Thank you Ben! The 80% CI[1] is an output from the model.
Rough outline is.
Start with an uniformative prior on the rate of accidental pandemics.
Update this prior based on the number of accidental pandemics and amount of “risky research units” we’ve seen; this is roughly to Laplace’s rule of succession in continuous time.
Project forward the number of risky research units by extrapolating the exponential growth.
If you include the uncertainty in the rate of accidental pandemics per risky research unit, and random variation, then it turns out the number of events is a negative binomial distribution.
Include the most likely number of pandemics to occur until the probability is over 80%. Due to it being a discrete distribution, this is a conservative interval (i.e. covers more than 80% probability).
Nice! This is helpful, and I love the reasoning transparency. How did you get to the 80% CI?(sorry if I missed this somewhere)
Thank you Ben! The 80% CI[1] is an output from the model.
Rough outline is.
Start with an uniformative prior on the rate of accidental pandemics.
Update this prior based on the number of accidental pandemics and amount of “risky research units” we’ve seen; this is roughly to Laplace’s rule of succession in continuous time.
Project forward the number of risky research units by extrapolating the exponential growth.
If you include the uncertainty in the rate of accidental pandemics per risky research unit, and random variation, then it turns out the number of events is a negative binomial distribution.
Include the most likely number of pandemics to occur until the probability is over 80%. Due to it being a discrete distribution, this is a conservative interval (i.e. covers more than 80% probability).
For more details, here is the maths and code for the blogpost and here is a blogpost outlining the general procedure.
Technically a credible interval (CrI), not confidence interval because it’s Bayesian.
Awesome, thanks!