The metrics used to rank the coders seem a little odd. The top leaders, who have cumulative scores of 95, have decent github participation (however that's measured) scores in the 80-range, but have Hacker News scores of 97...which means HN karma was weighted considerably more than repo participation.

It's a nice looking site but I think ranking coders will require more granular metrics...and in the end, the coders likely to dominate such metrics will already have no problem finding suitors.