Become a fan of Slashdot on Facebook


Forgot your password?
Math Politics

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)."
This discussion has been archived. No new comments can be posted.

Can Statistics Predict the Outcome of a War?

Comments Filter:
  • 0% (Score:5, Funny)

    by DigiShaman ( 671371 ) on Wednesday June 13, 2007 @02:59AM (#19487633) Homepage
    It's 0% if you play "Global Thermonuclear War". The only winning strategy is not to play.

    Sounds like fun. Let's test this theory =)
    • by adona1 ( 1078711 )
      Or else be really good at tic-tac-toe [] =)
    • Re: (Score:3, Funny)

      by EuroMike ( 810928 ) about a nice game of chess? :)
    • by Anonymous Coward
      Yes they can; but only 50% of the time.
      • Re: (Score:3, Informative)

        by AndersOSU ( 873247 )
        On Retro-active statistics...

        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.
        • by porcupine8 ( 816071 ) on Wednesday June 13, 2007 @09:59AM (#19490469) Journal
          But if you use, say, half the wars in the past 200 years as data to create the model, and then find that it predicts the other half (which were NOT used as data) accurately, doesn't that show some predictive power? What if you used it on some obscure wars whose outcome you didn't know until after you'd run the model - how would that be different from using it on a current/future war? The model doesn't know whether you know the outcome or not, unless you're being dishonest and futzing with it every time to get the results you want.
          • by AndersOSU ( 873247 ) on Wednesday June 13, 2007 @10:14AM (#19490669)
            Because we DO know the outcome, maybe not specifically, but generally, and those generalities are built into our assumptions, whether we use the specific data or not. That is especially true for historical studies. For example, we know that feudalism died out, so we're inclined to negatively weight a feudalistic society against a monarchy. Historically that is valid. However, if a future feudal society were to emerge so much would have changed that the assumptions for negatively weighting feudalism would no longer be valid - and the model would have no predictive power. It all has to do with subtle interactions between "independent" variables, and the impossibility from separating things like fascism from the other variables like depression in the 30s and 40s.
  • by quokkapox ( 847798 ) <> on Wednesday June 13, 2007 @03:00AM (#19487641)

    Is that lots of people are going to suffer and die, and lots of money will be spent, usually with detrimental results to all parties involved.

    Oh yeah, and the companies that make bombs and guns will get richer.

    • by Anonymous Coward on Wednesday June 13, 2007 @03:17AM (#19487723)

      Is that lots of people are going to suffer and die, and lots of money will be spent, usually with detrimental results to all parties involved.

      Yep. Those original 13 colonies are still licking their wounds.

      • Re: (Score:2, Insightful)

        by TuringTest ( 533084 )
        Just ask the settlers that lived there at the time the war started.
    • by pchan- ( 118053 ) on Wednesday June 13, 2007 @04:17AM (#19488017) Journal
      Can Statistics Predict the Outcome of a War?

      No. Statistics can never predict the outcome, they can only give you a probability of an outcome. That is, of course, unless the probability is 0 or 100%.
      • by Frequency Domain ( 601421 ) on Wednesday June 13, 2007 @09:12AM (#19489903)

        Can Statistics Predict the Outcome of a War?

        This may or may not be so.

        Statistics can never predict the outcome, they can only give you a probability of an outcome.
        Sorry, but you don't know what you're talking about. There are many kinds of statistical models. Models such as logisitic regression [] map inputs to probabilities of outcomes, but models such as linear regression [] map the inputs to the actual predicted outcomes, not their probabilities.

        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.

    • by dsanfte ( 443781 )

      Oh yeah, and the companies that make bombs and guns will get richer.

      Unfortunately for pacifists, people (when not neutered by an easy life with lots of food and entertainment, such as in the West) are vile, disgusting animals, prone to quarreling with and killing each other. Quite frankly, these types un-neutered by the 'easy life' we lead need to be kept from killing themselves and others.

      That's what "guns and bombs" are for.

      You may disagree with the motives for certain conflicts going on today (Iraq), a

    • by Solandri ( 704621 ) on Wednesday June 13, 2007 @05:44AM (#19488409)
      Are, the 2+ parties involved cannot or refuse to resolve their differences otherwise, and the perceived benefit of war is assessed to be greater than the suffering and death the war will cause. In other words, the suffering and death that will likely happen if you don't go to war would exceed the certain suffering and death of the war itself. Pretty much everyone agrees fighting WWII with a chance to win was a better choice than rolling over and letting Hitler take over Europe.

