Category Casualty estimation

Body Counts – What Can They Tell Us?

A new posting from William (Chip) Sayers. This is his tenth post here. He will be presenting at our Historical Analysis conference: Who’s Who at HAAC – part 1 | Mystics & Statistics (dupuyinstitute.org) and Schedule of the Historical Analysis Annual Conference (HAAC), 27-29 September 2022 – update 13 | Mystics & Statistics (dupuyinstitute.org).

——————-William (Chip) Sayers————————–

Body Counts – What Can They Tell Us?

There has been a lot of back and forth on the web lately about how many casualties Russia has taken in Ukraine, how much equipment they’ve lost and how they are going to replace those troops and weapons. I’ve decided it’s time to look into this subject and bring my own experiences to the table. 

In Vietnam, these kind of metrics seemed to have no utility other than to corrupt the officer corps with fake and exaggerated body counts, or worse, counting dead civilians as combatants. Or so the legends say. Gen. Westmorland demanded a high body count, so that’s what he got. And while the numbers looked impressive, it eventually became apparent that the numbers weren’t tied to anything concrete. Would another 50,000 North Vietnamese dead cause Ho Chi Minh to capitulate? Another 100,000? In the end, Hanoi admitted they had lost a million men in their bid to take over the South. Clearly, Uncle Ho was ready to fight to the last North Vietnamese soldier. When your opponent’s pain threshold is that high, the body count really doesn’t tell you anything. Or so it would seem.

In the aftermath of Operation DESERT STORM in 1991, various interested parties were pushing the Pentagon for a casualty count on the Iraqi side. I’m sure some wanted to thump their chests over the “kill ratio” between Iraqi casualties and the ludicrously low figures on the Allied side; while others apparently wanted to use the presumably large number of Iraqi losses to highlight the wanton brutality of the campaign we had pursued. The obvious source, Gen. Norman Schwarzkopf, commander of Coalition Forces, wasn’t interested in sharing numbers. Schwarzkopf was a field-grade infantry officer during the Vietnam War and he had seen the dark side of body counts and wanted no part of them. While there would be little of the corrosive effect on the officer corps after a short, victorious war, he also knew how badly the US Army’s credibility had suffered over the issue in Vietnam. There would be no body count from Operation DESERT STORM.

The question, then, got punted over to Defense Intelligence Agency where I was working at the time. I know the analyst who had to respond and I know that he basically made up a number out of whole cloth to get the action off his desk during a very busy time. If I recall correctly, that number was 100,000, with no distinction between killed or wounded, or if they were all KIA, as reported by some in the media. I had Col Dupuy’s model in hand and had used it to support my team’s effort during the war, so I could have come up with a far more reasonable estimate, but no one asked me. Some months later, another analyst I knew wrote an article for Foreign Affairs on the Iraqi casualty account. He rightly derided the 100,000 estimate. However, he made his estimate by counting vehicles destroyed and multiplying by crew capacity—certainly more reasonable than a complete WAG, but unlikely to be very accurate—particularly as most destroyed Iraqi vehicles had already been abandoned by their crews. So far as I’m aware, no one has ever done a serious, forensic study of the question since then and basically, nothing has been learned or understood about Iraqi casualties from ODS.

Flash forward 12 years: A few months after the invasion of Iraq in 2003, a couple of colleagues and I were asked to do a study of the nascent Sunni insurgency and come up with tips the U.S. Army could use to protect its personnel from ambushes. For my part, I went back to a series of “How we did it” monographs the Army put out in 1971-1974 in that brief period of time when we were looking at our efforts in Vietnam as a victory. After 1975, no one in the Army wanted to hear anything about Vietnam, and this series of monographs went into dusty archives, presumably never to see the light of day. In large measure, I was simply feeding the Army its own forgotten history, and they were very happy to receive it. 

Over the course of the project, I created a database from the Army’s information about various incidents they were involved in. In particular, I was interested in incidents which resulted in casualties to either side. For the U.S. side, I counted those personnel who were listed as KIA and those who were listed as Seriously Wounded in Action, or SWIA. I considered those personnel who did not return to duty within 72 hours as being WIA. I did not count those soldiers who put a band-aid on it and returned to duty within that 72-hour window. To count as an insurgent casualty, I only included those whose body was in our possession. i.e., enemy KIA where we actually had possession of the body and those who were captured and in custody. Some of the Enemy Prisoners of War (EPWs) may have been wounded, but there was no information with which to break that out. Any insurgent who was believed to have been killed or wounded but was not in our possession was not counted.

