The last two posts I made on force ratios was drawn from my book War by Numbers. There are additional posts I did early last year on the subject based upon my in-process follow-on book More War by Numbers. They are summarized here:
I have been fairly diligent about making sure the “categories” that are listed on the right hand column of the blog are maintained. Therefore, clicking on Force Ratio will lead you to all 29 Force Ratio related posts on this blog. There are 1,129 posts on this blog (as of this post).
Now, the purpose of War by Numbers was not to create Combat Results Tables (CRT) for wargames. Its real purpose was to test the theoretical ideas of Clausewitz, and more particularly, Trevor N. Dupuy to actual real-world data. Not as case studies, but as statistical compilations that would show what the norms are. Unfortunately, military history is often the study of exceptions, or exceptional events, and what is often lost to the casual reader it what the norms are. Properly developed statistical database will clearly show what the norms are and how frequent or infrequent these exceptions are. People tend to remember the exceptional cases, they tend to forget the norms, if they even knew what they were to start with.
Chapters 3, 4 and 5 of War by Numbers is primarily focused on measuring human factors (which some people in the U.S. DOD analytical community seem to think are unmeasurable, even though we are measuring them). As part of that effort I ended up assemble a set of force ratios tables based upon theater and nationality. The first of these, on page 10, was in my previous blog post. Here are a few others, from page 11 of War by Numbers.
Germans attacking Soviets (Battles of Kharkov and Kursk), 1943
Force Ratio Result Percent Failure Number of cases
0.63 to 1.06-to-1.00 Attack usually succeeds 20% 5
1.18 to 1.87-to-1.00 Attack usually succeeds 6% 17
1.91-to-1.00 and higher Attacker Advances 0% 21
Soviets attacking Germans (Battles of Kharkov and Kursk), 1943
Force Ratio Result Percent Failure Number of cases
0.40 to 1.05-to-1 Attack usually fails 70% 10
1.20 to 1.65-to-1.00 Attack often fails 50% 11
1.91 to 2.89-to-1.00 Attack sometimes fails 44% 9
Pacific Theater of Operations (PTO) Data, U.S. attacking Japanese, 1945
Force Ratio Result Percent Failure Number of cases
1.40 to 2.89-to-1.00 Attack succeeds 0% 20
2.92 to 3.89-to-1.00 Attack usually succeeds 21% 14
4.35-to-1.00 and higher Attack usually succeeds 4% 26
Page 10 for War by Numbers includes the following table:
European Theater of Operations (ETO) Data, 1944
Force Ratio Result Percent Failure Number of cases
0.55 to 1.01-to-1.00 Attack Fails 100% 5
1.15 to 1.88-to-1.00 Attack usually succeeds 21% 48
1.95 to 2.56-to-1.00 Attack usually succeeds 10% 21
2.71-to-1.00 and higher Attacker Advances 0% 42
Many commercial wargames have something called a CRT or Combat Results Table. It is based upon force ratios. For example, this was one of the earliest CRTs used on Avalon Hill Games in the 1960s.
As can been seen from this Combat Results Table, at 1-to-1 the chances of an attack winning is one-in-three. At 2-to-1 odds the chances of the attacker winning is either the same as the defender winning or is a two-thirds chance of winning. At 3-to-1 odds, the attacker will always win.
Now the variable factor is the exchange result, which is defined that the defender removed everyone and the attacker removes as much as the defender. This usually results in an attacker win, if the attack has the right “spare change.” If the attacker was attacking with a single 7 strength unit against a 3 strength defender and they roll and exchange, then both units are eliminated.
Compare that to the table from my book based upon 116 division-level engagements from the European Theater of Operations (1944-145).
Needless to say, some elements of my book War by Numbers are of interest to the commercial wargaming community.
Poking around the internet, I ran across an article from 1985 by the British journalist and historian Max Hastings, rather provocatively titled “Their Wehrmacht was Better than our Army.” It was published in the Washington Post. I had not seen it before (as I went to work for Trevor Dupuy in 1987):
The language in the first couple of paragraphs is also pretty provocative.
The discussion then goes to Liddell Hart.
The discussion then goes to Trevor Dupuy and Martin Van Creveld.
