Why don't you read a book or something: http://www.amazon.com/Short-Term-Trading-Strategies-That/dp/0981923909
Connors and Alvarez are good tarders but some of the strategies they have in their books have stopped working as shown by Mike Harris in his blog. Google "Analysis of Three ETF Counter-Trend Trading Strategies"
For the OP to use as a practice tool because Sharpe is 60 plus. A. Construct filter. Use 7 criteria: exchange location; float range; price range; institutional holdings; minimum liquidity; minimum EPS; and minimum RS. B. Color stock symbols to designate the sentiment. Use function to designate color of symbol. Only trade "long" sentiment. C. Wait rule Do not enter if stock has pro rata volume less than 0.25 of 65 day average. D. Entry rule. Enter when stock on list has first volume of 0.25 of 65 day average within 1.5 hours of open. Technical equivalent if you wish to cut and paste coding: enter if stock has pro rata volume greater than 0.65 of 65 day average after prior wait condition. E. Hold rule. Hold for as long as the stock has pro rata volume greater than 1.80 of 65 day average day after prior entry condition. F. Exit rule Exit the stock when pro rata volume is less than 1.80 of 65 day average day after prior hold condition. Summary. These rules yield results to meet goals. (rules according to MarkBrown). The gola of this system is to never have losses. It can also be traded short by changing the sentiment rule. you will have trades 4 to 5 days long. At a minimum you will make 10% a trade and double capital every 40 days maximum. This means you double 6 times a year. The main purpose of this method is to allow anyone to start with very little money. A first improvement would be to convert this simple program to crossover trading of portfolios. If you are orieinted to the OODA of CW you will not be able to understand how to trade with leading indicators of price. Also coding is very difficult. In part A you need 7 criteria coded and put into a filter construct. For part B the deception that begins to coding is that this is easy. It is not easy and I rarely meet professionals who can code sentiment, ever. In trading, coding has to be failsafe. This means part C is a gate/kill mechanism. Most people do not have complete codes as you see by the posts in ET. In part D I gave you two means. I also introduced how quickly work can be done to code by using cut and paste. Overall I showed you how to have a high Sharpe Ratio (tested by third parties for objectivity). I used these rules at a tradersExpo in an exhibitor's booth. It took only a matter of minutes for them to do the coding for B through F. A was impossible since they did not have the data. A substitute was used instead and we only got the 60 plus Sharpe result. You have to picture the limitations of programmers and traders. All are oriented to the betting and hoping and not doing research or critical thinking. This means they operate on beliefs in a conventional handed down tradition. The people who did the code (drag and drop) needed no time they had the coding tooling down cold. AND THEY WERE INCAPABLE OF BELIEVING THEIR RESULTS. They were defeated immediately since the results did not fit their beliefs and conventional goals. This is not as glib as the MarkBrown airhead contribution. These people repeated the coding and got the same test result for a year's data. Then they saw a new world for what the market offers. What happens now. This post will be trashed and ignored. Hey ,that is how it goes. So have a laugh at the trashing. That is what I am going to do, too.