Let a gambler break down knives, gun ownership, and nuclear weapons game theoretically:
We will make several assumptions in the following...that of the two actors in every example, one is a person tending toward violent behavior. We also assume the other agent believes in self defense. Logically two pascifists will not lead to a violent outcome EVER, and if one is a pascifist not a criminally violent actor, again, there will be no violent outcome.
Imagine you have two people who are aware of each other's arms. In the first example we will have unarmed agents, in the seond knife wielders, in the third gun carriers...then we will deduce the effect of nukes.
If two agents are aware of each other's arms, and they get into a dispute, what is the safest situation overall?
We will define "safest overall" as being a weighted average of each outcome, the possibility of any violence altogether, and the deterent factor. We will express serious harm (death and maiming), and moderate harm (rape, assault, or lesser harms that are not necessarily permanent physically).
We will assume that everyone realizes the safest result for you against an unarmed violent attacker is to have a weapon. By this reasoning, if they have a knife, you a safer to have at least a knife, or preferably a gun.
We will also assume everyone knows that criminals inherently do not follow laws, and violent criminals inherently would rather attack someone when they (the attacker) have perfect knowledge of the victim's arms. They would prefer that the victim be unarmed, then to hold a knife, and finally a gun, in that order. They would rather be unsure as to the status of arms of their victim than to KNOW the victim is armed. The fact both weapons are exposed make it easier for the sake of the math, however, when the information is imperfect it is actually safer, because of game theory. When you have 4 possible victims and tell a violent agent there is only one armed (while not saying which one), there is a 75% chance of success, and depending on the value (marginal utility) of the victim, this may be enough to make it "worth it" for the criminal. When you say 2 of the 4 possible victims are armed, but conceal which ones, the criminal has no incentive to victimize these people, and it cannot be "worth it", barring a survival motive like starvation. This is called being "unexploitable". This balance works better, because obviously if the violent actor had perfect knowledge of which 2 possible victims were armed, he would choose to victimize the other two.
We'll also assume the victims are smaller than the violent actors, and for the sake of simplicity, women. This is often the case in real life, and illustrates the importance of arms in society.
Now let's answer our question:
If two agents are aware of each other's arms, and they get into a dispute, what is the safest situation overall?
When two actors have no weapons and are aware of it, the violent actor will use their size advantage to rape or otherwise cause moderate harm to our victim. At an even lack of arms, the vitim is at a serious natural disadvantage. The chance of moderate harm (when harm occurs) will be estimated at 75%, as the violent actor is prone to such behavior. The chance of serious harm is 25%. The chance of no harm will be 10%, given 1 in 10 times the victim will luckily defeat the attacker. The total chance of harm is 90%.
Now, the two actors have knives. The penalty for failure is now increased for the violent actor. The chance of harm altogether is decreased to 80%. The risk of moderate harm is decreased to 50%, but serious harm risks increase to 50%. The chance of no harm is now 20%, 10% when a successful defense happens, and 10% because the blade is a deterent.
The two actors now have guns. The penalty for the violent actor's failure to succeed is further increased. The chance of harm altogether is 60%. The risk of moderate harm is decreased to 25%, while the risk of serious harm increases to 75%. The chance of no harm is now 40%, 10% self defense success, 30% deterent.
So, who is better off?
Let's assume the cases each had 10 trials:
unarmed: 9 of 10 times there was a violent outcome. 1 out of 10 times the victim and criminal escaped rape, assault, or worse. Of the 9 times violence happened, almost 7 of them were rapes or lesser harms (accounting for injuries to the assailant), and a little over 2 of them were maiming or death (accounting for the death of the assailant).
knives: 8 of 10 times there was a violent outcome. 2 out of 10 times the victim and criminal escaped rape, assault, or worse. Of the 8 times violence happened, 4 of them were rapes or lesser harms, 4 of them were maiming or death.
guns: 6 out of 10 times there was a violent outcome. 4 out of 10 times the actors both escaped assault or worse. Of the 6 times violence happened, 1.5 of them were moderate harm, 4.5 of them were serious harm like maiming or death.
Let's compare:
chance of no harm/ chance of harm total/ chance of moderate harm occuring/ chance of serious harm occuring / deterent factor
unarmed 10%(1)/ 90%(9)/ 75%(6.75)/ 25%(2.25)/ 0%
knives 20%(2)/ 80%(8)/ 50%(4)/ 50%(4)/ 10%
guns 40%(4)/ 60%(6)/ 25%(1.5)/ 75%(4.5)/ 30%
The Weighted average:
WA = ((the % of average non-violent outcomes per category)(the # of non-violent outcomes per category) + (the % of avg non-v outcomes per next category)(the # non-v outcomes per category)) / ((the # of non-v outcomes per category) + (the # of non-v per next category))
So,
WA = ((.1 * 1) + (.2 * 2) + (.4 * 4)) / (1 + 2 + 4)
= 2.1 / 7
= 30% (.3)
The total with all is 30% chance of non-violent results.
