After reading Quantifying Counterfactual Military History by Brennen Fagan, Ian Horwood, Niall MacKay, Christopher Price, and A. Jamie Wood from CRC Press (2024) I cannot decide if they are fans of wargames…or not. This line in particular, appearing early in the book, started me thinking:
For Evans, for most historians and for us, “exuberant” counterfactual history is no more than implausible extrapolation, akin to playing a board game with history-shaped pieces.
Fagan, et al, p. 3
That line was enough to keep this grognard reading further as I attempted to understand what the authors have to say about counterfactual analysis and military history…as read by a grognard from a wargaming perspective.
“Exuberant” vs. “Restrained”
The authors of Quantifying Counterfactual Military History start off by exploring their definition of counterfactual history. Their literature review, though written with regard to counterfactual historians, aptly describes wargamers as well:
In the historical profession the discussion of counterfactual history – a type of historical writing which creates an alternative historical narrative following an imagined difference in events at a given point in time – has generated more heat than light. At the extreme, it does little more than provide a battlefield for historians who dislike each other’s positions, with the middle ground vacated in favour of lobbing shells across it from entrenched positions. Richard Evans, for example, accuses Niall Ferguson and a group of like-minded historians of using “wishful thinking” to create utopias or dystopias which chime with their political views and in some sense seek to validate them. Evans argues that this historical revisionism is symptomatic of the weakness in historical speculation.
Fagan, et al, p.3
Quantifying Counterfactual Military History talks about two categories of counterfactual history that, again, are relevant to wargaming. The “exuberant” deals with “chains of multiple or alternatives past events and outcomes that never happened” whereas “restrained” is the category that “considers only minimal and clearly possible changes to historical events” (Fagan, et al, p. 3). In my view, most wargames fall into the “restrained” category as they often start with a historical scenario and then allows players to build their own counterfactual. Alternatively, some wargames—especially card driven games (CDGs)—are more akin to an “exuberant” counterfactual. Titles such as those in the COIN series from GMT Games lend themselves, by design, to creating an “exuberant” counterfactual. As Dan Thurot wrote in a recent review of Red Dust Rebellion (designer Jarrod Carmichael, GMT Games, 2024):
I have a little ritual for card-driven wargames. As events resolve, I place them to the side in a face-down pile, taking care to preserve the order in which they appeared. Sometimes, if an event is meant to stay on the table for a while, I’ll jot a note on my phone, sandwiching the offending snippet so that I can restore its proper sequence. When the game ends, I carefully store the deck, facing its cards opposite the rest. Later, after everyone has gone home and I have a few minutes to myself, I’ll pull out the deck again and flip through the miniature history book of our session.
The COIN Series has always been amenable to this ritual. Its events index to the playbook, where the designer is permitted a few sentences to discuss the inclusion of that particular happening. In some cases, I’ll learn something new. Even when I don’t, the process is instructive. What might have happened if our historical procession had gone some other way, if Batista had fled two years earlier or Tet crashed against better-prepared forces. These are, at best, historical fiction. But they’re also the core of systems like COIN, little fantasies that let us test our model against the barriers of what actually happened.
Quantifying Counterfactual Military History takes a different approach to counterfactual history. As the authors explain:
Our purpose in this book is to take a few tentative steps towards using modern statistical theory and computer power to add a quantitative dimension to this debate – and to appreciate some of the dangers in doing so. We are interested in the role of chance in history, which means we are equally interested in the probabilities of the events that happened – the facts – and their complement, the counterfactual.
Fagan, et al, p. 4
As simple as ABC
To quantify their counterfactual cases, the authors of Quantifying Counterfactual Military History use “Approximate Bayesian Computation” (ABC). The authors go to great lengths to explain how ABC is better than what they see as the traditional methodology for combat modeling, Lanchester’s model. The authors describe the Lanchester model as:
…it amounts to no more than the assumption that each of two forces engaged in battle causes damage in proportion to its numbers. Most analysts don’t believe this is true for much of warfare – it gives different results from other models in which a battle is just the sum of individual duels, or long-range fire is unaimed… . It is relatively easy to play out the battles…using the Lanchester model, but the crucial subtlety is in how to calibrate the parameters of the model, some of which we think we know and others we son’t, to the historical outcome. Suppose we do so naively, and the outcome of the simulation is wrong – it does not match the historical outcome. Then how do we interpret our findings?
