Hacker News Reader

Comment by aeternum | Hacker News Reader

From a game-theory POV, all wars are caused by misunderstandings / partial information.

If each side really knew the true strength of the other side, it would be clear which would win and therefore actual war is illogical and unnecessary.

View on HN

That's a fallacious line of reasoning on several fronts.

A) The beneficiaries of a war need not be the ones who pay the cost. If you personally stand to benefit from the war itself, regardless of win or lose, you want the war.

B) It's applying a rational economics approach when it's pretty clear that humans aren't rational. Japan kept fighting even when it was clear they would lose. Pride, image, your future after the war. All things that impact the decision that gets made.

C) War isn't typically total like it was in WWII. Most wars are just skirmishes or prolonged skirmishes that don't meaningfully change borders. In such a scenario, a single "war" loss may still be an important fight to have in the middle of the overall conflict (i.e. to signal to your opponent that you're not a pushover and that continuing on their current path has consequences).

D) Win/lose isn't the only outcome of war. Lose/lose is also another scenario. Additionally, you might lose the war but winning may not be the goal. For example, maybe it only takes me 1/10th the resources for me to wage the war that it takes for you to win. A single war might cost me 100 units, but it's important for me to be able to cause you to spend 1000 units because then I have a 900 unit advantage in some other area.

There's many more, but I don't think imperfect information is the only reason war exists.

(1) Externalities are handled in the game theory literature

(2) Preference definitions are not in scope for most game theory lit; what you're describing are preferences and expected utility, which ARE in the literature and metamodel

(3) This is not really true. It depends on how big you are and how big your coalition is. Consider that Saddam Hussain is no longer in charge of Iraq.

(4) Yes, potential outcomes are considered under game theory.

I get your point that OP's statement feels reductionist, but I wanted to explain that the OP is actually correct.

You didn't work at Rand in the 1960s, did you? :-)

Anyway, the OP may be correct, but is also completely useless since it is impossible to know your own true strength, much less that of the other side.

LOL, no, but, reading the research of that period, it would have been fun!

Bell labs in the 1970s and 1980s as well.

It's not totally useless; the game theory framework is still helpful if you define your expected ROI (and your competitors). It is decision and strategy under uncertainty, of course, but its still a useful decision approach.

I took OPs statement at face value:

> If each side really knew the true strength of the other side, it would be clear which would win and therefore actual war is illogical and unnecessary.

No, knowing the true strength is insufficient. I laid out other kinds of pieces of information that also influence the war being logical and necessary as well as the war being illogical and unnecessary and still happening.

Now, you could strengthen the argument and say "what OP meant was that if we had all the information needed to build a model to accurately predict the exact outcome to a high degree of confidence". To me that's the same as the "god of the gaps" argument though because at the most reduced state you end up with the classical position of "well, if I had the state of every particle in the universe I could predict what would happen". The simulation aspect is physically impossible as far as we know according to [1] because the amount of compute needed would be the universe itself. So now you have to start building a model that trades off error and complexity so that you could actually make a decision before the heat death of the universe.

[1] https://arxiv.org/abs/1704.03880

I really don't think the GP is wrong here. Yes there are ideologues, but there were plenty in the Japanese high command who understood logistics and understood that war with the U.S. would end the Japanese empire. It was the same way with Germany. The people who understood logistics knew it would be absolutely impossible to win a two front war and nearly impossible to win a one front war with the U.S. having a thumb on the scales just with lend lease alone. Germany sent their best generals to the east and they knew they had no hope but to delay the inevitable, and no alternative but to listen to Hitler. They continued to fight because to deny Hitler's madness often meant less than favorable outcomes for these commanders than making peace with Allies in 1945. The war in Japan ended when it became impossible for the emperor to continue with the cognitive dissonance and ignore the fact that the fight had ended years before it even began. Logistics have won every war there ever was.

> Germany sent their best generals to the east and they knew they had no hope but to delay the inevitable, and no alternative but to listen to Hitler. They continued to fight because to deny Hitler's madness often meant less than favorable outcomes for these commanders

That's not entirely true, though. Modern WW2 historiography acknowledges that this vision of WW2, that "Hitler was mad and his generals were afraid to contradict him", is mostly the version the surviving German generals gave in their self-serving memories, as a way to save face and paint themselves as more useful to the West during the Cold War.

More recent assessments contradict this version and show many German generals were enthusiastic about war with the Soviet Union, they fully expected to win, and in many cases goaded Hitler into making fatal mistakes he himself wasn't as sure about (an example: Operation Citadel, the battle of Kursk). The ones who knew the enterprise was doomed? The logistics guys, who did the calculations and predicted when the Wehrmacht would overextend itself and run out of supplies -- not the generals!

Professor Jonathan House, an American military historian specializing on WW2's Eastern Front, has a whole lecture about this, available on YouTube.

> If each side really knew the true strength of the other side, it would be clear which would win

This is too simplistic. Often wars are heavily influenced by things other than just the strengths of both sides. For example - What if the ground hadn't been soaked from days of rain at Agincourt and Waterloo? What if the Germans hadn't held back their armor reinforcements for so long on D-Day?

They'd still lose. Wars are fought not from individual battles but from logistics. Modern military's don't even engage if there isn't a lopsided power imbalance favoring their success.

Yeah like in Black Hawk Down.

In real life the battle of mogadishu saw Americans with 10x fewer casualties. Most modern battles have numbers like this with an order of magnitude fewer casualties on the American side, because American logistic planners try to ensure a massive power advantage.