Eventually, I ended up with a database of some 27,000 incidents evaluated for 80 different conditions over a period of a year and a half of operations. Some of the conditions I tested for included things like what kind of attack it was: small arms, Improvised Explosive Device (IED), mortar, complex, etc. The Army counted an attack as “complex” if two different types of weapons were used, e.g., small arms and Rocket Propelled Grenades (RPGs). I believed the Army’s definition was not helpful, as the purpose of that categorization was to indicate a higher level of training and tactical competence. So, I only counted an attack as being in the complex category that actually required integration of differing elements, such as small arms and mortars or a vehicular IED and coordinated sniper fire, etc.

In analyzing these incidents, certain trends seemed to fall out. The Iraqi insurgents were never able to successfully stand up to Allied troops in an infantry fight. Insurgent mortarmen were hopelessly inept. Even IED attacks rapidly dropped in effectiveness. Some of these things were due to the quick adaptation by Allied troops. They got better defensive armor and tactics, but experience was the key. It was difficult to watch Allied casualties spike with the rotation of a new unit into combat, but this was followed rapidly by a drop in losses. Often, the final rate was lower than that of the previous unit, indicating an effective process of disseminating lessons learned through the force. In contrast, there was no sign whatsoever of a similar process on the insurgent side. Their casualty rates increased incessantly to the point that something had to give. That something was the “Sunni Awakening,” when they realized that it would be better to seek our protection from the Shia militias than to fight us.

One of the things that convinced me that this result was inevitable was what happened with casualty rates on both sides with regard to IEDs. In the first months of the insurgency, it took only three IED emplacements to cause a U.S. KIA or SWIA. By the end of my study, this was up to over 30 to 1 and the insurgents were actually losing more personnel involved with IEDs than we were (not all of insurgent losses were IED trigger-men scooped up during an attack, or from emplacement teams who were caught in the act—some were from raids on IED factories, or from the movement of bombs where careless insurgents blew themselves up, etc.). Much of this data was collected before our realization of the extent our troops were suffering from traumatic brain injuries, so the real cost to our personnel was not completely captured. However, the enemy did not know this, either, so it didn’t influence their decision to change sides.

One of the most important findings I made was of a major inflection point in insurgent combat effectiveness in November, 2004. It appeared quite clear that many of their best people—if not the very heart of the insurgency—died in the fight for Fallujah. This should come as a surprise to no one, given the fight they put up, and the lack of escape routes we afforded them. From that point on, insurgent effectiveness dropped and continued a literal death spiral as moderately experienced insurgents were killed or captured only to be replaced by personnel with less experience and a lower life expectancy on the battlefield.

Ironically, the insurgency didn’t recognize what was happening and intensified its attacks. While the burnout of the Sunni insurgency was predictable by the end of 2005, it took another year for it to fully manifest itself. In the meantime, the increasing numbers of attacks—resulting in increasing Allied casualties, despite plummeting insurgent effectiveness—caused U.S. analysts to believe Iraq was a lost cause when, in fact, Allied forces were on the cusp of victory. Eventually, Sunni insurgents were forced to face the inevitable truth that that they could not win in a two-front war against both us and the Shia militias and to continue was tantamount to sect suicide. They knew that they would never be able to return to their position of domination over Iraq and would suffer mightily at the hands of the Shia they had previously oppressed. So, they took the only reasonable option available: they sought our protection from the Shia majority. This startling outcome was entirely predictable, if one paid close attention to the data.

A further myth I was able to disprove was the threat of “bleed out.” Counter-terrorism analysts were highly concerned that trained Iraqi military personnel turned insurgents would begin to exit Iraq with their skills and perpetrate acts of terrorism around the world. However, their bombmaking skills had no application anywhere outside of Iraq as it was almost entirely based on the use of artillery shells—not something you would find lying around Western Europe. Their small arms skills were almost non-existent, the vast majority of their “sniper” attacks were almost certainly lucky shots (I was able to identify the work of only one or two actual snipers from the data), fewer than one in ten of their RPG attacks hit anything, and they seemed completely baffled by the mortars they employed, often failing to put rounds inside the fence at a sprawling logistics base or walking their rounds off a target they came close to hitting with the first shot. Either their former soldiers didn’t join the insurgency, or they were completely incompetent. Either way, the “bleed out” threat never materialized.

Certainly, I would never advocate for a simple body count. If you ask for bodies, that’s exactly what you will get. And as we saw in Vietnam, that sometimes became problematic. However, the careful collection and analysis of combat loss statistics can be of great value.

To bring this back home, can body counts and other loss statistics be of value in analyzing the war in Ukraine? Yes, but with this caveat: those of us operating in the unclassified realm have very little access to good, useful information necessary for such analysis. Furthermore, the Ukrainian government not only recognizes the utility of good propaganda, they are masterful in its employment. And that’s a complement. However, it makes our job more difficult in sorting out truth from fiction. 