      I completely agree there are wars started by psychopaths who just want to spread death and destruction while profiting as a consequence (both monetarily and through conquest). But the question then becomes, how exactly do you stop such a psychopath if not through war?

      Wow, I get to play the Hitler card and still be on topic.

      • by nephridium ( 928664 ) on Wednesday June 13, 2007 @10:12AM (#19490641)
        What if - and I know how obtuse it may sound, but bear with me - there was a government, say of a country named Utopia, that had a certain amount of direct and indirect control of the military-industrial complex, the media and a huge chunk of the country's financial assets. Now what if the head of that government (let's call him G.W. Lush) was brought up in the believe that energy is one of the most important commodities and oil was the best available and usable source of energy. Now imagine the guy behind him, Mick Deney, had strong ties (financial and otherwise) to a company Ballimurton that specializes in building oil pipelines and other infrastructure. Furthermore there is a country called Biraq that not only has huge amounts of easily obtainable oil and a very weak military, but is also ruled by dictator Habbam Bussein that nobody really likes and who uses every chance he gets to piss off Lush and his buddies.

        And now, yes I know how ridiculous this scenario may seem to some, imagine a terrorist attack on Mr. Lush's country, which could, with the right amount of propaganda (albeit blatantly dishonest), be blamed on Habbam in order to justify a retaliation in form of an invasion that would make Lush (securing of oil-rich country and building military bases in a region where a lot more oil lies around for the taking) and Deney (no-bid contracts for Ballimurton, revenues are soaring) and their buddies (private contractors that do a lot of things the military does, get paid by the government far more than the 'official soldiers' get and here's the kicker: have no accountability whatsoever, nor are casualties reported to authorities, which always make for bad press) very happy. - If it went the way they would want it to go, that is..

        In this, I admit, a bit of a far-fetched scenario, "the suffering and death that will likely happen if you don't go to war would not exceed the certain suffering and death of the war itself", because the casus belli here was not national security, not a direct thread of any sort against Lush's country's population, but simply an abuse of the power invested in that government in order to make themselves and their buddies richer and more powerful. - In this unlikely scenario your conjecture does not hold true.
  • Makes perfect sense (Score:5, Interesting)

    by MillionthMonkey ( 240664 ) on Wednesday June 13, 2007 @03:00AM (#19487643)
    I notice that the probability of success in Iraq correlates well with George's approval rating.
  • by grahamsz ( 150076 ) on Wednesday June 13, 2007 @03:00AM (#19487647) Homepage Journal
    She took a bunch of historical information about wars, built a model and then when run on that historical information it was 80% accurate.

    Amazing stuff.
    • by Jah-Wren Ryel ( 80510 ) on Wednesday June 13, 2007 @03:12AM (#19487705)
      Why is this amazing? Are you being sarcastic? I can't tell.

      People are always trying to build models based on historical data, especially for things like the stock market. But, as they say, "past performance is no guarantee of future results" - and one big reason is that all it takes is for one significant new factor to come into play that didn't exist in any of the historical data and the model becomes useless.
      • When in doubt, assume i'm always being sarcastic.

        The current iraq war seems to be a new frontier in warfare style. I'm not sure we've ever fought an insurgency quite like that before. Without that knowledge and model calibration we'd probably fail to come up with an accurate number when trying to estimate success. I would imagine the model would show that the US military could easily overpower saddams military - mission accomplished.
        • Re: (Score:3, Informative)

          Note that the model does show this; an initial 70% chance dropping to 26% as the mission changed.

          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.)
        • by cp.tar ( 871488 )

          Even without new frontiers and warfare styles, you can really not predict anything.

          The analyses performed at the outset of the war in Croatia predicted total defeat of the practically non-existent and unarmed Croatian army in mere days.


          You can never factor in everything. And sometimes, things can turn over faster than the weather.

          Besides, counting pure firepower, the US not only has the ability to totally destroy Iraq, it has the ability to destroy the world three times over. And yet...

      • by SerpentMage ( 13390 ) on Wednesday June 13, 2007 @03:22AM (#19487759)
        No kidding since Taleb (Fooled by Randomness, and Black Swan) has explicitly said it is nearly impossible to predict the outcome due to the Black Swan.

        >> Sullivan analyzed all 122 post World War II wars and military interventions in which the United States, the Soviet Union, Russia, China, Britain or France fought a weaker adversary. She examined factors such as the type of objective (on a continuum from brute force to coercive), whether the target was a formal state, guerilla or terrorist group, whether the target had an ally and whether the more powerful nation had an ally.

