# [ExI] The Chinese bullshit theorem

Stuart LaForge avant at sollegro.com
Sun May 24 18:23:29 UTC 2020

```Claim:
The novel coronavirus most likely came from the lab in Wuhan.

Let L be the hypothesis that nCoV19 came from a lab.
Let M, for market, be the hypothesis that nCoV19 came about by chance
evolution in a wet market of mixed live animals.
Let W be the observation that the virus arose in Wuhan, China.

By Bayes’ Theorem, the likelihood or posterior probability that the
coronavirus originated in the lab given that the initial outbreak
happened in Wuhan is P(L|W)  =  P(W|L)P(L) /(P(W|L)P(L) + P(W|M)P(M)).

Similarly the likelihood that the coronavirus originated in the wet
market given that the initial outbreak happened in Wuhan is P(M|W)  =
P(W|M)P(M) /(P(W|L)P(L) + P(W|M)P(M)).

In order to find likelihood ratios for the two hypotheses we simply
divide their likelihoods

P(L|W)  /P(M|W) =[ P(W|L)P(L) /(P(W|L)P(L) + P(W|M)P(M))]/ [P(W|M)P(M)
/(P(W|L)P(L) + P(W|M)P(M))].

The denominators are equal, so they cancel out to give us a likelihood ratio:

R := P(L|W) /P(M|W)  =P(W|L)P(L)/( P(W|M)P(M))

Now we just need to assign our probabilities. If the virus came about
because of random evolution in the wet market, then it could have
evolved in any wet market in China. Since wet markets have been part
of Chinese culture for centuries, every city in China has a wet
market. There are 686 cities in China and therefore 686 wet markets.
That means that the probability that the outbreak happened in Wuhan
given that the coronavirus came from a wet market is 1/686.

P(W|M) = 1/686

On the other hand, there is only one biosafety level 4 virus lab in
all of China and it is in Wuhan. Therefore, the probability that the
outbreak happened in Wuhan, given that the virus came from the lab is 1.

P(W|L) = 1

probable, then P(L) = P(M) and so they cancel out, giving us a
likelihood ratio

R = P(W|L)/P(W|M)  = 1/(1/686) = 686.

Or in other words, given uninformative unbiased priors, Bayes theorem
demonstrates that it is 686 times more likely that the novel
coronavirus that causes COVID-19 came from the lab in Wuhan rather
than from the wet market there.

Q.E.D.

Note that even if you assign the wet market story a prior probability
of .95, Bayes still says that the lab hypothesis is still 36.1 times
more likely.

Stuart LaForge

```