Thanks for the reference. VORP is probably close enough to VAR. I can see how this might work for baseball since the games are played in public. I can also see how this would fail for evaluating work done in an office since the games are played in private. A toxic player would be obvious in front of a crowd but invisible in an office or factory floor.
I'm surprised they didn't use a Monte Carlo method. When I was a kid there was a baseball game based on real players' statistics. It was like fantasy football except for baseball and they didn't have fantasy football l back then. They could just run a million games with a particular player and compare the results to a million games substituting a synthetic average player. An Eastern European friend of mine once said that the Monte Carlo method was actually the American method since it relied on brute force, so maybe I'm just being "patriotic" here.
I think this might be what Brad was thinking of: https://en.wikipedia.org/wiki/Value_over_replacement_player
Thanks for the reference. VORP is probably close enough to VAR. I can see how this might work for baseball since the games are played in public. I can also see how this would fail for evaluating work done in an office since the games are played in private. A toxic player would be obvious in front of a crowd but invisible in an office or factory floor.
I'm surprised they didn't use a Monte Carlo method. When I was a kid there was a baseball game based on real players' statistics. It was like fantasy football except for baseball and they didn't have fantasy football l back then. They could just run a million games with a particular player and compare the results to a million games substituting a synthetic average player. An Eastern European friend of mine once said that the Monte Carlo method was actually the American method since it relied on brute force, so maybe I'm just being "patriotic" here.