> Game-theory software can even help locate a terrorist's hideout [... by estimating] on a 100-point scale the importance a wanted man attaches to his likes (fishing, say) and priorities (remaining hidden or [...] recruiting suicide-bombers). Such factors determine where and how terrorists decide to live. Game-theory software played an important role in finding Osama bin Laden's hideout in Abbottabad, Pakistan, says Mr Owen
After a crescendo of unsubstantiated claims, the article finally reaches a well-known case. The claims here seem laughable. First, by all accounts, bin Laden was pretty stationary in his compound and any hot-air model of his psychology regarding fishing vs. recruiting would have been irrelevant. Second, the US government at the time would have been ecstatic to push a story about game theory finding bin Laden - it would have protected sources on the ground, the guy doing a fake vaccination campaign etc.
Regrettably, the article's agenda seems to be no more than pushing a gaggle of unsubstantiated claims that would help consultants with their magical game-theory software make money off gullible companies / governments.
Hi @arcanus,
sorry approaching him here.
I am an author of a paper you commented on a past thread here (https://news.ycombinator.com/item?id=11977747), and we would like to share your feedback.
Would you please contact me at igorsteinmacher [] gmail?
It's a little frustrating how this article uses "game theory" as an abstract term that covers pretty much all of de Mesquita's analysis without really explaining how it is used.
I am guessing much of de Mesquita's predictive power is less classic game theory and more of an application of his selectorate theory. [0] To oversimplify leaders stay in power through winning coalitions that are much smaller than the populace as a whole. If you can identify that small group that keeps a leader in power and analyze shifts in support within that group, you can make some interesting "predictions".
That being said BBDM while very smart is also a great marketer and he's wrong sometimes, too. [1]
Reminds me of my mother's preferred online seer. With my skepticism, she told me to go to his site and see what he predicted. "Corruption scandal in Italian soccer league". Come on, man... pick a hard one.
But yeah, there's a lot of selection bias in these things. People remember the predictions that worked, and forget the ones that didn't. The grandaddy of these predictions: how many times has the Second Coming failed to materialise so far?
> Bruce Bueno de Mesquita [...] has made some impressively accurate political forecasts
In stories like these I always like to know how many incorrect predictions they made as well. I mean, if a squid can predict the results of a sports game, how impressive is what Bruce Bueno de Mesquita did?
A good proxy measure for the accuracy of BBdM's prediction would be how rich he got out of them. Given that BBdM's mostly concerned with macroscopic predictions about politics his claim [1] that he was right in 2/3 of his predictions, should have translated into immense wealth.
[1] Source: an old interview with him, maybe 10 years ago. I might be wildly misremembering here. Sorry, no better sources.
Being correct in 2/3rds of predictions isn't enough to get rich - he could very well have gotten more things wrong than the conventional wisdom that earns $0 on average.
Poker. Playing poker takes me strait back to micro classes. EV, game theory, all that.
It's also useful for describing a whole range of human interactions. Downside - when you realize people are all cynical fucks, and they all behave in a sub-optimal, selfish way.
It's a branch of maths that formalized a bunch of things and allows lovely proofs for certain interesting settings that economists, sociologists and politologists observe and analyze.
Other than that it currently isn't used that much in computer science.
Getting all the selfish incentives of stakeholders and attackers to enable and secure something like Bitcoin is the example. Whether it works out in the end or not is an open question. The current blockage could a be a failure of the game theory to predict the short term tactics of miners at the expense of the long term.
Competitive strategy relies largely on game theory. Market entry games, where an incumbent has the option to retaliate or accommodate a new entrant is a classic example.
The key aspect in this case (and many others) is one people usually have a hard time grasping: take in account that your opponent is as rational and intelligent as you are, and can predict your actions just as well as you can predict theirs.
In other words, it's like playing chess against yourself.
> a branch of mathematics called game theory, which is often used by economists, to work out how events will unfold as people and organisations act in what they perceive to be their best interests.
If it were really used by economists in power (e.g., in central banks or government), they would have foreseen the end of Bretton Woods, the early 80s recession, the 2000 and 2008 economic crisis, the currency race to the bottom, the strictly monotonically increasing gov't debt, the current housing price increases, ...
They might have also figured out that taxes to prevent smoking are likely to have the same effect if you apply them to employment and other productive applications.
Yet, here we are. The only question that remains is: Stupidity or malevolence?
> The only question that remains is: Stupidity or malevolence?
Stupidity. On your end. Let me address your claims one by one.
> the end of Bretton Woods
Ending Bretton Woods was voluntary. The gold standard leads to volatile inflation; look at the graph in [1]. Volatile inflation is bad because it makes transactions that take time to repay uncertain, see this video [2] for a layman explanation
> the early 80s recession, the 2000 and 2008 economic crisis
If economists at a central bank can foresee a crisis, the crisis doesn't happen. Full stop. What you are left, is, by definition, the ones people who work in regulation didn't see coming. Also note that those crises weren't foreseen by almost anyone, because of the nature of markets.
If you want to prevent crises, you need to put regulation up front that discourages the kind of short-sighted and reckless managerial and shareholder behavior that leads to those.
Hinging the stability of the economic system on the ability to catch an upcoming crisis in the making is doomed to fail, because you need a 100% accuracy (like making a system unhackable -- negative goals are much harder than positive goals).
> the currency race to the bottom
Do you mean low positive inflation, or are you buying into Donal Trump's talking points? Because if it's the former, it's intended, if it's the latter, I'm sorry to say you'll need sources to convince anyone that that's indeed something that exists.
All in all, what your post translates to is "I don't understand any of these things but I'm angry", which is not the right way to advise policy.
"You can't fire me, because I quit" - It's as voluntary as that, i.e., not. There were good reasons to have something close to a Gold standard in place but it just could not be maintained as the quantity of money in circulation increased continuously.
> If economists at a central bank can foresee a crisis, the crisis doesn't happen.
That does not address the fact that they didn't foresee it.
> If you want to prevent crises, you need to put regulation up
Or you just don't finance crises by having a naturally determined interest rate (instead of one determined by the central bank).
> the currency race to the bottom
I mean the fact that every major central bank inflates its balance sheet in order to decrease the value of its currency and "stay competitive".
Decisionmakers don't live outside the system, but within the system. So apply the game-theoretic framework to them, and you'll see why stupid or suboptimal decisions are often the outcome...
After a crescendo of unsubstantiated claims, the article finally reaches a well-known case. The claims here seem laughable. First, by all accounts, bin Laden was pretty stationary in his compound and any hot-air model of his psychology regarding fishing vs. recruiting would have been irrelevant. Second, the US government at the time would have been ecstatic to push a story about game theory finding bin Laden - it would have protected sources on the ground, the guy doing a fake vaccination campaign etc.
Regrettably, the article's agenda seems to be no more than pushing a gaggle of unsubstantiated claims that would help consultants with their magical game-theory software make money off gullible companies / governments.