        >> She tested her model and found that it was accurate in 80 percent of conflicts. It predicted a seven percent chance of success for the Soviets in the 1979 to 1988 war in Afghanistan and a 93 percent chance of success for the U.S. in the 1991 Gulf War.

        Just from reading the abstract what concerns me is hindsight bias. Hindsight bias is when you build a model, based on some data. Then to test the validity of the model you test the data. You can't do that because the model is based on data that you are trying to test.

        To properly test a model you need to use data that is completely out of the blue. For example I would love to have seen her test the model against the American civil war.
        • by grahamsz ( 150076 ) on Wednesday June 13, 2007 @03:28AM (#19487793) Homepage Journal
          And if that new data happens to relate to a new style of conflict, then I doubt the model will accomodate.

          Pre-vietman we were generally exposed to "traditional" wars. Part of the disaster there was that I'm not sure we really gauaged the enemy correctly going in.

          Iraq has a different insurgency again, and we were almost certainly expecting to have to defeat saddams army (which was relatively easy), but we overlooked the "terrorist" contingent.
          • Give the man a cigar! You just hit the nature of the statistical problem point blank. And this is why I think papers that try to wrap themselves in the security of statistics are misleading decision makers.
        • Re: (Score:3, Informative)

          by NonSequor ( 230139 )
          If you've got enough data you can get around this problem by using only half of your data to estimate your parameters. Then you do a statistical test to see if your parameters fit the other half of the data. If you get a good result then you should be more or less safe using the whole data set to retune your parameters.

          I'm not sure if 122 observations is enough data to support this kind of analysis though.
      • by witte ( 681163 )
        Also, war is not the kind of activity for which everybody candidly shares their data, during or after the facts.

        This model may be useful for strategic think-tanks when running through different tactical scenarios, though.
    • She took a bunch of historical information about wars, built a model and then when run on that historical information it was 80% accurate. Amazing stuff.

      The idea of studying history, and learning from it in order to avoid mistakes, is a good idea. The trick is if you can define sufficiently and accurately, many of the significant condtions and possibilities. For instance, the Vietnam war is a success if you consider the objective to be standing up to and halting the spread of worldwide communism, but co

      • by Moraelin ( 679338 ) on Wednesday June 13, 2007 @04:33AM (#19488091) Journal
        Ah, the joys of revisionism...

        Let me see, what happened in Vietnam (and eerily enough, Korea didn't go that differently either.)

        1. Actually refused to allow elections and backed an inept dictator that was hated even by the south. That's a funny way to spread democracy, you know.

        2. It lost its chunk of Vietnam to the communists.

        3. It actually created such an anti-american sentiment in Laos and Cambodia that they went Communist too. You know, let's bomb some countries which aren't our enemies, just because the communists smuggle arms and supplies through their territory. In fact, let's bomb a country that's our _ally_ FFS. If you trace the rise of the Khmer Rouge in Cambodia, it went from a fringe group that noone really supported, to _massive_ support in the zones bombed by the Americans.

        At any rate, voila, two more countries lost to communism as a result of inept American meddling in the area. Way to stop the spread of communism, buddy.

        4. That and Korea scared China into flipping from a country just licking its wounds and wanting to be left alone, to becoming a lot more politically and militarily active. Just because some idiot generals wanted to push the border all the way to China, and at least one idiot actually advocated attacking China.

        5. It takes some massive dose of revisionism to call it some spread of communism in the first place, when it was just a country (two, if you count Korea too), that just wanted to reunite. And that the _only_ reason it escalated to war is because the USA didn't allow elections.

        Contrary to Domino theory bullshit and McCarthist propaganda, the USSR was _not_ your enemy at that point. The only reason why there was, say, a north and south Korea was because the Russians actually stopped their advance at the exact spot where the USA asked them to ask. The USSR was still licking its wounds after WW2 anyway, and it knew it's in no condition to start a world war.

        You know what the USSR and China wanted at that point? They just wanted to have no border with NATO, if possible, because the USA had suddenly flipped from being their ally to treating them like mortal enemies. That's one reason why, for example, Stalin actually proposed to let Germany reunite if it stays neutral and doesn't join either pact. The wars in Korea and Vietnam just convinced them to rearm faster and help start the Cold War sooner.

        And if you want something which stopped both sides, that was the rise of long ranged nuclear weaponry and the mutually assured destruction.