From Dupuy: “On a man for man basis, German ground soldiers consistently inflicted casualties at about a 50 percent higher rate than they incurred from the opposing British and American troops under all circumstances. This was true when they were attacking and when they were defending, when they had a local numerical superiority and when they did not, when they won and when they lost.”
From Hastings: “A spirit of military narcissism, nourished by such films as “The Longest Day,” “A Bridge Too Far” and “The Battle of the Bulge,” was perpetuated mythical images of the Allied and German armies.”
From Hastings: “Yet to be a soldier in America has never been the honorable calling, outside a few thousand Army families. It has traditionally been the route by which young men of modest origins…may aspire to build a career.”
It is worthwhile to read the entire article.
Now, these claims were controversial in the 1980s, and a number of U.S. Army officers and people out at Leavenworth personally and professionally went after Trevor Dupuy over this issue. There was a long unpleasant discussion of that story written up by the lawyer Thomas Nutter. He was going to turn into a book, but I gather that effort was never completed.
I do address the subject of the relative performance of armies in combat in Chapters 4 through 7 of my book War by Numbers.
The Dupuy Institute does occasionally make use of a combat model developed by Trevor Dupuy called the Tactical Numerical Deterministic Model (TNDM). That model is a development of his older model the Quantified Judgment Model (QJM).
There is an impression, because the QJM is widely known, that the TNDM is heavily involved in our work. In fact, over 90% of our work has not involved the TNDM. Here a list of major projects/publications that we done since 1993.
Based upon TNDM:
Artillery Suppression Study – study never completed (1993-1995)
Air Model Historical Data feasibility study (1995)
Support contract for South African TNDM (1996?)
International TNDM Newsletter (1996-1998, 2009-2010)
TNDM sale to Finland (2002?)
FCS Study – 2 studies (2006)
TNDM sale to Singapore (2009)
Small-Unit Engagement Database (2011)
Addressed the TNDM:
Bosnia Casualty Estimate (1995) – used the TNDM to evaluate one possible scenario
Casualty Estimation Methodologies Study (2005) – was two of the six methodologies tested
Data for Wargames training course (2016)
War by Numbers (2017) – addressed in two chapters out of 20
Did not use the TNDM:
Kursk Data Base (1993-1996)
Landmine Study for JCS (1996)
Combat Mortality Study (1998)
Record Keeping Survey (1998-2000)
Capture Rate Studies – 3 studies (1998-2001)
Other Landmine Studies – 6 studies (2000-2001)
Lighter Weight Armor Study (2001)
Urban Warfare – 3 studies (2002-2004)
Base Realignment studies for PA – 3 studies (2003-2005)
Chinese Doctrine Study (2003)
Situational Awareness Study (2004)
Iraq Casualty Estimate (2004-2005)
The use of chemical warfare in WWI – feasibility study (2005?)
Battle of Britain Data Base (2005)
1969 Sino-Soviet Conflict (2006)
MISS – Modern Insurgency Spread Sheets (2006-2009)
Our bread and butter was all the studies that “did not use the TNDM.” Basically the capture rate studies, the urban warfare studies and the insurgencies studies kept us steadily funded for year after year. We would have not been able to maintain TDI on the TNDM. We had one contract in excess of $100K in 1994-95 (the Artillery Suppression study) and our next TNDM related contract that was over $100K was in 2005.
I really have no idea who JSTOR is or what they do, but I stumbled across my book War by Numbers listed on JSTOR along with a quote taken from each chapter. Not sure who picked these quotes or why. I had nothing to do with this. The link is here: https://www.jstor.org/stable/j.ctt1wvwdp8
The Military Conflict Institute (TMCI) decided after a decade of existence that maybe the best use of their time was to write some papers and reports. So this they did over the years, all volunteer (unfunded) work done by some of the “graybeards” of the operations research and combat analysis community.
These three reports are provided here and will be available for as long as this blog exists:
First is called Anatomy of a Combat Model and was written in May 1995 by Lawrence J. Low. It was “prepared as part of a long-term contribution to the field.” It is 81 pages.:
The second paper was from 1997 and is called A Concise Theory of Combat. It was written by Edmund L. DuBois, Wayne P. Hughes, Jr., and Lawrence J. Low “…in collaboration with The Military Conflict Institute.” It is 172 pages.