Fagan, et al, p. 14
ABC, the authors of Quantifying Counterfactual Military History argue, “allows us to effectively distinguish between two possibilities – that we got the values of the parameters wrong, or that the historical outcome was unlikely” (Fagan, et al, p. 14). The authors are proud in their claim that, “ours is the first use of ABC for historical analysis” (Fagan, et al, p. 15). The approach used, the authors admit, is almost wargaming:
In this book, we explore various ways in which we can rethink some military engagements and decisions, quantifying the chances of what rally appended, and what did not and what might have happened had different decisions been made. To the extent to which this is counterfactual, it is restrained rather than exuberant – as already noted, we won’t pursue too far the implications for alternative histories. This approach is close to wargaming; indeed, some of our computer simulations are like conducting simple wargames not a few times but many thousands or millions of times and examining the statistics of the outcomes. But we don’t want to play wargames without understanding precisely the implications of their (necessarily) artificial rules, and the extent of the randomness and arbitrariness built into them.
Fagan, et al, pp. 13-14
“Some Mathematical Background (with No Equations!)” is the title of the only appendix in Quantifying Counterfactual Military History. Here the authors try to explain probability and game theory and ABC to the non-mathematic layperson. Suffice it to say that though the explanation is short on math it is not automatically intuitive. To better understand ABC one needs to read the case studies, of which there are four:
- “Could the Germans Have Won the Battle of Jutland?”
- “Could the Germans Have Won the Battle of Britain?”
- “Could the United States Have Prevailed in Vietnam?”
- “The Road to Able Archer: Counterfactual Reasoning and the Dangerous History of Nuclear Deterrence 1945-1983”
Dogger Bank and Jutland
I am not going to spoil the conclusion of “Could the Germans Have Won the Battle of Jutland?” in Quantifying Counterfactual Military History. I will, however, quote the final sentence:
Rather our restrained detailed counterfactuals push us back towards the importance of individuals, their personalities and their decision-making, including long before the fighting – a theme we shall see again in the next chapter on both sides of the Battle of Britain.
Fagan, et al, p. 64
The role of individuals in history is described in Quantifying Counterfactual Military History in a way that wargame practitioners can likely relate to:
From nineteenth-century military players of Kriegspiel to twenty-first century policy makers using matrix games to attempt to work out the consequences of their decisions, their play-acting is conducted in deadly earnest with disbelief fully suspended, in the hope of reacting confidently to any possibility that becomes real and above all to avoid being surprised.
Fagan, et al, p. 9
Information warfare in the Battle of Britain
While the authors of Quantifying Counterfactual Military History use the Battle of Britain to explain in detail how ABC helps qualify the battle, at the end of the day they pivot a bit and claim that ABC does not necessarily reveal history but instead helps explain it:
We can, however, say with some quantified justification that, for most historians and certainly of those who believe in the “narrow margin”, it was materially possible for Britain to have lost the battle, and therefore to have been invaded – although whether an invasion would have succeeded is another question entirely. Moreover, bootstrap methods enable us to quantify comparisons of contrasting views on differing decisions, providing a jumping-off point for quantitative analysis rather than reducing the debate to mere clashes of opinion. This is what mathematics does, of course – its truth is found in the argument connecting assumptions to conclusions, not in the conclusions themselves.
Fagan, et al, p. 93
Numbers behind Fire in the Lake
In “Could the United States Have Prevailed in Vietnam?” in Quantifying Counterfactual Military History the authors state:
Our counterfactual has two components: increased pacification and a declared readiness to profit from precipitate Communist offensives such as occurred in 1972, when the ARVN, supplemented by US airpower, provided sufficiently to defeat the Communist offensive without the presence of American ground troops. The final allied objective would still have been negotiations. The best we can say is that if our counterfactual had become fact, South Vietnam may have survived, for longer, allowing a more successful nation-building effort, and or a negotiated solution under more favourable circumstances than in 1973.