        Redefining it as, basically, "nah, see, they calmed down because of the war (USA lost) in Vietnam" is pretty laughable.
        • by calculadoru ( 760076 ) <> on Wednesday June 13, 2007 @10:19AM (#19490745)
          Where to start? You obviously only read about communism. You never actually lived under it. But no matter, it would take aeons to explain to you just how evil it truly was. Instead, let us consider this:
          the USSR was _not_ your enemy at that point. The only reason why there was, say, a north and south Korea was because the Russians actually stopped their advance at the exact spot where the USA asked them to ask. The USSR was still licking its wounds after WW2 anyway, and it knew it's in no condition to start a world war.
          You seem to forget that by 1946 half of Europe was under Stalin's boot, and all of the USSR's satellite states were feeding their citizens massive propaganda against the West. Remember the Berlin blockade? Two full years before the Korean War.
          You think the US 'flipped'? How about the USSR invading Poland at the same time as the Germans, then taking half of it for themselves? How about destroying every single democratic regime in Eastern Europe, despite repeated promises to allow free elections? Do you really believe the USSR was a benign colossus? Do you honestly believe the 1952 Stalin note was anything but a bullshit ploy to divide the Western powers as they were setting up NATO?
          The wars in Korea and Vietnam just convinced them to rearm faster and help start the Cold War sooner.
          Mate, the Cold War started in 1946. Seriously. Look it up. It ended in 1989, when the Russians and the system they imposed through violence and fear were kicked out of Eastern Europe with no military assistance from the West, just by people who were fed up with Communism.
          Just because the right these days is utterly insane and likes to start wars all over the place doesn't mean there were always wrong - just as the left might be right these days, but they have an awful lot of skeletons in their collective closet. /end of rant. Ready when you are, sir.
          • No offense, but you missed the point. The point wasn't whether communism was good or evil. Yes, communism was evil and Stalin was an evil fuck. Ok, now that that's out of the way, let's talk the actual point.

            The point was that US's war in Vietnam was a failure, not some glorious act that stopped communism. It was something self-produced and then lost abjectly. It actually pushed two _additional_ countries to communism: Laos and Cambodia.

            How the heck can _that_ count as having won the fight and stopped commu
            • Re: (Score:3, Insightful)

              by gilroy ( 155262 )
              This is really just the expected, but warped, end result of a morality system that is Manichean: Either you're a saint or you're Satan.

              It is entirely possible for the US to have done bad things and still have been superior to the USSR. I would in fact argue that's actually the case. The US did terrible things that were both mistaken (i.e., counterproductive) and wrong (i.e., unethical). There's a lot of blood on our hands and it will take decades to get clean, if it's even possible. All of that said, yo
    • by jmv ( 93421 ) on Wednesday June 13, 2007 @03:24AM (#19487771) Homepage
      That was my thought. Especially since I'm sure there must lots of parameters that can be tweaked to make the model fit. I can't help but think that "We achieved 80% accuracy learning our training set" isn't a very sexy thing to report in a paper :-)
    • She took a bunch of historical information about wars, built a model and then when run on that historical information it was 80% accurate.

      Sounds like all those black box stock market predictors to me, none of which are better than a dartboard, but it is very useful for news articles and the media loves numbers like this one: Iraq War has 26% Chance of Success".

      However, it's total crap to predict something so complex and varied as a war. Think of the confounding factors! How can there be unbiased inclusion of geographical, political and an infinity of unknowable factors such as which side has the upper hand on intelligence?

      How can we assum

    • by Hal_Porter ( 817932 ) on Wednesday June 13, 2007 @03:52AM (#19487885)
      Whatever dude. I'd like to date a woman like this. We could argue about Prussian foreign policy in the 19th Century, then fuck like mink, then maybe write some code. Then argue about US Foreign policy, then watch some documentary on strategic bombing where I'd play devil's advocate to conventional wisdom, then argue a bit more and have great make up sex.
  • by Xero_One ( 803051 ) on Wednesday June 13, 2007 @03:01AM (#19487653)
    Statistics are like bikinis. What they reveal is suggestive, but what they conceal is vital.
  • by JonathanR ( 852748 ) on Wednesday June 13, 2007 @03:01AM (#19487655)
    As in all projects, when you let the scope blow out, then the costs blow out proportionately. In Operation Iraqi Freedom, the initial scope was to topple Saddam Hussein. Scope then changed to include installation of democracy.

