Comments for Dart-Throwing Chimp

Comment on Lost in the Fog of Civil War in Syria by dartthrowingchimp

dartthrowingchimp — Thu, 23 May 2013 12:29:11 +0000

Thanks, Nils. You’re right, of course, that the fall of Assad wouldn’t directly end the war. I can’t be sure because I didn’t keep careful notes, but I think that awareness of that point might actually have been part of how I rationalized pegging those probabilities so high in spite of what I knew about the base rate of civil-war duration. That loose connection, in turn, hints at one of the challenges of probabilistic forecasting: the base rate of what? Good topic for a future post, that…

Comment on Lost in the Fog of Civil War in Syria by Nils Gilman (@nils_gilman)

Nils Gilman (@nils_gilman) — Thu, 23 May 2013 11:13:16 +0000

Great post, dartthrowingchip. Just to amplify your point about how most civil wars are very long-running, the same is even more true for “subnational conflicts.” The World Development Report from a couple years back estimated that the average duration for subnational conflicts is 41 years. Since Syria could easily degenerate into a subnational conflict, that should be sobering for anyone who thinks the conflict there is likely to end soon. Thinking more in terms of regional history, the civil war in Lebanon, where the ethnic mix is closest to Syria’s, lasted 15 years. Finally, there’s an embedded assumption in several of those charts that if “Aleppo falls” or if “Assad goes” that this means the end of the civil war — which is some kind of a joke, albeit in poor taste.

In sum, there’s every reason to think the civil war could last a long time.

Comment on Lost in the Fog of Civil War in Syria by dartthrowingchimp

dartthrowingchimp — Thu, 23 May 2013 09:55:37 +0000

(c:

Comment on Lost in the Fog of Civil War in Syria by Oral Hazard

Oral Hazard — Thu, 23 May 2013 00:35:24 +0000

Interesting! And congratulations — I imagine it’s a very exciting project to be associated with as a researcher. Supposedly IARPA wants to increase the # of questions next season, so I’m sure they’ll make good use of your skills. Too bad, in a way, because you would have made an awesome teammate sharing personal insights and how they apply to forecasting principals in an interesting and accessible way. It was how I found your site, actually. Best of luck. Now I’ll have someone to blame for the particularly dastardly ones.

Comment on Lost in the Fog of Civil War in Syria by dartthrowingchimp

dartthrowingchimp — Wed, 22 May 2013 20:22:40 +0000

I’m actually working as a consultant to the project for the next season, helping to identify and write the questions.

Comment on Lost in the Fog of Civil War in Syria by Oral Hazard

Oral Hazard — Wed, 22 May 2013 19:33:15 +0000

Jay, were you invited to be a “super forecaster” next season?

Comment on Lost in the Fog of Civil War in Syria by Grant

Grant — Wed, 22 May 2013 16:38:13 +0000

I try to avoid specific numbers, especially since I’m still not a full political scientist. Also we can’t overlook the facts that reportedly Qatar and Saudi Arabia have been supplying rebel groups with at least some weapons, the sectarian nature of the war likely means that the rebel groups don’t lack for recruits and the fact that Iran, Hezhbollah and Iraq (or at least someone in Iraq) wouldn’t be sending fighters to help the Assad government if they thought he was going to win this soon. Additionally we can’t predict international events like a general Sunni/Shia war in the Middle East that might change fortunes.

Still, with the better training of the Syrian military, the support (monetary, military and political) coming in from friends fearful of Assad’s defeat and a Sunni takeover and the fractured nature of the rebels it seems to me that unless a major event occurs in aid of the rebel groups (a politically unified front, the U.S. supplying them with weapons, a NATO intervention etc), time is more on Assad’s side. I will admit though, that I’m actually rather surprised that the rebellion has lasted as long as it has. I could be biased into presuming an Assad victory because of the 1970s Syrian civil war*.

*It should be fascinating to see if Syria suffers more civil wars in the future from its sectarian splits regardless of the outcome of this one.

Comment on Lost in the Fog of Civil War in Syria by dartthrowingchimp

dartthrowingchimp — Wed, 22 May 2013 10:03:08 +0000

Thanks, Richard. What you say about overcompensating is quite possible and had occurred to me, too. That’s actually what I had in mind when I picked a title for the post.

Comment on Lost in the Fog of Civil War in Syria by dartthrowingchimp

dartthrowingchimp — Wed, 22 May 2013 10:00:23 +0000

Well, yes, it's certainly possible. The hard part about probabilistic forecasting is figuring out what you think phrases like that mean and then quantifying your judgment about the likelihood of the event of interest. I'm guessing you mean that the probability of Assad losing the war declines with time. So, then, if you were asked how likely it is that he will still be president of Syria at the end of 2013, what number would you give? And then, conditional on his surviving that long, end of 2014?

Comment on Lost in the Fog of Civil War in Syria by Richard Bridger (@RichardBridger)

Richard Bridger (@RichardBridger) — Wed, 22 May 2013 07:38:44 +0000

Admirably honest Jay. I remember thinking that if even you thought Assad was going to fall, then surely he would.

There is a chance that you are now (subconsciously) over-compensating for your earlier overreaction by downplaying Assad’s advances, which would be another example of motivated reasoning. Not suggesting you are, but it would be horribly ironic.