When comparing selected forecasts made this time last year to actual results it seems like they must have been looking at completely different sets of data.