War by Numbers: Grounding Wargames in Hard Data

In a recent session hosted by Fight Club International, Christopher A. Lawrence, Executive Director of the Dupuy Institute and author of War by Numbers, joined the conversation to pull back the curtain on the "science" of wargaming. As a historian by trade with decades of experience building massive combat databases, Lawrence offers a sobering look at how modern wargames—both professional and hobbyist—are often built on a "base of sand."

Here are the key highlights from his discussion on the intersection of historical data, human factors, and the future of combat modeling.

1. The "Base of Sand" Problem

Lawrence highlights a critical issue in the Department of Defense (DoD): the loss of institutional expertise. He noted that when the Marine Corps sought to revitalize their wargaming in 2016, they had to reinvent the wheel because previous experts had disappeared due to budget cuts [10:52].

More alarmingly, most combat models used today are unvalidated. Out of over 150 casualty estimation methodologies, Lawrence found that fewer than six have ever been validated against real combat data [46:53]. Instead, many rely on "face validation"—an officer looking at a model and saying, "this looks right" [47:08].

2. Debunking the 3-to-1 Rule

One of the most pervasive myths in military planning is the necessity of a 3-to-1 force ratio for a successful attack. Lawrence’s research suggests otherwise:

• The Reality: The average historical attack is conducted at a ratio of 2:1 or less [36:32].

• Statistical Significance: Most attacks actually fall within the 1.5 to 2:1 range at the division level [36:39].

• The Defender's Edge: Mathematics consistently shows the defender is statistically stronger than the attacker [12:36].

3. The Mystery of "Breakpoints"

A "breakpoint" is a forced change in posture—when a unit stops attacking or begins to retreat.

• The 40% Myth: Many games use a rule that a unit "breaks" at 40% casualties. Lawrence traces this back to a single 1956 study by Dorothy Clark [24:43].

• The Catch: Clark herself noted that casualties were likely not the primary cause of breakpoints [27:23]. Despite this, the 40% rule has been copied and recycled in manuals for decades without new systematic research.

4. Human Factors: The 3-to-1 Advantage

While the DoD is often loath to measure "soft factors" like training and leadership, Lawrence’s data shows they are decisive.

• Historical Evidence: In 1943, the German army showed a combat effectiveness advantage over the Soviets of up to 3-to-1, largely due to accumulated combat experience [15:27].

• Modern Extremes: In some post-WWII engagements, such as South African forces in Angola, the effectiveness ratio climbed as high as 6-to-1 [15:49].

5. Lessons from Ukraine (Mariupol and Donbas)

Lawrence’s recent work includes analyzing the siege of Mariupol and operations in the Donbas.

• Mariupol: Contrary to the "urban legend" of intense, high-casualty urban fighting, Lawrence found that the Russians were in no hurry. The operation was "laxadaisical," with low daily casualty rates as the Russians waited for a political collapse in Kyiv that never came [39:31].

• Donbas: He describes the operations there as having moments of intensity but generally being slow and non-aggressive [41:55].

The Path Forward: A Scientific Discipline

Lawrence’s closing message is a call to action: wargaming must mature into a scientific discipline. Instead of "plagiarism as flattery"—where game designers copy mechanics from previous games—we need large-scale data collection [59:15].

To build a reliable model for the future, we must first ensure it can accurately replicate the past. If a model cannot replicate the history we know, its odds of predicting a future conflict in Ukraine or the Pacific are slim to none [47:57].

Watch the full webinar here: Data for Wargames with Christopher A. Lawrence