One final vignette: When the U.S. gave Stinger missiles to the Mujahideen insurgents to fight the Soviets in Afghanistan, a coworker of mine was the point-man on watching the results. After a week of successful action, he wrote a piece saying that the Soviets had lost seven aircraft, or about one per day. That evening, Dan Rather read his report, verbatim on the CBS evening news, and in the succeeding months, the media and other analysts based their estimates of Russian aircraft losses on the “one per day” comment. In fact, you can find claims of Soviet losses to this day, based on that report. In reality, the Soviets took quick action, lowering their losses significantly, albeit at a high cost to the effectiveness of their Air Force. However, I have often wondered if perhaps Mr. Gorbachev trusted the Western news media more than his own generals on this issue (a not entirely unreasonable position), and eventually came to the determination that Afghanistan wasn’t worth it based on a little piece of unintentional propaganda.

—————————————–

 

My comment: Chip Sayer’s postings are completely independent of The Dupuy Institute. He emailed this to me last week and I did not get around to reading it until this morning, as I was copying and pasting it to the blog. There is a lot of significant statements in this posting, which I was tempted to place in bold. Some of these reinforce statements I have made in my books, in particular Modern American Wars. This is definitely a blog post worth reading slowly twice.

The Severskii Donets River Crossing Operation

This is another piece of TNDM analysis from William (Chip) Sayers. He will be doing a presentation at our Historical Analysis conference: Who’s Who at HAAC – part 1 | Mystics & Statistics (dupuyinstitute.org). Now, I have questioned the Ukrainian estimates of overall casualties: The Ukrainian casualty claims are inflated – part 1 | Mystics & Statistics (dupuyinstitute.org). This post by Chip Sayers makes the point that the estimates of 400+ Russian casualties in this river crossing operation is believable. 

The William (Chip) Sayers piece:

The objective of this set of model runs was to explore the failed Russian river crossing operation of the Siverskiy Donets.  Ukrainian photos of the battlefield aftermath showed vividly the carnage wrought by modern defenses when properly warned and deployed. 

Ukrainian reconnaissance detected the Russian preparations to cross the Siverskiy Donets at least 3 days prior to the river crossing operation, which gave them time to bring up reinforcements and prepare a defense.  According to uawardata.com, The 79th Ukrainian Air Assault Brigade was defending the sector, while one or two Battalion Tactical Groups from the 35th and 74th Mechanized Brigades of the 41st Combined Arms Army were in the area of the attack.  The Institute for the Study of War reports that it was a BTG from the 74th Mech Brigade that conducted the attack.

The vulnerability of attacking forces in a river crossing operation is such that surprise and support are necessary to preclude catastrophic losses.  Clearly, the Russian BTG failed to achieve surprise — it is unknown how well supported they were, but the scale of their losses indicates that whatever support they had was probably inadequate.

I set up the battle with the following forces:

Russian: one mechanized BTG from the 74th Mech Brigade with support from half the guns of the 120th Artillery Brigade (the 41st CAA’s sole artillery brigade) and 30 Mi-24 attack helicopter sorties.

Ukrainian: one battalion from the 79th Air Assault Brigade plus a battalion slice of brigade assets (the brigade apparently was covering a large frontage and probably couldn’t afford to devote more to the battle).  Assuming the Ukrainian Army productively used the warning time they had to bring up supporting fires, I gave the defenders an Urgan MRL battalion from the 27th Rocket Artillery Brigade.  Just for fun, I also gave them ~10% of the Switchblade “kamikaze” UAS that the United States has sent (100 x Switchblade 300 anti-personnel drones and 20 x Switchblade 600 anti-tank drones). 

The Siverskiy Donets is only about 50m wide at the crossing point, and the ground is heavily forested, though relatively flat.  As best as can be determined from aerial video imagery, the battlefield is actually too small to support an attacking battalion.  Soviet motorized rifle battalions attacked on a minimum frontage of 1 kilometer, while a Russian BTG is somewhat beefier organization.  The Siverskiy Donets makes a partial loop at the point of attack and a kilometer attack zone would have taken the entire frontage of the loop such that Russian troops on one side might have been firing into the faces of their compatriots attacking at the other end of the line.  This suggests that they probably attacked on no more than a single company frontage (400m), with four companies in echelon.  This is not a formula for success in an opposed river crossing.

For the base case, I did not give the Ukrainian side a CEV advantage, nor surprise to either side.

Results:

After a full day of battle on 11 May, the Russians were decisively repulsed, having lost 27% of their personnel, 80% of their armored fighting vehicles, 50% of their attack helicopters and 38% of their overall combat power.  This compares favorably to the ISW’s estimate of 458 casualties (the TNDM predicted 521) and 80 pieces of equipment, compared to the TNDM report of 89.  50% losses to the attack helicopter force is probably a bit overstated and could bear closer examination.

