TLIs and Gun Control

Well, Trevor Dupuy’s work on the Theoretical Lethality Index (TLI) that was done back in 1964 has entered into the U.S. gun control debate, not by our choice.

We discussed an earlier work that addressed this at Common Use, Lineage, and Lethality | Mystics & Statistics (dupuyinstitute.org). This first came to our attention through an article posted by CNN that generated thousands of pingbacks to our site: Opinion: Now that guns can kill hundreds in minutes, Supreme Court should rethink the rights question | CNN.

Even though I have my doubts about the utility of using the Theoretical Lethality Index for discussing gun control, I did attend and present at the conference “Current Perspectives on the History of Guns and Society” in mid-October. See: Conference: Current Perspectives on the History of Guns and Society | Mystics & Statistics (dupuyinstitute.org)

Attending this conference did lead to some useful discussions about collecting data on lethality and weapons effects in a civilian environment, similar in some respects to what I had in Chapter 15 (Casualties) in War by Numbers. This has been discussed before on this blog: Wounded-to-killed ratios in Ukraine in 2022 | Mystics & Statistics (dupuyinstitute.org) and Two proposals on Combat Casualties | Mystics & Statistics (dupuyinstitute.org). I am currently not actively trying to market an effort to further explore wounded-to-killed ratios in modern combat (although I think this is sorely needed) and I am not marketing any efforts to look at lethality in a civilian environment.

Now, there is an article contesting the original articles on the subject on the website The Volokh Conspiracy by David Kopel called “The Theoretical Lethality Index is useful for military history but not for gun control policy.” This blog post, which is rather long, is here: The Theoretical Lethality Index is useful for military history but not for gun control policy (reason.com).

Part IV of the article actually “estimates” the TLI of an assault rifle at 640. This seems a little low. It is clear that TLIs of assault rifles are 800-900 or higher, depending on the model of the rifle and how they are calculated. David Kopel’s article provides the following figures:

18th Century Flintlock: 43
1903 Springfield bolt-action magazine-fed rifle: 495
Modern AR semiautomatic rifle: 640
Modern 9mm semiautomatic handgun: 295

Not sure why they needed to “estimate” the TLI of an assault rifle (AR), as it can be calculated using the formulae in Numbers, Predictions and War. We do have lists of various TLIs for a wide variety of weapons. We do have a complete list on the DOS version of the TNDM which I am too lazy or too busy right now to get up and running. But, we did do have some old listings and spreadsheet calculations sitting around on my computer from past model validation runs. So, let me quote some figures from those efforts:

From Excel spreadsheet:
Soviet Tula-Tokarev 33 Pistol = 297.36 (7.62mm)

From WPN_LIST-WWII:

Soviet AK-47 Assault Rifle = 831.685

7.92mm FG 42 Assault Rifle: 789.823

7.92mm MP 43/StG 44 Assault Rifle: 904.045

We did check back with Chip Sayers who keeps his own listings he has calculated, and they show:

Tokarev TT-33 semi-auto 7.62mm pistol – 265

9mm Parabellum-Pistole Luger P08 – 228

9mm Walther P.38 – 229

AK 7.92mm Assault Rifle – 813

Sturmgewehr 44 – 868

Calculations vary a little from using to user depending how they determine what the practical sustained rate of fire for a weapon on a per-hour basis, maximum effective range and accuracy (often an estimation) of the weapon is (see pages 187-199 of Numbers, Predictions and War).

I guess we could go back and do the calculations for a whole range of assault rifles, but I have a lot on my plate at the moment. Certainly, someone else could do this with a little investment of time. The formulas for the TLI are public.

Anyhow, I am not going to enter this gun debate. The TLIs were designed for use in analyzing combat. While they are not directly applicable to the civilian world, they are illustrative. How relevant they are for discussions on gun control I will leave for others to argue. That is not our business.

But… there is one statement is David Kopel’s argument I must take issue with, which is “Extrapolating from the historic arms that Dupuy studies to present-day arms is questionable.”

Now, we have used the TNDM, which uses the formulae for the TLIs, as part of our effort to both analyze combat in the past and to analyze combat in the present. This construct developed in 1964 was used as part of our predictions for the Gulf War in 1991 (see: Forecasting the 1990-1991 Gulf War | Mystics & Statistics (dupuyinstitute.org) and Assessing the TNDA 1990-91 Gulf War Forecast | Mystics & Statistics (dupuyinstitute.org) and Assessing the 1990-1991 Gulf War Forecasts | Mystics & Statistics (dupuyinstitute.org). Needless to say, these predictions did better than most predictions at the time, and certainly did better the recent U.S. “intelligence communities” predictions for Afghanistan in 2021 or Ukraine in 2022. As I note: I like to claim that we are three-for-three in our predictions… | Mystics & Statistics (dupuyinstitute.org) or maybe four-for-four: Does this mean that we are four-for-four in our predictions? | Mystics & Statistics (dupuyinstitute.org).

The TNDM was also used for part of our prediction efforts on Bosnia in 1995 (see Forecasting U.S. Casualties in Bosnia | Mystics & Statistics (dupuyinstitute.org), reports B-0 and B-1 here TDI – The Dupuy Institute Publications and America’s Modern Wars) and has been used for others for a number of their own efforts in 2022 (see An Independent Effort to Use the QJM to Analyze the War in Ukraine | Mystics & Statistics (dupuyinstitute.org) and A Second Independent Effort to use the QJM/TNDM to Analyze the War in Ukraine | Mystics & Statistics (dupuyinstitute.org)). So, we are using TLIs for present day armies. We also did studies comparing proposed modern armor brigades with WWII armor divisions, which we have never blogged about (see FCS-1 and FSC-2 TDI – The Dupuy Institute Publications), although the corps and division-level model validation charts from that effort are in Chapter 19 (Validation of the TNDM) of War by Numbers and is reference here: Validating Trevor Dupuy’s Combat Models | Mystics & Statistics (dupuyinstitute.org). Some of our discussions on model validation are here: Summation of our Validation Posts | Mystics & Statistics (dupuyinstitute.org).

But, there is also Dr. Alexander Kott’s work which extrapolates weapons developments into the future, using a set of formulas similar to the TLI. This is discussed here The Evolution of Weapons and Warfare? | Mystics & Statistics (dupuyinstitute.org) and here Data Used for the ARL Paper | Mystics & Statistics (dupuyinstitute.org) and here Data Used of the ARL Paper – post 2 | Mystics & Statistics (dupuyinstitute.org) and here Technological Advancement Lessons from History? | Mystics & Statistics (dupuyinstitute.org) and here Two ARL Reports | Mystics & Statistics (dupuyinstitute.org). So, if David Kopel’s statement is correct, then the work Dr. Alexander Kott is doing at the Army Research Lab (ARL) is not valid. Dr. Kott did present his work at the Historical Analysis Annual Conference (HAAC) in late September and his briefing will be posted to this blog.