I did a backtest to see if eliminating signals based on the previous day's candles would improve or hurt performance. Entry is $10K worth of shares at the open of the 15 min bar when there is a cross of the 9/39 EMAs on the previous bar. For the "no gap " test the earliest entry is the open of the 10am EST bar. Results indicated that eliminating the early bar entries hurt the performance. Net long 115% profit reduced to 111% and Net short profit 133% reduced to 99%. Test from 6/1/2007 to 11/7/2008.