In contrast, the Ukrainians lost 6% of personnel (85 men), 2 AFVs and 6% of their combat power. 

This disparity in the loss of combat power to the two sides would have undoubtedly made another attempt to cross the river on the 12th a non-starter, even though some sources have reported it as a two-day battle.  If these losses are indeed indicative of what really happened, the second day’s attack could only have been carried out by the 35th Brigade’s BTG.  This would imply that not one, but two BTGs may have been wrecked in the Siverskiy Donets operation.

Run #2: While the force-to-space ratios don’t really support at two-BTG attack, it is possible that when the 74th Mech’s BTG was rendered combat-ineffective, the 35th’s BTG was passed through the 74th to renew the attack.  By including both BTG’s in the Russian Order of Battle and throwing in the balance of the 120th Artillery Brigade, we can see the impact of a rough 2:1 numerical advantage for the attacker’s side.

As it happens, even this reinforcement only yielded a combat power ratio of 1:1, which did not allow for a successful crossing of the river.  Russian casualties were somewhat lower, but still higher than the defender’s losses.  Both sides took about 7% personnel losses, but the Russian units had roughly double the personnel, so took 271 casualties as opposed to only 110 by the Ukrainians.  The Russians lost 29 AFVs to the Ukrainians’ 3, but this is deceptive as the defenders had only a small number of tanks to lose — their 3 AFV losses represented 45% of their starting force.  The bigger story is that the Russian’s lost 15% of their combat power, while the Ukrainians lost just 8%, thus ensuring the attackers would never make it across the river without significant reinforcement.

Run #3: Some sites have reported that the operation continued into a second day, so the third run was essentially the first run with a second day’s extension where the 35th Brigade’s BTG took over the fight.  I also postulated that it would be too expensive for the Ukrainians to devote a second package of Switchblade drones to battle.  Therefore, the defenders’ combat power was weakened in comparison to the first day’s battle, despite their low losses.

After extracting the losses to the defenders and not adding the Switchblade package back in, the fresh 35th Bde BTG replaced the broken 74th’s BTG with exactly the same combat power.  Thus, we have a fresh second echelon committed to battle against a moderately depleted defender. 

Results:

Without their Switchblade drones, the 79th Air Assault Brigade defenders were unable to repulse the second echelon, though the attack was only able to make 350 meters beyond the river’s edge.  In doing so, the Russians lost another 323 men, a third of their AFVs and 14% of their combat power.  The defenders lost 130 men, 2 AFVs and 1% of their combat power.  All told, in the two-day battle the Russians lost nearly 850 men, 45 AFVs, had one BTG rendered combat ineffective and barely gained a toehold on the south side of the Siverskiy Donets. 

Run #4: Clearly, a CEV advantage to the Ukrainian side would merely make the Russian defeats in Runs #1 and #2 worse, so they needn’t be explored further.  Run number 4 was run as run #3, but with a Ukrainian CEV of 1.2.

Results:

On day 1, the Russians lost 57% of their combat power and 124% of their AFVs, indicating they their armored vehicles wouldn’t have survived to the end of the day, thus the battle would have likely ceased less than 16 hours in.  After such a slaughter, it is questionable if the Russians would have tried it again the next day with the second BTG.

On day 2, the Russian second echelon managed to take the crossing site to the same depth as run #3, but lost 50% of their AFVs and 20% of their combat power in doing so.  The casualty list after two days would have totaled over a thousand Russian soldiers.  Technically, this might be called a Russian “victory,” but without a substantial exploitation force behind them, it would surely be a pyrrhic one.

Run #5: The next few runs explored the idea that the Russians not only failed to achieve the needed surprise, but were, in fact, surprised by the resistance they ran into.  Again, runs #1 & #2 were sufficiently adverse to the attacker that they needn’t be explored further.  This is a re-run of run #3 with minor surprise by the defenders.

Results:

While the 74th’s BTG was thoroughly destroyed (41% personnel losses, 161% of AFVs and 71% of its combat power lost), interestingly, the 35th’s BTG made the same gains and losses were almost exactly what they were in run #4.  This surprising result is because surprise degrades over time, and the second day of minor surprise essentially equates to the Ukrainian CEV of 1.2 from run #4.

Run #6: Given this, run #6 was done with substantial surprise.  Not surprisingly, the 74th BTG was obliterated in short order, losing over 54% of its personnel, 271% of its AFVs and 115% of its combat power.  It could be scratched out of the Russian OOB after half a day’s fight.  The 35th BTG, however, still made it across the river, albeit at some cost.

