Last week, I presented some tiers for individual positions. This week, I applied the algorithm across all positions. The result is nice visualization of the tiers for overall players from ranks 1-200.
If this is the first time seeing this model, here is a brief description: (for everyone else, jump to the charts below)
I take an aggregation of expert ranks (via fantasypros), and pass it into a machine learning clustering algorithm called a Gaussian mixture model. This algorithm finds natural tiers among players by consider which players were ranked similarly to other players. The result is charts which help you visualize the tiers and help you draft players.
As always, feel free to direct comments, questions and suggestions below or on twitter.