Is this less or more as weaponry technology increased? Let's exclude guns to see:
WA = ((.1 * 1) + (.2 * 2)) / (1 + 2)
= .5 / 3
= 16.67%(.1667)
The total with unarmed and knife holding situations is 16.67% of non-violent results, nearly half that of the percentages with guns.
Is there less or more unarmed?
the unarmed percentage was 10% for non-violent results.
The increase from 10%, to 16%, to 30% shows how introducing higher levels of weaponry created a larger and larger incentives for non-violent outcomes.
Observations:
It's obvious as we increase the level of mutual arms we increase the chance of nonviolent, not violent, outcomes. What happens is the criminal agent has a lesser and lesser incentive to try and victimize our potential victim. The reason the chances are arbitrarily increased for deterent in guns is the huge increase in death.
In poker there is something called EV, or expected value. In cash games, no one cares if they lose their whole stack of chips, you can just buy-in again. In tournaments, once you lose all your chips (usually once, but certainly finite), you are done and out of the tournament, and your money is lost. In cash you take any close edge if you're ahead, or think you're ahead, because you can buy back in at anytime. Real life doesn't reflect cash games. In tournaments and in real life you get one buy-in (life), and a fatal mistake ends all future possibility for EV. For this reason you will pass on small edges when ahead, in order to "accumulate EV" in the future, when it's more valuable.
So, our criminal agent jumps from a deterent of 10% with unarmed agents, to a deterent factor of 40% with guns, precisely because guns are more likely to kill them. The "mutually assured destruction" is the most powerful deterent to violent behavior. Other EV for the criminal would be victims with lesser levels of arms (or none), perfect info of those arms (like in the example), and getting "the drop" on their victim (catching them by suprise, which we ignore in our examples). The reason I ignore this particular issue is because it's irrelevant in nuclear war, the destruction is mutually assured by computers versed in game theory mathematics.
Conclusion:
Guns=more safety according to game theory. Concealing these weapons make you even safer. Any higher level of weaponry for nonviolent actors also increases the benefit. We also notice law is not a deterent largely for the violent actor, therefore not worth consideration overall. No criminal in prison thought "I'm going to be caught". Well, few anyway.
Also, since mutually assured destruction leads to less and less violent outcomes, it would be predictable and expected that when every nation on Earth has nukes, assuming nukes are still the highest form of arms at that point, there will cease to be major military actions. War, essentially, will be non-existant. It's also predictable that armies, navies, and air forces incapable of stopping these types of attacks will become obsolete, unless used for OFFENSE largely. In any case, to be purely defensive they will require immense downsizing.
No wonder so many nations want nukes...especially those opposed to us! Ever notice that few (Japan is one) of our allies want nukes? When the violent attacker isn't conflicting with you, you don't feel the need to have a nuke to inact game theory. Miracles! When they do want one, it's to defend against a neighbor with them who they are at odds with!
Nuclear weapons are defensive weapons more than they are offensive weapons. The "seen" of Bastiat's Broken Window Fallacy (to use the economic parable for a non-economic metaphor) is the attack we did on Japan. The "unseen" is the peaceful resolution of the Cold War, the existance at all of a "cold war", and the steady decline in military fatalities across the globe as better and higher levels of weapon technology becomes available. True, we killed 130,000 civilians in Iraq, Afghanistan, and Pakistan over the last 10 years...we killed millions of Vietnamese in less time...and that's 13,000 deaths of civilians a year, the exact same number of Americans killed by guns every year...the flu kills 30,000 Americans a year, and 15,000 American arthritis patients alone die of SAIDs (anti-inflamatory drugs, both over the counter and prescription) as weak as asprin every year in the U.S.
Wars are becoming largely non-threats to American soldiers, and to a lesser extent civilians in the countries we attack. The same is true of our foes legitimately in the fight against us. Yes, we killed a lot of people, but compare the stats to past wars of similar scope....it's decreasing. We've lost far less soldiers in the last ten years than we lost in some BATTLES of the Civil War. This is also why the anti-war movement faces such apathy...the population rises, the number of dead shrink, and so the percentage of impacted citizenry falls (not a reason for draft, please don't start).
I hope everyone can see, nukes are inevitable, and good. Unless you're anti-gun, there is no reason to be anti-nuke. Well, unless you're a primitivist who wants to ban technology. There is no putting the genie back in the bottle. True, it only takes one nut, and those nuts will exist (and have), but look at the numbers above. Do you find them illogical? They are made up, arbitrary, but are they not rational? Could you not look up stats at the CDC and Department of Justice to plug in and prove my point?
I'm not out on a limb here. Nukes are defensive weapons...but only when not used. When used they are either offensive or simply retaliatory. If they can be made precision enough to only attack a certain square milage where government is, like say Iran's government, and not to target civilians like they mostly do now, then I'd say they are rather likable. Unfortunately, they target civilians, and give a threat of total annihilation necessary to game theory and being unexploitable mathematically...so they are good, defensive mostly, but not likable.
I better get rep'd for this one...lol.