Fagan, et al, p.131
It is interesting to read the conclusion of the authors of Quantifying Counterfactual Military History that “big data”—to use the modern terminology—was a problem during Vietnam. As the authors state: “…Vietnam and MACV’s quest for data demonstrates the problem of having so much data that is blinds you…“ (Fagan, et al, p. 132). Wargame designers can likely relate as they know that simplified models often can provide just as much, if not more, insight into a problem as a highly detailed approach can deliver.
Shall we play a game?
The final case study in Quantifying Counterfactual Military History touches on a subject of deep personal interest to me. There are many takes on Able Archer and how close we came—or did not come—to nuclear war in 1983. In this chapter the authors dive deep into game theory to explore nuclear deterrence as a game, but with limits. As the authors write:
Instead, we return to our point about the streetlight effect – that there is a danger of using models because they are available rather than because they are right or enable a better decision to be made. We would certainly argue that game theory at least did not help to avert nuclear war. Rather in the end, the lesson is a humane one – that mutual awareness and understanding have to be worked for but are always beneficial. Even when deception, guile, and bluff are real possibilities, it helps to talk, and to think oneself into one’s adversary’s position. The long road to trust – the only sue way out of a dangerous game – can only be travelled through doing so.
Fagan, et al, p. 179
“All models are wrong, but some are useful”
It must of been humbling for the authors of Quantifying Counterfactual Military History to write, “we have reached the limits and been reminded to be conscious of the limitations of quantitative modeling” (Fagan, et al, p. 187). The lure of technology, they write, is like a drug, “lest the sheer power of modern computing seduce an inexperienced user into thinking that modeling is simply a matter of inputting parameters into a ‘black box’ to discover the answer” (Fagan, et al, p. 187).
As much as Quantifying Counterfactual Military History focuses on the math, it is people that make the difference. Here the authors write what I suspect many wargamers already understand:
Finally, we are struck, despite our intentions, by the extent to which personalities matter once one moves beyond the most restrained and cautious counterfactual possibilities. If nothing else our examples have enabled some level of unbiased identification of who those people are, and which decisions mattered at the critical juncture and by how much.
Fagan, et al, p. 188
Roll the die
While I enjoyed reading Quantifying Counterfactual Military History at the conclusion of the book I felt stronger than ever that wargaming is a great—even superior—way to explore history. While mathematicians and data scientists and operations research analysts always try to model a battle it is the human element—like a player of a wargame—that makes the decisions that ultimately decide the outcome of a battle/game. In more than a few ways the authors final conclusion is actually a call for wargaming:
People together make history as we know it, whether that means making decisions which turn out to be correct and are rewarded, or making wrong decisions – wrong because they were ill-considered or ill-informed, and thus always unlikely to turn out well, or merely because they were unlucky. In the end, everyone has to accept the roll of the die.
Fagan, et al, pp. 188-189
Feature image by RMN
The opinions and views expressed in this blog are those of the author alone and are presented in a personal capacity. They do not necessarily represent the views of U.S. Navy or any other U.S. government Department, Agency, Office, or employer.
RockyMountainNavy.com © 2007-2025 by Ian B is licensed under CC BY-SA 4.0
Rocky Mountain Navy Gamer 
Along the same lines, the Cuban Missile Crisis was precipitated because Nikita Khrushchev took the least rational option…putting offensive nuclear-capable missiles and bombers (and as the U.S. found out decades later, their nuclear warheads) in Cuba, which the United States would ultimately detect and demand their removal. The counterfactual is that the Soviets do not introduce nuclear weapons, but continue to provide defensive weapons to the Cubans (which is what CIA expected from a rational actor) https://www.cia.gov/readingroom/docs/CIA-RDP80M01009A000300420003-8.pdf. Being irrational cost Khrushchev his job, but surprised the heck out of the CIA.
Sometimes the historical outcome was the outlier, rather than the median ¯_(ツ)_/¯
…and a challenge to wargame designers. How do you design a game with plausible outcomes yet still account for the Black Swan event?