Run #7: Run 7 was done with complete surprise — not a likely situation, but one explored for completeness.  The results were much the same, with the 74th annihilated after just a few hours, while the 35th persisted in gaining the far bank of the Siverskiy Donets.  However, at a cost of 1/3rd of its personnel, all of its AFVs and 35% of its combat power, the BTG would have been hors de combat until a rest and refit period were accomplished.

At this point, I ran several sensitivity exercises without using the Switchblade package from the first day of the battle, but including variations of surprise and CEV.  No amount of surprise was capable of repulsing the attack and only a Ukrainian CEV of 1.5 proved sufficient to keep the Russians on their side of the river without the use of the kamikaze UAVs.  As a large influx of combat power in some form was needed to rebuff the attack, I made a final run with the entire 79th Air Assault Brigade in place of the Switchblades.  This did, indeed, defeat the single BTG on day 1 of the battle with losses comparable to those estimated to have actually been incurred by the Russian force.  In other words, it took a force three times the size of our estimate to successfully defeat the river crossing without extraordinary support.  This would, of course, have been advantageous for defending a day 2 attack by the 35th BTG had it occurred, since the entire brigade would remain in place, as opposed to the Switchblade package that we posited would be too expensive to recreate for the second day’s battle.

Conclusions:

Once again, the TNDM has demonstrated that the reported results of this battle are entirely consistent with historical outcomes — nothing too unusual or particularly remarkable happened here.

Also, Russian tactical incompetence is once again the key to interpreting the results of this battle.  Had the Russians achieved the prerequisite surprise, things would have gone very differently.  A final model run with the shoe on the other foot confirms this.

Finally, the impact of a single weapon system — in this case the Switchblade kamikaze UAS — proved pivotal.  Essentially a round of ammunition, these killer drones significantly boosted the combat power of the defenders.  However, once expended, the defending units went back to their organic firepower, which was insufficient to hold the line.  This underscores the importance of a steady, sustainable supply of weapons and logistical support from NATO and other countries sympathetic to the Ukrainian cause.

Two proposals on Combat Casualties

In my last post I referenced two proposals I prepared for the CCCRP (Combat Casualty Care Research Program) on combat casualties. This was built off Chapter 15 “Casualties” in my book War by Numbers and some combat casualty modeling work I had done at the Marine Corps Warfighting Lab (MCWL). The working level of CCCRP was very supportive of this effort, but their proposal process went through a group of reviewers, and in each case, one reviewer saw no reason to do this work. As I think the reviewers were medical doctors, and the proposals were not directly related to casualty care, it is not surprising that a couple of them would feel that way. It obviously did not fit any of their immediate needs or concerns. So, this effort died in the fall of 2020.

The two proposals were:

  1. Combat Analysis Proposal of three tasks:
    1. Ratios for operations: A brief exploration of existing sources to determine what the expected wounded-to-killed ratios are for different types of operations. This would include a projection for 1) modern (post-World War II) conventional combat scenarios, 2) modern counter-insurgencies, 3) modern training and assistance missions, and 4) special operations and other active assistance programs. The goal would be to assemble a set of figures that can then be applied to combat model outputs, planning, and other studies and analysis.
    2. Ratios by Posture: Conduct a detailed analysis of division-level conventional combat to determine the wounded-to-killed ratios dependent on outcome and posture. This will be a comparative analysis based upon the coded outcomes of the engagements in the database and whether the force is conducting an offensive, defensive or other type of mission. The goal is to determine if wounded-to-kill ratios vary in these conditions (which they certainly do) and to determine to what extent they do and what are those values.
    3. Lethality of IEDs: Develop a set of values for the lethality of modern IEDs (Improvised Explosive Devices), booby traps, mines, claymores, truck bombs, suicide bombers, and other such devices. The lethality figures will have to be developed from a systematic review of a variety of combat and medical reports. Ideally, this data can be assembled from the experiences in Iraq and Afghanistan, but research will also examine various operations in the Middle East, Northern Ireland, Vietnam and other post-WWII insurgencies. This would also include a review of the Wound Data and Munitions Effectiveness Team (WDMET) data from the Vietnam War and the ongoing Joint Theater Trauma Registry (JTTR).
  2. Combat Modeling Proposal of three tasks:
    1. Lethality Rates: A brief exploration of existing sources to determine what the lethality rates are for various weapons. This effort will tap existing open sources and is an extension of the work outlined in Chapter 15, “Casualties” in the book War by Numbers.
    2. How Models Measure Casualties: Conduct a review of around a half-dozen active U.S. Army and DOD conventional combat models and determine how they measure casualties and how they determine wounded versus killed (and missing). This review will probably include meetings with the modelers, in additional to conducting an analysis for the differences between their various approaches.
    3. Medical Deployment Options for Conventional Operations: Conduct a comparative simulation of medical care, evacuation and return-to-duties over an extended conventional warfare exercise. This would be an examination of the optimal medical support structure of the sake of sustaining combat power over an extended conventional deployment.
      1. The three recommended scenarios are 1) Emergency deployment to the Baltic States, 2) intervention in collapsed state of North Korea, and 3) the validation scenario for Iraq in 2006. Other scenarios may be added time permitting, or as a possible future effort.

