The Economist has an interesting piece on military psychology – specifically, the research going into predicting when an opponent who is outgunned and outmanned will just keep fighting:
Military history is, as a consequence, littered with disastrously wrong assumptions about belligerents’ will to fight. America, for instance, famously underestimated the determination of Vietnam’s National Liberation Front when it involved itself in that country’s civil war in the 1960s and 1970s. Similarly, in 1916, during the first world war, Germany underrated France’s will to defend its fortress at Verdun against what the Germans hoped would be a war-winning assault. Casualties in that battle exceeded 300,000 on each side.
Assessing enemy morale is crucial to warcraft. Bruce Bueno de Mesquita, a political scientist at New York University, reckons human will matters enough for four wars in ten to be won by what starts off, in strict military terms, as the weaker side. Behavioural scientists are now, however, bringing the power of modern computing to bear on the question.
…
The researchers charted participants’ responses on a seven-point scale of ascending willingness. The responses suggested that, among other things, those who declared themselves willing to sacrifice the most were the ones who also seemed least interested in material comfort and economic prospects. The researchers then embedded themselves with troops from the interviewed groups (save the IS prisoners), in part to seek differences between stated and actual willingness to fight.
,,,
Artis’s researchers identified fighters who had mentally downgraded their families to second or third place. Some were Peshmerga, who most valued “Kurdeity”—a love for the homeland steeled with commitment to fellow Kurds and Kurdish culture. Many IS captives, for their part, had shunted their families into third place behind the caliphate and sharia. Units girded with those beliefs had fought on effectively even after seven-tenths of their comrades had fallen.