In fairness Feb was the only partially correct forecast. November cold and sunny! October windy! Snow for late Jan! Mild Christmas! You get the message. Probabilistically, 'Joyce' was likely to get something right. Thus the fact that she got everything else wrong (especially nearer the forecast time!) provides strong evidence for a very poor forecast. So indeed, interesting that some will get sucked into thinking the forecast was any better than the one I could have picked out of a hat.