    Nobody wrote up a scope change request, let alone getting it signed off...
    • by dbolger ( 161340 ) on Wednesday June 13, 2007 @04:27AM (#19488059) Homepage
      I know I'm going to get modded offtopic at best here, but wasn't the point of the invasion to stop Iraq from deploying its extensive stockpiles of WMDs? Wasn't it supposed to be a pre-emptive strike to get him before he got America and its allies? Toppling Hussein was supposed to be a byproduct of that, but it was not the primary goal.

      You're right, statistical prediction is unlikely to suceed if every couple of months the goal of your war changes, but its next to impossible if even the initial point of the war gets retconned. How far can this go?

      We're going into Iraq to stop Saddam and his WMDs.

      No WMDs found? Oh, then we came to Iraq to stop Saddam and free the Iraqi people.

      Saddam gone and there's still fighting? Then we came to Iraq to fight the terrorists there so we dont have to fight them here.

      Terrorism worldwide increasing despite, or possibly /because/ of the invasion? Then, erm... SUPPORT OUR TROOPS!

      How far can you push this? No statistical model, no battle plan can succeed if the people in charge can't even make up their mind what they are fighting for.
  • Ha! (Score:2, Funny)

    by Anonymous Coward
    That's nothing! I can predict the outcome of a war with 100% accuracy when applied retrospectively.
  • I will predict the winner of the World Series, for any year up through 2006. Looks like the White Sox had a 96% chance of winning in 2005. I'm not gaming the numbers, I swear. I've got a perfect system for retroactive "prediction."

    John Edward has a better act, Ms. Sullivan.

  • Psychohistory (Score:5, Insightful)

    by ttys00 ( 235472 ) on Wednesday June 13, 2007 @03:04AM (#19487675)
    Hari Seldon [], is that you?
  • Future Wars (Score:3, Insightful)

    by PrinceAshitaka ( 562972 ) on Wednesday June 13, 2007 @03:07AM (#19487683) Homepage
    Hindsight is 20/20. I would be more interested in what the odds of Future wars would be. Like against Iran or North Korea, simultainiously.
  • This is (Score:2, Insightful)

    by cojoneees ( 916665 )
    pure historical data. At least it looks that way to me. I find it hard to believe that one can predict accurately the outcome of a war. Think about the super technologies that the involved parties may keep in secret just to have the surprise factor in a war. That could definitely screw up the statistics :)
    • by anagama ( 611277 )
      She makes some good points. One point is that if an objective can be achieved by brute force alone, i.e., without cooperation from the conquered people, that bodes well for success by the brute. As an example, she points to the first gulf war which had the objective of expelling Iraqi troops from Quwait. In contrast, the current mess, i.e., installing a government, requires the cooperation of a population interested to a large degree in not accepting whatever we try to impose. Hence the poor chance of s
  • by kahei ( 466208 ) on Wednesday June 13, 2007 @03:11AM (#19487701) Homepage

    As far as I know nobody has formally specified the 'win' outcome for the war -- so I'm a bit doubtful that anyone has worked out an EXACT 26% (not 25%! That number would sound like a guess! But 26% sounds like SCIENCE!) chance of the US side achieving it.

    If the 26% really was worked out with a reasonable methodology, then the interesting part isn't the number so much as whatever definition they came up with of 'victory'.

    That said, giving ridiculously exact answers to impossibly vague questions is fun and harmless. 92.8% of the time.
    • by bersl2 ( 689221 )
      Somebody is missing the point of the point estimate.
    • Considering the current situation, not letting the country fall into Iran/Syria sponsored total chaos nor being taken over by an Al-Queida affiliated islamic dictatorship until the next US presidential election will be a victory. Doing so without the need of massive reinforcements would be a large victory, being able to also evacuate in good order is very unlikely.
  • You don't need fancy statistics to predict that a foreign military invasion will ultimately fail. In the meantime you can also be sure of what a previous poster suggested [].
    Invasions may work when they are combined with a civilian invasion if the population numbers allow for it (like China in Tibet and the like). And even in that example, China has a hard time completely assimilating Tibet and it's not quite done with it yet, after half a century and no armed resistance.
  • Psychohistory anyone?

    That said, odds are someone will stumble across a retrospective analysis of this sort. Wake me after they have correctly predicted the outcomes of the next few wars, then color me impressed.
  • Wrong (Score:5, Insightful)

    by nagora ( 177841 ) on Wednesday June 13, 2007 @04:02AM (#19487937)
    If she thinks there was only a 70% chance of regime change in the early part of the Iraq shambles then she needs to go back and see where she dropped that other 30%.


    • If she thinks there was only a 70% chance of regime change in the early part of the Iraq shambles then she needs to go back and see where she dropped that other 30%.

