For a long time, many have tried to predict the stock market and identify its trends and patterns. Using Support Vector Machines (SVM), we predict various N days ahead of the NASDAQ Index from Jan 2000 to Dec 2018 using various selective predicting patterns. We only use the given stock opening, closing, high, low, and volume data to derive its corresponding technical indicators to predict N days ahead. This study utilizes selective thresholds to predict stock markets, which selects certain stock days to train and test the model based on whether a certain technical indicator meets the criteria of values. We ended with 71.36% accuracy with Momentum 20 Selection (greater than 0.2 and less than -0.2 after normalization) when predicting 25 days ahead. This accuracy is significantly higher than previous studies that used only technical indicators mainly due to the selective predicting methods. Contrary to the direction of turning towards more niche data sets, we show that traditional technical indicators are sufficient in achieving high accuracy predictions in the stock market.