This was the last time I attempted to market DOD. It took some man-months of effort to assemble and submit these proposals along with all the required supporting documentation. There is a trade-off of whether my time is better spent looking for the next contract or writing my next book. I decided after this experience to focus on writing my next books, which is why I now have two completed books going through editing and am about to contract for two more books. It is difficult for a small company to work for the DOD. The number of hoops and time consumed in marketing efforts are the same whether it is Lockheed Martin or TDI. Anyhow, if CCCRP would not approve such an effort, then I do not know who in DOD would. As it is, I gave up trying to further market it and prioritized my time to other projects.

Wounded-to-killed ratios in Ukraine in 2022

I made a post on Friday that unsurprisingly got some pushback on twitter. It is here: The Ukrainian casualty claims are inflated – part 1 | Mystics & Statistics (dupuyinstitute.org).

One of the twitter complaints is that I used a wounded-to-killed ratio of 4-to-1 for my challenge of the Ukrainian kill claims. This was based upon the only set of aggregate wounded-to-killed statistics offered in the conflict, the reports from the Donets Peoples Republic (DPR) who are reporting as of 26 May that they had 1,912 soldiers killed and 7,919 wounded (a 4.14-to-1 wounded-to-killed ratio). 

This also generated another twitter criticism because I was using DPR data. Yet, this last week Zelensky has stated that they are losing between 60 – 100 Ukrainian soldiers a week. An additional 500 are wounded. This is at least 5-to-1 wounded to killed ratio (or 8.33-to-1). 

Apparently, some other people have been using a 3-to-1 wounded to killed ratio (not that this lower figure changes my argument one iota). What is the basis for people using 3-to-1? WWII data? Lack of artillery?

It almost certainly is a traditional figure drawn from Trevor Dupuy or WWII data or convention. Now, wounded-to-kill figures vary by 1) definition, 2) posture, 3) medical care, 4) how Died-of-Wounds (DOW) are counted, 5) mix of wounding agents, 6) number MIA or captured, 7) use of body armor and 8) the nature of the targets (airplanes or tanks compared to infantry). This is discussed in depth in Chapter 15 of my book War by Numbers.

Let my post a table from my book:

German

                                                                        U.S. Army      UK Army       Army[1]

Ardennes from 16 – 23 December                  4.85-to-1         1.67-to-1         3.25-to-1

Ardennes from 24 December – 1 January      5.65-to-1         2.89-to-1         3.08-to-1

Ardennes from 2 – 16 January                        5.12-to-1         4.59-to-1         2.99-to-1

 

                                                                        German Army[2]          Soviet Army[3]

Kursk from 4 – 11 July                                   5.11-to-1                     2.29-to-1

Kursk from 12 – 18 July                                 4.54-to-1                     2.68-to-1

See my book for the footnotes.

So, 3-to-1 or greater for WWII data. The Soviet Army at Kursk, when on the defensive and having divisions being overrun with large numbers of MIA and captured had lower than 3-to-1. But, when the Soviet missing was less than 10% of the casualties, the wounded to killed ratio rose to 3-to-1. This is discussed in more depth in my book along with supporting data. Anyhow, 3-to-1 is kind of the minimal figure to be used for most ground combat in WWII and as you can see from the table above, it is sometimes higher. 

The wounded-to-killed ratios since WWII have been higher. In some cases much higher (like 13-to-1 for the U.S. Marine Corps in Afghanistan). All this is discussed in depth in my book. 

From a practical point of view, I do not know what the wounded-to-killed ratio will be for the fighting in Ukraine. I expect it to be more than 3-to-1. This is the old WWII figure and medical care has improved since then and many people are now wearing body armor. Body armor certainly increases survivability from blast and fragmentation wounding, which is the majority of wounding on most battlefields.

Therefore, the DPR wounded-to-killed figures of 4-to-1 looked reasonable to me. They may be low, depending on how they are counting wounded (not all lightly wounded are counted as wounded). Zelensky’s off-the-cuff figure of 5-to-1 (or higher) also seems reasonable. 

On other hand, for those using a 3-to-1 figure, then I have to ask what is their basis for using this figure (WWII data or tradition?) and what is their basis for not revising it to reflect better medical treatment improvements over the last 75 years and the rather extensive use of body armor now?

Now, as this war continues, I expect the wounded-to-killed ratio to actually increase. This is because both sides are making more use of artillery and more of the engagements are using dismounted infantry. This changes the mix of causative agents and targets in a manner that should result in more people wounded per person killed (or to express it another way, less mortality per casualty). 

