacles, such as an arm. In addition, they often do not penetrate enough to reach vital structures. For example, when testing the .357 Magnum Glaser, Fackler finds that it only penetrates 4.7". [18] About the likely performance of a Glaser against a human aggressor, Fackler says the following:

As we were doing tests on the Glaser, one of my colleagues asked me to draw on my experience as a combat surgeon and estimate the survival time of someone shot from the front in the mid-abdomen with this bullet. My answer was, "About three days, and the cause of death would be peritonitis," since only the bowel is likely to be injured. [19]

One should note that some frangible rounds might produce adequate penetration. For example, some types of cartridges manufactured by MagSafe penetrate 12" or more. Also, MagSafe claims that certain types of their rounds defeat some kinds of intermediate cover, such as wallboard. Given that one of the selling points of frangibles has been that they do not "overpenetrate" and given the poor performance in testing of frangible rounds in the past, one should be extremely cautious before staking one's life on such exotic rounds. Perhaps there are frangibles with good performance. The performance of some of MagSafe's cartridges is very impressive. However, since the ability of such cartridges to produce wounds is so hard to model analytically, only empirical tests will give an answer.

5.0 Problems with Combat Data and Statistics Drawn from Combat Data

Combat data gets a huge amount of attention when people discuss handgun bullet performance. However, there are many problems with it. Duncan MacPherson describes some of these problems at length as follows:

It might be supposed that the complexities of analytically modeling [incapacitation by handgun bullets] could be easily avoided by merely assembling data from combat shootings and examining the results. Several individuals have attempted to compile a data base of combat shootings, but these efforts have suffered because available data is more anecdotal than scientific. This data has primarily been used as a "sanity check" on various modeling efforts, because most individuals involved know (or knew) from experience that a model cannot be directly derived from this data. In point of fact, there are clear technical reasons why a model cannot even be fully validated (much less derived) from this data; these reasons will be discussed in some detail because it is important to understand the difficulties intrinsic to this natural approach. . . . [T]hese efforts cannot be valid (despite the extravagant hyperbole that sometimes accompanies the results).

The problems with deducing [a handgun bullet incapacitation] model from combat data are principally statistical problems, and are enormously compounded by the large number of uncontrolled (and uncontrollable) variables. The practical complexities associated with the physiological and psychological factors of the individual wounded and the complexities in the details of bullet penetration make valid statistical analysis impossible, but even an idealized situation has severe statistical problems. Most individuals without a background in mathematics are surprised at the variation that pure statistical chance can introduce into an empirical result. The magnitude of the statistical variation can be illustrated by simple examples. We will use the example of a flipped coin, a scenario immeasurably simpler than a combat shooting. The coin flip has no variation due to the subject's state of mind or physical characteristics, no variation due to the chance location of the exact bullet impact location, and no variation due to assessment of borderline cases (is there incapacitation or not?).

If a coin is tossed 10 times (or 10 coins tossed once) the expected result is 5 heads and 5 tails. This expected result is more likely to occur than any other outcome of the trial. The average result obtained by performing this 10 coin experiment a great many times will be close to this expected result. This does not mean that exactly 5 heads will result from any one trial, and most laymen are surprised to learn how low the probability of getting exactly 5 heads is. There is a little less than 1 chance in 4 that exactly 5 heads will appear in any one trial, and more than 1 chance in 3 that in any single trial the number of heads will be 0, 1, 2, 3, 7, 8, 9, or 10. This means that if the probability of getting heads was unknown and was being estimated experimentally by performing this coin tossing trial, there would be more than one chance in three that the probability of tossing heads would be found to be either less than or equal to 0.3 (30%) or greater than or equal to 0.7 (70%), and less than one chance in four that the correct result (0.5 or 50%) would be obtained. This computation clearly shows that 10 samples are not nearly enough to experimentally determine the actual probability with any reasonable confidence and accuracy. The statistical variation is reduced as the number of trials is increased, but not very rapidly. If the number of tosses is increased to 30, there is about one chance in three that the number of heads will be either less than 12 (40%) or greater than 18 (60%). A much larger number of samples is required to get high confidence that an experimental result is close to the correct value.

