Statistics of Deadly Quarrels

CarlNorthcore writes "Brian Hayes published this paper in the Computing Science chapter of Jan-Feb's American Scientist. It provides a fascinating and [sadly] relevant statistical exploration of our world's deadly conflicts. Look out for the excellent "Web of Wars" diagram."

Comments

How To Stop Wars

[Mittermeyer]

Here is a less obtuse and more practical summation of an 80s book that somehow got short shrift. How To Stop A War is more usable then that article.

Re:How To Stop Wars

[pmancini]

That is interesting. I am a big fan of Dunnigan. He is an impossible person to deal with in real life, so I hear, but he has a brilliant mind. His other related books make good reading as well. I especially liked his book on how the digital revolution has changed warfare over the years.

He and Keegan share a similar idea that is echoed in the article mentioned: "this respect, Richardson's data suggest that wars are like hurricanes or earthquakes: We can't know in advance when or where a specific event will strike, but we do know how many to expect in the long run. We can compute the number of victims; we just can't say who they'll be." Keegan in particular writes in "The Face of Battle" that war is very similar to natural disasters and lists the ways. A good read.

Finally, if one is interested in this sort of thing, Dunnigan and Austin Bay wrote "The Quick and Dirty Guide to Warfare" which makes predictions. The first book in the series was quite accurate 10 years later. The last update appears to be the 1996 third edition.

already gone

[GutBomb]

think of the deadly conflict that is happening in this guy's server room right now. slashdotted in 1 minute.

Mirror of Web of Wars image

[JayAndSilentBob]

Just the image, not the surrounding page..... http://www.sellingmysoul.com/web.jpg Hope someone else grabbed the rest of the site before it died....

Data covers too much, too old

[mybecq]

Although this diagram looks nice, it doesn't present a clear view of what is happening now. It consists of all conflicts between countries from Richardson's statistics (1820 - 1950), with refinements from Wilkinson.

Consider the graph (when it eventually comes up). All the red-lines represent Category 7, which is only the two world wars (the most recent of which was 50+ years ago). Category 6 is for deaths of from 500,000 to 2,000,000.

It would be nice to have information regarding something in more recent history, such as the last 10 - 20 years.

I didn't see anything about

[JeanBaptiste]

completely one sided wars, like Iraq v US (round 1)
wouldn't that be a high magnitude for Iraq and a low magnitude for the US?

and yes I do know that this study did not include that war... were there any completely one sided wars involved in the time frame studied?

Shouldn't this be scaled to the population?

[rebill]

One comment at the end of the article caught my attention:

We now have it in our power to have a magnitude-8 or -9 war. In the aftermath of such an event, no one would say that war is demographically irrelevant. After a war of magnitude 9.8, no one would say anything at all.

This points out a comparison problem within the original research - it does not take into account the population increases over time. For example, somewhere in the mid-1600s, London had a population of 600,000 people, while it currently has a population near 7,000,000. That is the difference between a magnitude-5.7 and a magnitude-6.8 event, using the given scale.

Would factoring in the population growth curve enhance or reduce the apparent randomness of the data?

Napoleon's march into Russia

[jcsehak]

There's a cool map of Nappy's march into Russia, which shows, visually, the losses suffered by the thickness of the line, among other things. Really beautifully rendered. Edward Tutfe (master of information design) is a big fan of it, understandably so.

i can't believe this:

[cosmo7]

doesn't any country realise that the best way to win is to conquer australia first? you get five extra units and you only have to defend two territories.

Re:A bit biased

[jonathanjo]

I didn't like the fact that the thing has no indication of time. What about the fact that the US has been around for little over 200 years while other countries (especially the European ones with lots of lines) have been around for much more than that. Maybe they should limit this thing by time or something.

This covers the period from 1820 to 1950, as explained in the article. And the caption states, "The diagram ignores many changes in national status (such as the assembly and disassembly of Yugoslavia)." Since they used TLD country codes presumably they are ascribing conflicts to the current nation on the soil of the nation that engaged in it, for convenience. "They" *did* "limit this thing by time or something."

RTFA.