 

 

P.S. Final note: in the fall of 2020 I was hoping to obtain contracts based upon two proposals submitted to CCCRP (Combat Casualty Care Research Program) to further examine wounded-to-killed ratios and the analysis of combat casualties, including a combat casualty modeling effort. This effort died because some reviewers did not feel a strong need to further examine this subject. I still feel this subject needs to be fully explored. Tempted to do a further blog post on this. To date, my book is the most complete discussion on the subject that I am aware of.

The Ukrainian casualty claims are inflated – part 1

Just to state the obvious, the Ukrainian Army claims of Russian casualties are inflated. Still, I find people on twitter who are repeating the Ukrainian claims like they are fact, including people with significant followings (100K+). They are even doing some estimates of Russian dead per day, based upon this questionable data. They should know better.

Let us look for a moment at the Ukrainian Army claims as of early 3 June of 30,950 Russian soldiers killed. Now, the wounded-to-killed ratio for operations in this war looks to be around 4-to-1. This is based on the reporting of the Donets Peoples Republic, which oddly enough, has been consistently and steadily reporting their losses. Also see: Wounded-To-Killed Ratios | Mystics & Statistics (dupuyinstitute.org) and Chapter 15 of my book War by Numbers.

So 30,950 x 5 = 154,750 casualties (not counting captured and missing-in-action). This is out of a Russian force in Ukraine estimated to be 150,000 to 190,000. Something does not match up here.

The Russians are advancing, so they are either 1) advancing against a Ukrainian army of 200,000+ with the few remaining tens of thousands of troops they have, or 2) They have deployed over 100,000 brand new replacements (for which there is no evidence) who are advancing against this motivated and experienced Ukrainian army, or 3) the Ukrainian estimates are high. The latter seems the most logical explanation.

It does appear that the Ukrainian claims are off by a factor of 2 to 4. So, maybe 7,738 to 15,475 killed if we divide the Ukrainian estimates by 4 or 2. The NYT claimed that as of 19 April U.S. intelligence estimates were between 7,000 to 10,000 Russians killed. This seems like a more reasonable estimate.

The same NYT report on U.S. intelligence estimates were that Ukrainians had suffered 5,500 to 11,000 killed as of 19 April.

Vincent B. Hawkins has passed away

I just heard that Vincent Hawkins, former HERO and DMSI employee, passed away yesterday. Born in August 1955 he was 66 years old.

He came to work for Trevor Dupuy’s HERO/DMSI around 1986 and was there when I arrived. Always outgoing and welcoming, he had a master’s degree in history, was fluent in German and led the German research team for the Ardennes Campaign Simulation Data Base (ACSDB). He was the co-author of one chapter in Trevor Dupuy’s Understanding Defeat.

He continued at DMSI until around 1991, when it closed down. He was the editor for weekly defense newsletter they were doing, which I thought was a worthwhile effort. He then did some historical research for George Kuhn’s casualty estimation efforts at LMI.

Always outgoing and humorous, Vince was a good friend at DMSI. His office was always one of my stops when I took a break for coffee.

I like to claim that we are three-for-three in our predictions…

We have discussed on this blog before our analysis for the Gulf War, the casualty estimate for Bosnia peacekeeping mission, and the casualty and duration estimate for Iraq. In each case, we were either the closest public estimate or pretty much dead on. Just as a reminder:

Predictions | Mystics & Statistics (dupuyinstitute.org)

Now, back in late 1990 Trevor Dupuy made his predictions on the Gulf War. They are discussed here:

Forecasting the 1990-1991 Gulf War | Mystics & Statistics (dupuyinstitute.org)

Assessing the TNDA 1990-91 Gulf War Forecast | Mystics & Statistics (dupuyinstitute.org)

Assessing the 1990-1991 Gulf War Forecasts | Mystics & Statistics (dupuyinstitute.org)

In 1995 we provided the Joint Chiefs of Staff (JCS) with our predictions for Bosnia. This was, as far as I know, the first formal attempt to make a prediction of casualties for an “operation other than war.” This prediction is in Appendix II of America’s Modern Wars and is discussed here:

Forecasting U.S. Casualties in Bosnia | Mystics & Statistics (dupuyinstitute.org)

In 2004 we provided the Center for Army Analysis (CAA) and OSD Net Assessment our predictions of casualties and duration for the war in Iraq. Again, as far as I know, this was the first formal attempt to make an analytically based prediction on casualties and duration for a insurgency. This prediction is discussed in depth in Chapter 1 and Appendix I for America’s Modern Wars and is discussed here.