      No kidding. With the US behind it, a regime change was inevitable. Anyone with knowledge of United States military might and the shabbiness of Iraq's military would surely give a much greater than 70% probability. It's no secret that we were up against an easily-defeatable nation.

      Regime change has definitely occurred; the former head-of-state

  • by rew ( 6140 ) <> on Wednesday June 13, 2007 @04:03AM (#19487941) Homepage
    The problem with this is that the model was "trained" on the same historical data on which it is eventually tested. This doesn't prove anything.

    As an example, a defence contractor once built a system that would recognize wether or not a tank was in a picture. First the system was trained on half the "with tanks" and half the "without tanks" pictures. Next the system got a good percentage correct on the second half of the pictures. It turns out the "with tanks" pictures had been taken on a sunny day, and those without on a cloudy day. So the system was actually telling "sunny" or "cloudy".

    In this case, it could very well be that her system predicts the outcome of the war, based on the weather in tokyo 6 weeks before the start of the war. This example was chosen so that you, not an expert in this field, immediately can dismiss this as a nonsense predictor. But as the model gets more complicated, and you feed it lots of parameters that might seem relevant, even the experts will no longer be able to see the value of such complicated predictions. At some point you just have to "trust the computer".

    Aerodynamics: Yes. We understand the underlying principles, we've verfied the predictions made by the models in real life, and found that it matches very good.

    In this case: No. Before I trust such a model, it would need to be verified (as is, no modifications allowed!) against say at least 20 wars that haven't started yet. If it preditcs the outcome of those correctly, the model has merit.

    I'm not going to wait around (I hope).
  • Anyone who has played an RTS knows this...
  • So I am stupid (Score:4, Insightful)

    by Rumagent ( 86695 ) on Wednesday June 13, 2007 @04:08AM (#19487959)
    Can someone explain me how she can go on about conflicts between states, when the majority of conflicts are characterized by the opposition not being a state? I also have a hard time accepting the definition of victory. She defines it as "A state can attain its political objectives in war by rendering its opponent physically incapable of continuing to fight" (or make them believe that such an outcome is unavoidable). Given these criteria, how can an asymmetric war be won? Is it possible to render every terrorist/freedom fighter "physically incapable of fighting"? It probably isn't, so how many attacks are "just" violence and how many attacks constitutes an opposition?

  • Tags? (Score:3, Interesting)

    by evanbd ( 210358 ) on Wednesday June 13, 2007 @04:15AM (#19488003)
    If ever there was an article that wanted the "No" tag... Damnit Slashdot, give us the old style tags back!
  • So, what are the chances that the Cavs are going to come back from their three-game deficit in the finals and put the Spurs in their place?

  • Anyone remember the Powell Doctrine? Nope? Well it was designed to meet this lady's stated points for sucess and it served us well from Vietnam till Iraq. Then a Bunch of draft dodging Dixiecrats and Nixon era hawks ignored those lessons in a classic case of groupthink and now we are back in Vietnam.

    I think we should impose a punitive tax on anyone who contributed to the Republican party during the last 6 years equal to the cost of the war.
  • You can make an equation that'll fit any known values. Trick is to apply her equations to a future war and see if her predictions measure up with reality. Iran. North Korea. Maybe China as she gets bigger and bolder, or Russia if Putin finds Yeltsin's Kremlin Vodka stash.

    > Full elaboration of the forecasting methodology is laid out in a new paper (subscription required -- link goes to the abstract)

    Too bad she's not smart enough to post to an Open Journal. More government-funded research being resold by p
  • by Flying pig ( 925874 ) on Wednesday June 13, 2007 @04:37AM (#19488109)
    This is Slashdot, we are geeks, please can we correctly distinguish between forecasting and statistical analysis? Forecasting is an activity in which you develop mathematical models to describe a system based on the analysis of existing systems, e.g. if I find there is a -0.7 correlation between the global mean temperature and the estimated number of pirates in the Caribbean, that is analysis of an existing system using statistics, but I would not build a mathematical model of global warming based on that without applying a great deal of non-statistical input.