The third paper was from 2013 and is called A Philosophy of War Its primary authors were Frank Benedict, Rosser Bobbitt, Ted DuBois, Chuck Hawkins, John Honig, Wayne Hughes, John McIver, Roger Mickelson, Clayton Newell, Itzhak Ravid, Russ Vane, Gene Visco and Greg Wilcox. It is 245 pages.
I would also argue that the books Understanding War and Attrition by Trevor Dupuy and my books America’s Modern Wars and War by Numbers are also very much in line with the original mission of TMCI , which Trevor Dupuy co-founded.
Niklas Zetterling’s revised and update version of his excellent book Normandy 1944 is being re-issued. According to Amazon.com it will be available January 10, 2020. The link is here: Normandy 1944
It is set up to “look inside” so you can get some idea what is in there. It is of course, not another war story but a two part discussion on “Campaign Analysis” and “German Combat Formations.”
The “look inside” feature did not include an ability to search the text, so I was not able to check the really important stuff, like how many times Trevor Dupuy and I are mentioned in the book. I am graciously acknowledged in the introduction (as is Richard Anderson). Now, I did write an appendix for the original book. Always the gentleman, Niklas did ask my permission to remove it from this edition.
The book does include a discussion of the relative combat efficiency of the German forces compared to British and U.S. units, always a sensitive subject. We have never invested a lot of time in analyzing Normandy. Most of our analysis of this subject is from Italy 1943-44, Ardennes (Battle of the Bulge) 1944-45 and Kharkov and Kursk 1943 (and shown in War by Numbers). So this is a nice independent look at the subject using additional data from a different campaign by a different scholar.
As a result of a comment by Tom from Cornwall, we ended up adding three posts to this discussion that looked at terrain and amphibious operations and river crossings in Italy:
The previous posts on this discussion on force ratios are presented here. These were the posts examining the erroneous interpretation of the three-to-one rule as presented in Army FM 6-0 and other publications:
We are going to end this discussion for now. There is some additional data from the European Theater of Operations (ETO) and Ardennes that we have assembled, but it presents a confusing picture. This is discussed in depth in War by Numbers (pages 32-48).
I am assembling these discussions on force ratios and terrain into the opening chapters for a follow-on book to War by Numbers.
Polish Sherman III after battle on Gothic Line, Italy, September 1944
Having looked at casualty exchanges from my book War by Numbers and in the previous post, it is clear that there are notable differences between the German and Soviet armies, and between the Israeli and Arab armies. These differences show up in the force ratio tables, in the percent of wins, and in the casualty exchange ratios. As shown above, there is also a difference between the German and the U.S. and UK armies in Italy 1943-44, but this difference is no where to the same degree. These differences show up in the casualty exchange ratios. They also will show up in the force ratio comparisons that follow.
The Italian Campaign is an untapped goldmine for research into human factors. In addition to German, American and British armies, there were Brazilian, Canadian, French, French Algerian, French Moroccan, Greek, Indian, Italian, New Zealander, Polish, and South African forces there, among others like the Jewish brigade. There was also an African-American Division and a Japanese-American battalion and regiment actively engaged in this theater. Also the German records are much better than they were in the second half of 1944. So, the primary source data these engagements are built from are better than the engagements from the ETO.
We have 137 engagements from the Italian Campaign. There are 136 from 9 September to 4 June 1944 and one from13-17 September 1944. Of those, 70 consisted of the Americans attacking, 49 consisted of armed forces of the United Kingdom in the offense, and 18 consisted of the Germans attacks, often limited and local counterattacks (eight attacks against the United States and ten attacks against the UK). So, let us compare these based upon force ratios.
American Army attacking the German Army, Italy 1943-44
(70 cases in the complete data set, 62 cases in the culled data set)
Force Ratio……………Percent Attacker Wins………………..Number of Cases
There were seven cases of engagements coded as “limited attacks” and one case of “other”. These eight cases are excluded in the table above on those lines in italics.