Forecasting the Iraqi Insurgency | Mystics & Statistics (dupuyinstitute.org)

And then there are these posts:

President Obama’s Casualty Estimates | Mystics & Statistics (dupuyinstitute.org)

Casualty Estimates for a War with North Korea | Mystics & Statistics (dupuyinstitute.org)

The CRS Casualty Estimates | Mystics & Statistics (dupuyinstitute.org)

Casualty Estimates for Conflict with Iran | Mystics & Statistics (dupuyinstitute.org)

Casualty Estimates for Conflict with Iran – Killed and Wounded | Mystics & Statistics (dupuyinstitute.org)

How Common are Casualty Estimates? | Mystics & Statistics (dupuyinstitute.org)

Casualty Estimates for Conflict with Iran – Summation | Mystics & Statistics (dupuyinstitute.org)

I always like to claim that we are three-for-three, in that we have published three predictions before conflicts occur that are fundamentally correct. As significant, in my mind, is that we were correct, based upon historical analysis and using combat models build upon history for not only a conventional war, but for an unconventional or guerilla war and for a peacekeeping mission. This is a wide range of scenarios. We are not aware of anyone else who has done this.

 

Beyond Lanchester

The publication of the book Beyond Lanchester last year had escaped me. See Beyond Lanchester: Stochastic Granular Attrition Combat Processes

His blurb on the book:

F.W. Lanchester famously reduced the mutual erosion of attrition warfare to simple mathematical form, resulting in his famous “Square Law,” and also the “Linear Law.” Followers have sought to fit real-world data to Lanchester’s equations, and/or to elaborate them in order to capture more aspects of reality. In Beyond Lanchester, Brian McCue–author of the similarly quantitative U-Boats In The Bay Of Biscay–focusses on a neglected shortcoming of Lanchester’s work: its determinism. He shows that the mathematics of the Square Law contain instability, so that the end-state it predicts is actually one of the least likely outcomes. This mathematical truth is connected to the real world via examples drawn from United States Marine Corps exercises, Lanchester’s original Trafalgar example, predator-prey experiments done by the early ecologist G.F. Gause, and, of course the war against German U-boats

This is an in-depth discussion of the subject of the use Lanchester equations by Dr. Brian McCue, previously of CNA (Center for Naval Analysis) and OTA (Congressional Office of Technology Assistance). We have also posted and written before about Lanchester (see War by Numbers). Some of our old blog posts on Lanchester are here:

Lanchester equations have been weighed…. | Mystics & Statistics (dupuyinstitute.org)

TDI Friday Read: The Lanchester Equations | Mystics & Statistics (dupuyinstitute.org)

The Lanchester Equations and Historical Warfare | Mystics & Statistics (dupuyinstitute.org)


The book is 121 pages. The Table of Contents for Brian McCue’s book includes:

Introduction

Lanchester’s Theory

A New Look At Lanchester

Trafalgar

Subsuface Combat in a Test Tube

Weaknesses of the Deterministic, Continuous-Variable Approach

A Probabilistic, Event-Driven Revision of Gause’s Work

Theory and Experiment

Implications for Military Operations Research

Applying Hughes’s “Salvo Equations” to Engagements between U-Boats and Convoy Escorts

Wartime Analysis

Using Simulated Annealing to Solve a Problem of “Ecological” inference

Results

Back to Attrition: The Salvo Equations

Results: Fitting HESSE to the North Atlantic Data

Goodness-Of-Fit

Final Thoughts

 

Anyhow, having just discovered it, I have not read it yet. Brian McCue is an old friend of mine and previously published U-Boats in the Bay of Biscay. See: U-Boats in the Bay of Biscay: An Essay in Operations Analysis


 

Wargaming in the U.S. Army

I was asked recently about wargaming in the U.S. Army, and I kind of punted on the question. I then got a friend of mine to answer it. The wargaming efforts we have been involved in have been primarily for analysis and casualty estimation efforts. We have not been involved in wargaming for training, officer development, or development of planning. Other people have been doing that, I gather with mixed results. Here is two recent articles on the subject that were forwarded to me:

1. Educational and Tactical Use of Wargames with the U.S. Army: https://warroom.armywarcollege.edu/wargaming-room/tactical-edge/

2. Wargaming the Gray Zone:

https://warontherocks.com/2021/03/whispers-from-wargames-about-the-gray-zone/

 

Attrition has been discounted to $877.95

One can find a copy of Attrition: Forecasting Battle Casualties and Equipment Losses in Modern War on Amazon.com for $877.95. It used to sell for $890.

Attrition for $900 | Mystics & Statistics (dupuyinstitute.org)

On the other hand, we are still selling new copies for the list price of $19.95. See here: http://www.dupuyinstitute.org/booksfs.htm

Ordering information is here: http://www.dupuyinstitute.org/order.htm