    That said, I find this very unconvincing. And why? Because it is actually very hard to measure the outcome of a conflict, especially when the actual strategic objective of the conflict may be a state secret on the side of the aggressor. Put simply, we do not really know, in the case of Iraq, what the real objective of the US Government is. Is it:

    • To stabilise Iraq with a government that will be a more reliable client of the US than Saddam was?
    • To destabilise Iraq and the Middle East to prevent power accumulations that threaten the US regional aircraft carrier, USS Israel?
    • To maintain high oil prices by creating instability, enriching the Bush family and their clients?
    • To keep up pressure on other states by showing that the US will intervene and create anarchy if it wishes (The old "remember what happened to XXX country?" "There is no XXX country." "Exactly, that's what you need to remember")
    My point is that for any desired outcome other than the first, the US Government would be achieving its strategic objectives. The fact that these objectives might be objectionable to the majority of the US population is irrelevant; for most of history, wars have been fought by military elites without reference to the interests of the majority of the population. In exactly this way, Vietnam can actually be seen as a victory for the US if the strategic objective was to stop the expansion of Communism. Personally I don't believe in the domino theory, but if you do you can argue that the example of Vietnam stopped other regional states from going Communist.

    In the past, wars usually ended when one side ran out of resources, whether provisioning, human, strategic or geographical. The constraint on warmongers in democratic societies is that society can ultimately strangle the resources of its internal warmongers without, necessarily, killing anybody. It is also possible for democratic societies to change the playing field so that strategic objectives change or become irrelevant. (e.g. by doing so much business with other countries that it becomes impossible to pursue strategic objectives without doing more harm to yourself - which you could say is happening with the US and China.)

  • by ElGanzoLoco ( 642888 ) on Wednesday June 13, 2007 @04:45AM (#19488149) Homepage
    Forget about the statistical model, the Slashdot blurb has completely missed the point (as usual) by emphasing it. The point that Mrs. Sullivan is trying to make - and it's a good point - is that the traditional criteria for assessing the outcome of the conflict and whether you have won or lost (such as the number of buildings blown up and enemies killed, number of square kilometers controlled, etc) have become irrelevant in new types of (asymetrical) conflicts, where the objectives are political more than geographical, and where sociological aspects (support of the population, curbing down radicalism or sectarianism, promoting a particular form of government) determine the outcome of the conflict more than raw firepower.

    The relevant part:
    Driving Saddam Hussein's army out of Kuwait in the 1991 Gulf War and overthrowing his government in 2003 was a brute force objective that was accomplished relatively quickly, for example, but quelling sectarian violence and building support for the current government has been much more difficult because it requires target compliance.

    "We can try to use brute force to kill insurgents and terrorists, but what we really need is for the population to be supportive of the government and to stop supporting the insurgents," Sullivan said. "Otherwise, every time we kill an insurgent or a terrorist, they're going to be replaced by others."

    So, don't panic. No one is seriously trying to "predict" the outcome of a war by statistics alone. It's about time the American academia and military ditch the Cold War mindset they've been stuck in since 1947, and start adjusting to the new realities of warfare and conflict resolution. This has happened in smaller countries (in Europe and elsewhere) some time ago, with varying degrees of success. French opposition to the war in Iraq, for instance, was largely based on a good understanding of which political and sociological forces would naturally prevail in Iraq once the artificial Baathist regime was terminated. In other words: yes, we can blow the country to bits, but once we've done that, there is very little that can be done to manage the country's politics afterwards.
  • by Triv ( 181010 ) on Wednesday June 13, 2007 @04:59AM (#19488215) Journal

    This has been covered rather thoroughly on Deep Space Nine [] - 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!'

    ...and the federation ultimately won the war, so...

    (if the researcher is allowed to compile averages and call it research, I can quote fictional sources and call them precedent. pbbbt.)


  • horse puckey (Score:3, Insightful)

    by misanthrope101 ( 253915 ) on Wednesday June 13, 2007 @07:33AM (#19488967)
    How can you measure success when there are no objectives? We could leave Iraq today, just pack up our crap and leave, and the President could still claim victory. Even if your goal is "democracy" (laughable, since our allies include Saudi Arabia, Kuwait, and other distinctly non-democratic) we can point to the fact that the Iraqis got to vote, and go home. Success!

    Well, success if you ignore the fact that the country is falling apart, has become a haven and catalyst for terrorism, has worse access to health care/clean water/safe streets/medicine/electricity than when Saddam was in power, and their current government also uses torture, detention without trial, death squads, etc. But even if what we're doing is making things worse, more of the same will no doubt constitute improvement.

    Okay, sorry for the diatribe. I'm sure you can use stats to analyze who will win a particular armed conflict--the fact that the USA represents half the global arms expenditures might be relevant. But Iraq isn't a war, but an occupation, which is of a completely different nature. If you just tasked the US military with killing Iraqis all day long, they could do that without any impediment. But asking them to make Iraq into a USA-loving western-style freedom-loving democratic republic, by the brilliant combination of the force of arms and handing out candy to kids, is a bit daft.