And yet...

And yet what? Are you not convinced by the casualty data?

I don't seem to recall Mogadishu being any kind of "victory". It's almost like the casualty data didn't tell the whole story.

It only takes one side to engage.

Look at the orders of magnitudes differences in casualties in desert storm, iraq, and afghanistan. When Americans battle they ensure they have a massive power advantage for every fight and this is abundantly clear just from the casualty data. Even if the other side engages first like in desert storm, Americans ensure that they don't land boots unless they have a tactical advantage which they had in Desert Storm thanks to technology that Saddam and his army, which was one of the largest and experienced land forces in the world at the time, could not begin to compete with.

if you find yourself in a fair fight it's time to re-think your strategy

So the rational decision is, if you are attacked then you should surrender immediately?

Is that true? What if the situation is that I am currently stronger then you but can expect to become weaker in the future? Why not force a war now if I can expect to win?

I still haven't finished it (too depressing) but my reading of the book The Sleepwalkers about the events leading up to WWI is that all the parties thought they were in that place; even if they all had a full picture it's certainly possible those beliefs would in fact be true.

If I know that you are currently stronger and would ultimately win a war, I should logically concede now since I will lose anyway.

If I know that you are currently stronger but we both also know that I can draw out a war such that I eventually become stronger, then you shouldn't go to war since you will ultimately lose.

> If I know that you are currently stronger and would ultimately win a war, I should logically concede now since I will lose anyway.

You'll often lose less if you fight to a loss.

In fact, you often won't have to fight at all if you make it clear to your adversary that you will fight to a loss.

If I'm a street thug, I'm not going to try to mug someone who clearly communicates that their response to a mugging is detonating their suicide vest. I will, however, repeatedly victimize someone who clearly communicate that they are a pushover.

This is an iterative game, against the same players. Loss aversion is not a winning strategy to pursue in the long-term. Something resembling tit-for-tat is.

This seems untrue.

There are also genuine uncertainties about the future: things that neither side knows but will impact the outcome. Because of that, there is always a range of outcomes for a conflict, and so both sides would still have an incentive to carry out the war.

Consider, for instance, if there are two countries (A and B) looking at conflict with perfect information. Given their relative strengths, the war will cost both $1 billion, but A will win and reap $1 billion + $1 in plunder. In your model, B will always surrender and give A $1, but the reality is there is a big spread about both the costs and the gains. Lots depends on how well each side fights and responds, even with perfect knowledge of the other.

So...you know, if it's the US versus...the Philippines[1], then sure, the uncertain range of outcomes is small enough that rational thing for the Philippines to do is surrender. But for countries that are even reasonably well-matched, they would want to fight with the goal of making the conflict unattractive for the other side. Probably the best historical example of this is the USSR in Afghanistan.

[1] I am sure the US would never invade and occupy them just because the US found it convenient ;)

> Consider, for instance, if there are two countries (A and B) looking at conflict with perfect information. Given their relative strengths, the war will cost both $1 billion, but A will win and reap $1 billion + $1 in plunder. In your model, B will always surrender and give A $1

In the perfect information scenario, B will surrender more than $1. B is better off at any tribute level up to $1,000,000,000.

Sure, I didn't go far into detail, but it's actually worse than either of us has said because once B surrenders anything ($1 or $1,000,000,000), they are at even greater disadvantage and "logically" should surrender more money. In fact, they should surrender the entire country, as once they are at a disadvantage A can keep coming back forever.

Effectively, the idea that war would not happen with perfect information suggests that the logical thing to do is for all countries to surrender control to the strongest country immediately.

The norm is that you pay regularly occurring tribute until such time as you feel you probably don't have to anymore. But there's no norm of surrendering control; that's a conquest, not a tribute.

There is a lot to disagree with here. First, the article itself provides an example of an army with a superior fighting force that had far fewer casualties than the other side, yet still lost the war. That was due to the civilians back at home lacking the stomach for a long, protracted war abroad. So even if one side has superior strength, and even if the side with the superior strength wins every battle, they might still lose the war.

Another issue is your implicit assumption that the side with the superior fighting force will always win the war, but I don't think you can make that assumption. Just like the better football team can lose to the underdog, there is a stochastic element to warfare. A sudden bit of bad weather can turn the tide of a battle. There are countless other elements that are unobservable and unpredictable that can decide which side wins.

There are also asymmetric payoffs to going to war. A nation might have a slim chance of victory, but the cost of defeat or surrender might be genocide or subjection. How do you assign a payoff to the choice "surrender" when the outcome is the destruction of your nation as it once existed?

[0] https://en.wikipedia.org/wiki/Strategy_(game_theory)#Pure_an...

Wait a minute. Shouldn't the goal of the defender be to make it clear to the attacker that they would incur losses that outweigh possible wins for the attacker, thereby making the attack a lose-lose scenario? This is only possible if the defender makes it credible to the potential attacker that they will defend their country even when the costs would be extremely high and they cannot possibly win the conflict. In other words, at least from deterrence point of view it's not about winning, it's about making sure the attacker would overall lose more than gain (which is not the same as winning against the attacker).

i'll add that there's a lot of people that see death as the door to paradise and fighting for their cause as the key to that door. No amount of logic or information is going to stop them.

when you're dealing with the mentally ill, all bets are off.

Reply on news.ycombinator.com