Standings

One score per model across every sourcing and analysis task it has been tested on: quality counts for 70%, cost 15%, speed 15%. Forecasting has its own boards, scored on resolved outcomes.