  • Doesn't Matter (Score:5, Interesting)

    by alexgieg ( 948359 ) <> on Wednesday June 13, 2007 @09:46AM (#19490327) Homepage
    I've posted the text below to today's thread on vegetable oil, but I think it's relevant for this thread, so here goes a copy:

    The actual reason your (I'm Brazilian) government doesn't do this [the message I was replying to suggested USA simply stopped using foreign fuels] isn't because of oil itself. It's because it has since the 1970s a deal with OPEC by which all OPEC countries would accept only US dollars as payment for their oil, no matter who was purchasing it.

    Now think about it: Germany, Brazil, China etc. want to purchase oil. Their currency isn't US dollar, it's something else. So, they must first acquire US dollars, and then use these dollars to purchase the oil. How do they do obtain US dollars? Well, the US government doesn't give US dollars to other countries for free, to get some they must sell goods to USA. At good prices, mind you, otherwise Americans won't purchase their goods, but those sold by some other country.

    All these countries get the US dollars they need to purchase oil. But not only this amount. Imagine what would happen if for some reason Americans slowed down the purchase of their goods? No US dollars, no oil. Pretty bad, eh? So, all countries build reserves with billions of US dollars, as a way to purchase oil when and if the need arises. Now, obviously, some of these US dollars do come back to USA, otherwise USA would have no exports at all. OPEC countries, for instance, import lots of things from USA. They have tons of US dollars available due to only accepting this as a means of payment. Even so, though, most of these US dollars remain outside USA. Everyone has it, and everyone needs it, so other countries also allow exchanging goods among themselves using US dollars.

    Now, US dollars reserves in foreign countries, as well as foreign exchange of goods using US dollars, both cause one important effect, more important than the above mentioned cheap import goods: less US dollars inside USA. And less dollars inside USA equals low inflation. In other words, this system allows USA to export its inflation to other countries, so that Americans themselves don't feel it. Were all the US dollars abroad come back to USA, and USA would feel a recessive inflation so extreme that 1929 would pale in comparison.

    So, as I said in the beginning, the problems isn't oil itself. It's the money supply. Were OPEC to begin accepting other currencies, all these US dollars floating outside USA would be far less needed, thus starting to flow back into USA. And, guess what? Some months before USA deciding to wage war on Iraq, Saddam Hussein had decided to accept other currencies. Recently Iran has also shown interest in doing so. And what we began to hear? That USA is thinking about waging war on Iran.

    So, don't be fooled. No matter whether the government is Republican or Democrat, any President of the USA will do the exact same thing. Because not doing, by allowing OPEC to accept other currencies, will mean years or even decades of extreme suffering to the American people. And no one has any idea how to solve the problem by any means other than bullying OPEC countries into conformance.

    On the other hand, China, Russia, the European Union, all of them hate this system, because it ties their development to whatever is happening inside USA. And all of them would love to have their currencies among those accepted by OPEC countries, for this would yield them the same benefits USA have: inflation export and direct, non-USA dollar backed, cheap goods imports from all those countries who would need to build reserves of their currencies.

    Do you smell 3rd World War on the air? I do. In a few years, decades if we're lucky, at everyone's backyard.

    [PS for this statistics thread: if you consider that the actual goal is to stop OPEC countries from accepting currencies other than the US dollar, then so far it's being successful. Maybe not 100% so, but nevertheless more than if there were no war. Add this consideration to the statistical analysis and I bet its result would be very different.]
  • by DragonWriter ( 970822 ) on Wednesday June 13, 2007 @11:51AM (#19492233)
    But this is not the first attempt (unsurprisingly) at trying to put together an objective system of predicting military outcomes. Actually, lots of models have been applied, such as the Tactical Numerical Determinisitic Model [] with varying degrees of success.
  • by josepha48 ( 13953 ) on Wednesday June 13, 2007 @01:40PM (#19494109) Journal
    When you look back at things it is so much easier to predict the past :D..

    So what does she say out odds are of winning the 'war on terrorism'?

    Hmm, first we had 76% chance of winning the war in Iraq, and now it is 26%?

    I'm not trying to be a troll, but this post sounds more like reviewing historical data and coming to obvious conclusions about what is already known. I've always heard that statistics can be used to say anything.

    What are the odds of my post being modded up to a 5? What are the odds of my post being modded down to a -1?

Can anyone remember when the times were not hard, and money not scarce?