I've been remiss in not noting the arrival of Military History Carnival #28 at Cliopatria. While it seems to be moving from a round-up of the best military history blogging to covering 'military history on the Internet' generally, there are still some good old-fashioned blogs therein. For example, Sellswords, mercenaries and condottieri presents a fascinating examination of the question: what was the reason for the inaccuracy of early modern firearms -- 'In other words, did soldiers use their firearms to its full potential?'
What I found particularly interesting were the details of experiments into musket accuracy conducted in the 18th century. For example:
Hanoverian experiments in 1790 showed that when fired at various ranges against a representative target (a placard 1.8 m high and up to 45 m long for infantry, 2.6 m high for cavalry) the following results were achieved: at 100 meters – 75% bullets hit infantry target, 83.3% cavalry, at 200 m – 37.5% and 50%, at 300 m – 33.3% and 37.5% respectively.
This statistical approach to thinking about combat seems close to what we would now call operational research, which has its origins in Britain in the Second World War (Bomber Command), the First World War (anti-aircraft gunnery), or maybe Charles Babbage's day (postal delivery), depending on who you talk to. But from my (admittedly limited) understanding of the methods of operational research, it probably could have arisen any time after the development of probability theory in the 17th century. The interest of 18th-century militaries in getting answers to questions susceptible to statistical analysis suggests that the impetus was there, so why didn't it happen sooner? For that matter (and it's a question I keep coming back to), why didn't the RAF develop them in conjunction with the bomber?