Mosquitoes cause Nuclear Wars…
Or a treatise in predicting Causality
We’re going to try to prove that mosquitoes can cause a nuclear war. We’re going to get there mathematically of course, because that’s the right way! We will chain together one piece of thought after another, and each step we will maintain an invariant — the probability of the sequence occurring in the real world, thus far. The way it goes, the longer the chain of “facts”, the less probable the sequence becomes.
As a brief example — “The lion lurking behind the herd of deers started running because a gunshot was fired somewhere close by.” This statement can be arrived at through two steps. The first one is that the herd of deers in question ran, because they heard a gunshot fired somewhere close by. The second, is that the lion ran after its prey. The first statement is then as probable as the product of the probabilities of the latter two (assuming that they are independent, which is true because the lion would likely run after its prey irrespective of whether the herd hear a gunshot before they run).
Now, on the matter of mosquitoes and nuclear wars. Let’s start one fact at a time.
- Malaria is caused by infected female anopheles mosquitoes. The probability of a mosquito bite causing malaria is then the product of the mosquito being infected, of it being a female, and of it being of the anopheles kind. We could write it succinctly as P(malaria) = P(infected).P(female).P(anopheles).
- People go on sick leave when they have malaria. But malaria is only one of the many diseases that keep people away from work. So, the possibility of a person going on sick leave because they have malaria could be expressed as the probability P(sick-leave) = P(leave).P(malaria). Chaining it together with the previous one, we get the possibility of a person going on sick leave because of a mosquito bite to be equal to about P(sick-leave) = P(malaria).P(infected).P(female).P(anopheles), where P(malaria) is the probability that the people on sick leave had malaria. With me so far?
- Okay. Now. We’re going to go fast track on the rest of the statements. People may get overworked when their colleagues go on sick leaves and they must serve as substitutes. P(Overworked People) = P(substitute) . P(sick-leave)
- P(Strike-in-manufacturing-industry) = P(Overworked People).P(Pay-related discontent)
- P(Expensive Farm Implements) = P(Strike-in-manufacturing).P(Farm implements not manufactured in China).
- P(Expensive Food) = P(Expensive Farm Implements).P(Food is from farm and not lab/factory).
- P(General Discontent with Government) = P(Expensive Food).P(Kinderino-shows-up-on-TV).
- P(Government Cries War) = P(General Discontent with Government).P(Smart Government)
- P(Enemy States Perceive Threat) = P(Government Cries War).P(Government Issues Secret Orders)
- P(Enemy Launches Nukes) = P(Enemy State Perceives Threat).P(Enemy has eccentric head-of-state)
- P(Government Launches Nukes) = P(Enemy Launches Nukes).1 (there’s nothing to lose any more).
- P(Nuclear War) = P(Government Launches Nukes) . P(Enemy Launches Nukes) = P(Enemy Launches Nukes)²
Now then. Let’s expand for P(Nuclear War).
P(Nuclear War) = P(Enemy Launches Nukes)² times P(Enemy has eccentric head-of-state) times P(Government issues secret orders) times P(Smart Government) times P(Kinderino shows up on TV) times P(Food is from farm and not lab/factory) times P(Farm implements not manufactured in China) times P(Pay-related-discontent-in-workers) times P(substitute) times P(leave) times P(infected) times P(female) times P(anopheles)
The above result then comes about (probably) by the above 14 clauses. Assuming that each thing is true at least about 2o% (1 in 5) of the time (granted that the Chinese manufacturing clause is a near impossibility), that number is at least 1 / 5¹⁴ . That’s a staggeringly small number already.. All hope is lost!
However, as in most causality studies, you compute the expectation. So what’s the expected number of mosquitoes that bite and cause a nuclear war? It’s the number of mosquitoes in the world times that number 1/5¹⁴.
Turns out, you can figure that {the number of mosquitoes in the world, of course… what were you thinking?} out too (esp. if you are in the tropics). Look around you. Do you see a mosquito within 50cm distance around you. Yes! Good. Now multiply (1 mosquito per meter squared) that with the area of the earth (about 3.78 x 10¹² m²). That’s about 3.78 x 10¹² mosquitoes in all.
Now, to get the expected value, you just multiply those two numbers 3.78x10¹² times 1/5¹⁴ .
That value is 619.315. Clearly greater than 1. You can even factor in the probability of Chinese farm implements being a hundred times lesser and you’d still get a number >1.
Hence, the expected number of mosquitoes that bite and will cause a nuclear war is > 1. All in all, mosquito bites cause nuclear wars. QED.
And that is the nature of Causality prediction…
(Written with a rather large piece of dark chocolate in my stomach).
