Been thinking about "easy mode" games a bit. In some competitive fields there is a period before many have figured out how to play, where people even doing really basic analysis can win big with little effort.
Show this thread
-
this is prob obvious but how would you apply a monte carlo here? simulate the results based on observed regional distributions, then see how often youd see a spurious relationship? sounds awesome
You know the expected value and sampling variation in each state from the poll results. Add some variables for correlated errors between states and demographics. Simulate the election ten thousand times and look at the results of your simulated elections to get a prediction.
