, , , , ,

Based on the official numbers from mid-August, our current system (started in February) has an overall performance 40% better than baseline.   The basline is an unweighted linear opinion pool (ULinOP).  Our goal is to be solidly 50% better than baseline for Year 2.

Here’s the monthly breakdown, including Dec & Jan where we had software problems and (consequently) very low participation.

No official “T&E” questions had resolved as of mid-August, but we expect the August number to change.

Good job.  With participation rising again, I expect to do even better this autumn.

For the technically inclined, “%gain” is the %difference, calculated as:

\frac{ULinOP-DAGGRE}{\frac{1}{2}(ULinOP + DAGGRE)}