Needless to say, this is a fairly good performance by the American Army, with them winning more than 40% the attacks below two-to-one and pretty winning most of them (86%) at odds above two-to-one.
British Army attacking the German Army, Italy 1943-44
(49 cases in the complete data set, 39 cases in the culled data set)
Force Ratio………………..Percent Attacker Wins………………..Number of Cases
There were five cases of limited action and five cases of limited attack. These ten cases are excluded in the table above on those lines in italics.
This again shows the difference in performance between the American Army and the British Army. This is always an uncomfortable comparison, as this author is somewhat of an anglophile with a grandfather from Liverpool; but data is data. In this case they won 44% of the time at attacks below two-to-one, which is similar to what the U.S. Army did. But then, they only won only 63% of the time at odds above two-to-one (using the culled data set). This could just be statistical anomaly as we are only looking at 30 cases, but is does support the results we are seeing from the casualty data.
What is interesting is the mix of attacks. For the American Army 77% of the attacks were at odds below two-to-one, for the British Army only 23% of the attacks were at odds below two-to-one (using the culled data sets). While these 99 cases do not include every engagement in the Italian Campaign at that time, they include many of the major and significant ones. They are probably a good representation. This does probably reflect a little reality here, in that the British tended to be more conservative on the attack then the Americans. This is also demonstrated by the British lower average loss per engagement.[7]
The reverse, which is when the Germans are attacking, does not provide a clear picture.
German Army attacking the American and British Army, Italy 1943-44 – complete data set (18 cases)
Force Ratio…………………..Percent Attacker Wins…………………Number of Cases
0.72 to 0.84………………………….0%………………………………………………7
1.17 to 1.48………………………..50…………………………………………………6
1.89…………………………………….0…………………………………………………1
2.16 to 2.20………………………..50…………………………………………………2
Gap in data
3.12 to 3.24………………………..50…………………………………………………2
The Germans only win in 28% of the cases here. They win in 13% of the engagements versus the U.S. (8 cases) and 40% of the engagements the UK (10 cases). Still, at low odds attacks (1.17 to 1.48-to-1) they are winning 50% of the time. They are conducting 78% of their attacks at odds below two-to-one.
In the end, the analysis here is limited by the number of cases. It is hard to draw any definitive conclusions from only 18 cases of attacks. Clearly the analysis would benefit with a more exhaustive collection of engagements from the Italian Campaign. This would require a significant investment of time (and money).[8]
Regiment de Trois-Rivieres tanks entering the ruins of Regabuto, August 4th, 1943. Source: http://www.sfu.ca/tracesofthepast/wwii_html/it.htm
————————–
[1] There were four limited attacks that resulted in three defender wins and a draw. There was one “other” that was an attacker win.
[2] There three limited attacks that resulted in two defender wins and a draw.
[3] There were four “limited actions” that were defender wins and one “limited attack” that was a defender win.
[4] There as one “limited action” that was a defender win and two “limited attacks” that were defender wins.
[5] There were two “limited attacks” that were defender wins.
[6] The author’s grandfather was born in Liverpool and raised in Liverpool, England and Ryls, Wales. He served in the British merchant marine during World War I and afterwards was part of the British intervention at Murmansk Russia in 1918-1919. See the blog post:
[7] See War by Numbers, pages 25-27. The data shows that for the Americans in those 36 cases where their attack was successful they suffered an average of 353 casualties per engagement. For the 34 American attacks that were not successful they suffered an average of 351 casualties per engagement. For the UK, in the 23 cases where their attack was successful, the UK suffered an average of 213 casualties per engagement. Of the 26 cases where the UK attacks were not successful, they suffered an average of 137 casualties per engagement.
[8] Curt Johnson, the vice-president of HERO, estimated that it took an average of three man-days to create an engagement. He was involved in developing the original database that included about half of the 137 Italian Campaign engagements. My estimation parameter, including the primary source research required to conduct this is more like six days. Regardless, this would mean that just to create this 137 case database took an estimated 411 to 822 man-days, or 1.6 to 3.3 man-years of effort. Therefore, to expand this data set to a more useful number of engagements is going to take several years of effort.
