The 5 groups:
1a. Never vaccinated, No prior infection
1b. Never vaccinated, prior infection
2a. Vaccinated, No prior infection
2b. Vaccinated, prior infection (Infected then Vaccinated)
2c. Vaccinated, prior infection (Vaccinated then Infected)
The 5 groups:
1a. Never vaccinated, No prior infection
1b. Never vaccinated, prior infection
2a. Vaccinated, No prior infection
2b. Vaccinated, prior infection (Infected then Vaccinated)
2c. Vaccinated, prior infection (Vaccinated then Infected)
I'll use some BS numbers to make the point. Let's assume: study population of 100, 20% of the population had a prior infection, half are given the snake oil.
Starting conditions:
1. Never infected, Never Snaked: 40
2. Prior infected, Never Snaked:10
3. Never infected, Snaked: 40
4. Prior infected, Snaked: 10
Now let's assume wave 2 happens. 20% of the never infected gets infected, and 0% of the previously infected get reinfected.
Both groups 1+2 combined and 3+4 combined have an average infection rate of 16% for wave 2. The groups are equal because snake oil is bullshit.
What the Ars author was suggesting is to compare group 1 with group 3+4 combined. Now you have a 20% infection rate vs a 16% infection rate. The snake oil appears to have an (20-16)/20 = 25% effectiveness just by including previously acquired immunity.