According to the Trevor Dupuy’s Quantified Judgment Model (QJM), as described in the book Numbers, Predictions and War (1985 edition), the value of defending is urban terrain is between 1.82 to 2.24. This is a multiplier to the combat value of the defender.
This consists of the value of a defense posture (Table 5, page 230)…
Force Strength
Attack: 1.0
Defense (hasty) 1.3
Defense (prepared) 1.5
Defense (fortified) 1.6
Withdrawal 1.15
Delay 1.2
…Multiplied by the value of terrain (Table 1, page 228):
Terrain Characteristics
Rugged-Mixed 1.5
Rolling-Mixed 1.3
Urban 1.4
Now, the entire terrain table is not included here, there are 14 terrain types in the table. I just included the three that were discussed in the previous post: The Defensive Value of Urban Terrain | Mystics & Statistics (dupuyinstitute.org). As can be seen, there is not a lot of difference between Rolling-mixed to Urban to Rugged-Mixed. Clearly there is an advantage to defending in urban terrain, but no more so than other good defensive non-urban terrain.
Keep in mind that with the QJM, it only requires superior combat value to move forward. So, if a force is doing a prepared defense (1.5) on urban terrain (1.4), which equals 2.1; then a force with more than 2.1 times more combat power will be able to advance against them.
Now, this is a game construct, not a piece of analytical work. It really has no other validity than any other game construct. For example, in “classic” Avalon Hill (AH) games the urban terrain multiplies the defense by 2. But their Combat Result Tables (CRT) are still based upon a three-to-one rule, so attacking urban terrain with twice the force on the AH 1-to-1 combat results table was a truly risky proposition. The attack still has a 50% chance of losing. The AH CRT is here: Force Ratios and CRTs | Mystics & Statistics (dupuyinstitute.org). Whereas with the QJM, the attack is going to go forward and with roughly equal casualties.
Still, the QJM is a model that has been extensively validated three times, so there are some reasons to believe that some parts of it may be close to reality. But we have not validated the model to a large collection of urban engagements. This would be useful to do but it does take a little effort.
I hadn’t ever realized that the QJM was explicitly a game construct!
I have never used the QJM but I understand it uses force rations to resolve battles. At a more detailed level (in reality) there is a tactical struggle and history suggests the better tactician can win even if they do not have a superior force. I understand the QJM has factors for the ability of commanders, but I wonder if this over simplifies the tactical struggle. For example, if I were to fight a battle with Alexander the Great I expect he would win decisively even if I had a significantly superior force. Does this mean the QJM is less useful in adjudicating battles that a system that operates at a tactical level (i.e. Individual units manoeuvring separately and in formations to engage other individual units)?
C.R.: “Does this mean the QJM is less useful in adjudicating battles that a system that operates at a tactical level (i.e. Individual units manoeuvring separately and in formations to engage other individual units)?”
-I think so, fwiw. The original sample was primarily regimental/brigade or divisional sized actions. You’d have to adjust the defensive modifier up for smaller actions and down for larger actions, i.e., you don’t need anything like a 2:1 edge to advance at army group level, but you probably do need that well-known 3:1 ratio at company level. I don’t remember T.N.D ever writing about that explicitly, but he did hint at it in Understanding War in relation to attrition the percentage of forces actually deployed.
Oh, I don’t remember any of T.N.D’s examples involving less than a brigade on the attack (at least, most of one), and he usually stopped at corps level, although some were at army level.
We did validate the model to battalion level. That is discussed in depth in our TNDM newsletters, discussed in my book War by Numbers and discussed on a few blog posts. See: https://dupuyinstitute.org/2019/11/01/tdi-friday-read-battalion-level-combat-model-validation/