If you made a forecast for week 19, based on a 3 week weighted moving average, with oldest data to most recent data weighted 0.20, 0.30, and 0.50, respectively, how much would your forecast be different from the actual jobless claims?