They guessed, and exaggerated, for the most part and there is no one with that role in the army. We have very few reliable sources on army compositions until the fourteenth and fifteenth-centuries, let alone for deaths in battle.

From the eleventh-century there could have been heralds with the army, but their precise role is somewhat indeterminable. The earliest heralds had begun distinct from heraldry: they were messengers and humble officials attached to armies, for example, they are described as waking the warriors before the battle of Las Navas de Tolosa (1212). Heralds' precise purpose in cataloging those present at musters is unclear (as there is no evidence of knights being required to serve, or being paid for service, citing a herald's record as proof). Keen argues that this was ceremonious (creating precedence from those present) rather than practical in intention. In the later Middle Ages they might have kept track of knights and men-at-arms who fell in the battle, but wouldn't have concerned themselves with non-aristocrats.

Chroniclers, lay or otherwise, wouldn't be able to give an account of the thousands that might die on a battlefield which could be a couple of square miles (larger, if the victors gave chase during a rout). So they estimated, and exaggerated, and therefore are rather tricky sources to use. They too would focus on the aristocracy.