This coin tossing example illustrates the problem with trying to assess [a handgun round's ability to incapacitate] from combat data--there isn't nearly enough data to provide statistically reliable results. The layman may think that 10 or 30 combat encounters with any one bullet and load are enough to provide statistics with great quantitative confidence (and many articles in gun magazines state or clearly assume this), but this belief is unquestionably false. This conclusion would be true even if all the combat encounters had bullet impact in the same area of the body; the fact that this data includes encounters with hits all over the body makes the statistical situation much worse. Impact location variability means that 150 cases with one bullet and load are at best something like 5 cases for each of 30 different impact locations (and really are 1 case each for 150 different locations). This impact location issue is further complicated by the very significant variation in psychological state of the person being shot. If two different cartridges are being compared, the bullet strike location and the person being hit should be exactly the same for each cartridge, but of course the data doesn't come that way.

Once in a while someone produces statistical data to support a strongly held view, and on inspection these results do not exhibit the expected statistical scatter. A well known case is the data of Cyril Burt (1883-1971), a British psychologist interested in IQ heritage; an examination of his data . . . indicates that the absence of expected statistical irregularities in Burt's data is so improbable that the data was almost certainly fabricated. In this usage "fabricated" includes "improving" data by removing irregularities or otherwise fraudulently manipulating it. This sort of thing is not common, but is by no means unheard of; the unusual factor in the Burt case was the fact that he produced data too good to be true despite his familiarity with statistics.

It is almost impossible for a layman with no knowledge of statistics to avoid the "too good to be true" trap in manufacturing or doctoring data. Consider a situation where there are 11 different bullets and loads in the same caliber, and these loads have true values (obtained in some magical way) of what we will call the "[Incapacitation] Index" of .55, .56, .57, . . ., .64, and .65 (these are numerical values assigned to [a handgun round's ability to produce incapacitation] on an arbitrary scale). If each of these loads is used in 100 combat encounters, the probability that these combat encounters will produce the correct value of the "[Incapacitation] Index" for all of these loads (to avoid jumbling up the effectiveness order) is about 1 in a trillion . . . This result is "too good to be true", or more precisely "too unlikely to be undoctored". This same situation (and the same infinitesimal probability of occurrence) applies if the "[Incapacitation] Index" was determined from one set of data and then remained in the same order for another set of data. This result also applies if the order stays fixed when the second data set is taken under slightly different conditions (e.g., a modest change in velocity due to a change in the barrel length of the gun).

There is another aspect of combat encounters that compounds the data gathering problem. Most combats are not single shot affairs because each combatant keeps firing at his adversary until it is clear that the adversary is incapacitated (at least that is his approach if he has any sense at all). For this reason the person shot usually has multiple wounds, and assessing the relative contribution of each to the incapacitation is often ambiguous or controversial. Thus, even well documented encounters may not give any useful data about "one shot" bullet effectiveness.

Statistical variations in data from combat encounters limit the practical use of this data to guidance in assessing the general plausibility of [estimates of a handgun round's ability to produce incapacitation], and that is the way it has been used by the most responsible investigators. Any claim that [a handgun round's ability to produce incapacitation] can be assessed within a few percent based on combat shooting data is based on ignorance, or fraud, or both.[20]

6.0 Even Evan Marshall Seems to Agree with Some of Their Basic Premises

In his report for a 1987 FBI conference on wound ballistics, Evan Marshall seems to agree with some of the basic points Fackler and others make about the way in which handgun bullets incapacitate and the need for adequate bullet penetration. Showing an understanding of how handgun bullets incapacitate, Marshall observes, "The only sure stop with a handgun round regardless of caliber or design is a central nervous system hit." [21] About the need for adequate penetration, he says, "We need sufficient penetration with handgun bullets to reach the major organs and vessels that, when damaged, will produce relatively rapid incapacitation." [22] It is unclear why he would let his work on "one shot stops" be used to argue for cartridges that do not produce good penetration.

Perhaps his work would be more valuable if, instead of only giving the simple ratio of "one shot stop" successes to the number of observed incidents from combat data, he provided detailed case studies with accompanying analyses, showing how bullets worked well in particular cases and not well in others. For example, he could say, "While many criticize frangible bullets for their inability to penetrate arms, etc., the .357 MagSafe shows that it can do so in case X." He could attempt to argue about the successes and failures of given cartridges using examples from real world data, of which he should have an abundance, and not merely the statistics that many criticize.

End Notes