# {{qctrl.question.title}}

{{qctrl.question.predictionCount() | abbrNumber}} predictions
{{"myPredictionLabel" | translate}}:
{{ qctrl.question.resolutionString() }}
{{qctrl.question.predictionCount() | abbrNumber}} predictions
My score: {{qctrl.question.player_log_score | logScorePrecision}}
Created by: googolplexbyte and
co-authors

### Prediction

This is the first question of the Fermi paradox series.

In a recent paper, Dissolving the Fermi Paradox by Anders Sandberg, Eric Drexler & Toby Ord of the Future of Humanity Institute, University of Oxford, the Drake's Equation was run as a Monte Carlo Simulation rather than a point estimate using the following distributions for the parameters of the Drake's Equation;

Parameter Distribution:

• $R_∗$ log-uniform from 1 to 100.
• $f_p$ log-uniform from 0.1 to 1.
• $n_e$ log-uniform from 0.1 to 1.
• $f_l$ log-normal rate, $1 − e^{−λVt}$ (giving $f_l$ mean 0.5 and median - 0.63).
• $f_i$ log-uniform from 0.001 to 1.
• $f_c$ log-uniform from 0.01 to 1.
• $L$ log-uniform from 100 to 10,000,000,000.

I thought Metaculus would be able to produce distribution more reflective of our current knowledge, and allow the possibility of running Monte Carlo simulation more reflective of the possible outcomes of the Drake's Equation.

Some of the paper's (and see also this presentation) choices for parameter distributions are surprising such as $f_p$ which is unlikely to resolve to be significantly less than 1, unless I'm mistaken.

It would also be fun to see if the distribution of resolutions to Drake's Equation derived using Metaculus-determined parameter distribution, would match the distribution produced by directly asking Metaculus how Drake's Equation will resolve.