Continuing Tom's post below

You keep saying "x has an exact meaning" about different things, Tom, but I don't think anything much has a truly exact meaning. In fact, a lot of words or statements can have very different or even opposite meanings depending on definitions, assumptions, prejudices, etc.

In the case of "war related" (the term used in the Lancet), the Lancet study described exactly what they were including, and it covered all violence as well as health factors, etc. The UN study called their figure "warfare" deaths, which is, without further explanation, anything but an exact term.

To me, "warfare" deaths would be quite different from "violent war related" deaths, the former being strictly resulting from battles fought between coalition forces and anyone else while the latter would include these plus deaths from assassinations, death squads, kidnapping-murders, inter-factional fighting, market square suicide bombs, IEDs used on non-military targets, etc. - a considerably higher figure, I think you have to agree. But deaths all the same.

You also didn't take into account the generally accepted increase in death rates since the time of the UN study.

So - as I said, the ranges do overlap quite widely if you take these factors into account. If you don't want to take them into account, then IMO you are being hypocritical.