Can Statistics Predict the Outcome of a War? 572
StatisticallyDeadGuy writes "A University of Georgia scientist has developed a statistical system that can, she claims, predict the outcome of wars with an accuracy of 80 percent. Her approach, applied retrospectively, says the US chance of victory in the first Gulf War was 93%, while the poor Soviets only had a 7% chance in Afghanistan (if only they'd known; failure maybe triggered the collapse of the USSR). As for the current Iraq conflict: the US started off with a 70% chance of a successful regime change, which was duly achieved — but extending the mission past this to support a weak government has dropped the probability of ultimate success to 26%. Full elaboration of the forecasting methodology is laid out in a new paper (subscription required — link goes to the abstract). Some details can be gleaned from her 2006 draft (PDF)."
Re:If i'm reading this correctly (Score:3, Informative)
It's not a new frontier in warfare style; insurgencies have existed back to Roman times and beyond. What is different are the methods used to combat them; for moral reasons we do not permit ourselves to use the traditional techniques. (Plus, its being a bloody mess of cronyism, profiteering, and sectarian violence hasn't helped matters any.)
Covered previously on... (Score:3, Informative)
This has been covered rather thoroughly on Deep Space Nine [memory-alpha.org] - the upshot was something like, 'there might be a 99% chance we're gonna get our asses kicked, but we're human (ie special) so to hell with statistics; bring on the Breen!'
(if the researcher is allowed to compile averages and call it research, I can quote fictional sources and call them precedent. pbbbt.)
Triv
Re:100% likely outcome (Score:1, Informative)
E.g. if I pick a real number randomly it has a probability of 0 % of being a rational number, yet if I pick 42, the real number is indeed rational.
Of course it is near impossible for humans to pick a random real number as most of us only know a few of them (pi, e, sqrt(2), and fractions thereof).
Re:If i'm reading this correctly (Score:3, Informative)
I'm not sure if 122 observations is enough data to support this kind of analysis though.
Re:0% (Score:2, Informative)
Re:Can Statistics Predict the Outcome of a War? (Score:3, Informative)
You can model any past data, but it really isn't a surprise that the model successfully predicts past outcomes, when that very data was used to generate the model. A model is useless unless it's predictive.
After only reading the summary this model will continue to be useless, because with outcomes such as "70% chance of winning" you need a large sample size (read: lots of death) to have any statistic certainty in the validity of the model.
Re:0% (Score:4, Informative)
(I hope most people realize the parent is quoting Dr. Strangelove.)
Re:100% likely outcome (Score:3, Informative)
No, wrong. If an event has probability 0 of occuring, it will never occur, even if you keep trying forever. (Mathematically stated: probability measures are countably additive.)
Re:100% likely outcome (Score:5, Informative)
A statistical model can be predictive without being causal, i.e., the inputs don't necesarrily cause the outcome, but they are observed to occur jointly. Hence the old saying "correlation is not the same thing as causality". There are lots of good examples of this, one of my favorites is that the number of deaths by drowning per month in Finland is highly correlated with the ice cream consumption per month. People don't drown due to the ice cream - the correlation is because the number of people drowning in a given month is proportional to how many people go swimming, and many fewer people go swimming or eat ice cream in Finland's winter months.
Re:100% likely outcome (Score:2, Informative)
Also, see http://instruct1.cit.cornell.edu:8000/courses/sta
non-exact quote.. (Score:5, Informative)
Just in case that hits the nail on the head - send your $5 to a Multiple Sclerosis research center plzkthx.
Re:Statistics, Schmatistics (Score:3, Informative)
Re:100% likely outcome (Score:3, Informative)
Here's a link to an excerpt, if you're interested:
http://www.washingtonpost.com/wp-srv/style/longte
That's not to say I completely disagree with your sentiment, just that it isn't a uniquely human trait. Perhaps a uniquely primate trait, but not strictly human.
Humans ceratinly seem to have the greatest capacity for war, and I don't think we'll be finished with it any time soon, but that doesn't mean our more violent nature isn't some kind of throwback to instinctive behavior.
Re:0% (Score:4, Informative)
wretched, n. "very unfortunate in condition or circumstances; miserable; pitiable." creature, n. "1. an animal, esp. a nonhuman: the creatures of the woods and fields; a creature from outer space. 2. anything created, whether animate or inanimate. 3. person; human being: She is a charming creature. The driver of a bus is sometimes an irritable creature."
People living in conditions such as those they were subjected to in the Third Reich's concentration camps are wretched creatures.
The dictionary. It is your friend. Hug it.
Re:No offense, but you missed the point (Score:2, Informative)
Yeah, Communism's only killed 100,000,000 people so far, let's give it another chance...
Curious that thousands of people were arrested or killed for fleeing from East Germany to West Germany, and really no one went the other way. Maybe the fact that thousands fled - at risk of death - to escape the USSR into the West and the US tells you something about the moral superiority of the US... Dissent here, you don't get shot.
And yes, I have first hand accounts from my mentor Alexandr who escaped the USSR in 1974, via smuggling himself and his son (left his wife and 3 daughters) in a boxcar carrying pigs from the Ukraine to the Balkans, then hiked it out to Austria over 6 weeks, and after 5 years finally made it to the US.
After the wall fell, he was able to go back and visit his wife's memorial (no grave allowed - erected by family), and spend time with his last living daughter. Wife and one daughter were executed for his escape, but it allowed his son to live (who was slated to go for front duty in supressing uprisings in the Ukraine where the casualty rate was near 80%, and fled with Alexandr at his wife's request, knowing she would probably die). Other two daughters were smuggled to relatives living in southern Ukraine. And one died in childbirth in the so advanced hospitals.
Visit an orphanage in the Ukraine sometime. Talk to the people who lived under the boot of the USSR, and get a real clue about how it was. When farmers had to turn over 80% of their crops to feed people in Odessa, while the farmers starved to death. Where you were sent to Siberia - if your family even knew that much - for simply having been seen with a "conspirator" who was accused of trying to overthrow the USSR. Where you would be shot for trying to cross a border, no questions asked.
Yeah, that's SO morally equal to the US...
Re:What are you trying to say? (Score:1, Informative)
As an example, pick a real number uniformly between 0 and 1. Whatever value you picked, you got, so clearly it is possible to get that number, but most people would agree that the probability of picking one particular value out of an infinite set is zero.
On the flip side, what's the probability that the value you pick will be a real rather than a rational number? It's one. It's not that rationals don't exist, but they're freakin' overwhelmed by the number of reals because the former are countably infinite in number while the latter are uncountably infinite.
This being slashdot, I probably have to say that computer programs can't show this stuff. Computer arithmetic doesn't accurately reflect mathematics, because you're always using countable representations of numbers, unless you happen to have a computer lying around with infinite registers.
Re:0% (Score:3, Informative)