It’s the kind of commonplace everyone feels safe saying: “The industry’s in a terrible state if they’re getting ideas from computer games.” Like many too-quick assumptions about cultural hierarchy, and intelligence, and potential, it’s wrong – and the mistake contains a rather specific kind of stupidity. Time Commanders (BBC2, Thursday, 8-9pm) is a history documentary in the form of a computer wargame, and it’s a sly revelation: the contestants – ordinary punters in groups of four – direct one side in a historical battle. Tonight was Boudicca’s revolt against the Roman Occupation in 60AD. The Britons – the tribes of the Icenii and the Catuvellauni – were the computer’s to control: the group of four modern Britons, two commanders with oversight of the whole field, two lieutenants in the thick of the fighting effecting troop movements, marshalled the Roman defence against the uprising, commanding some hills near Colchester.

The modern Britons are two millennia better informed than Boudicca or her tribal sub-commanders – and the woad-painted CGI Celts wiped the floor with them. Some of the problem was just extreme unclarity of orders: the lieutenants, a man and a woman, got increasingly annoyed at not knowing what their commanders, ditto, meant by “Move your cavalry to the left” Whose left? What direction do you think we’re facing? But just as much of the error derives from semi-digested assumptions about what battles are MEANT to be like. The observing experts, – who comment on the action in another room, then come in to show what really happened – pointed out (scathingly) that the tactics adopted suited troops with muskets facing cavalry, not legionaries facing tribesmen in chariots: the participants half-knew too much about later-hence-“superior” warfare, and – after being goaded out into the open and completely surrounded by vastly larger numbers – decided that savages would all just run away if their leader was killed. So they switched their concentration entirely to chariot-chasing Boudicca and her two teenage daughters round a far corner of the battlefield, with a view to assassination, while their legionaries were massacred to a man. Boudicca escaped: but even if her troops had (computer-magically) been able to see her killed, these were hard semi-autonomous tribal units fighting for their combined cultural survival, not frightened children.

The programme set-up neatly dramatises the problems of the flow of information in a battle: if speed and precision of command aren’t there, then everything turns to rubbish. The graphics are clunky and not realistic: but this actually allows for a clearer view of the overall situation, if you have the will to keep your distance. The close-up zoom – which is what most “better” knowledge really consists of – supplants a grasp of the whole with the glamour of fascinating detail. Tonight’s group were particularly poor, disastrously losing a battle the Romans had convincingly won – but last week’s group actually improved on the historical (Carthaginian) victory. In both instances, the amateur commander had fairly swiftly intuitively grasped what the strategy ought to be: fatally for her little pixelated robots, this week’s then allowed herself to be distracted and dissuaded by her charming fellow officers (human and equals outside the game) and forgot what the various robots, on her side and not, knew and didn’t know, could and couldn’t do. A slightly different version of the same, rather specific kind of stupidity.