In mid-April last year in this column, I tried to find true breakout candidates based on only two weeks of statistics. Only 29% of those selected maintained their breakout performances. So I thought, if we had a 29% hit rate after two weeks, could we do better after four?