I have noticed that a lot of cyclicals and garbage stocks made it on the A rank screener due to one time massive earnings. It's also very hard to figure out which is cyclical and which isn't. Apple is technically a cyclical, but I don't think they will suffer much of a downturn. (They are a one of a kind.)
I suggest you build special evaluation screeners for different industries. Like for example, for a financial screener, you could make banks and insurance A ranks based on a high ROE and low P/B where as a cyclical can be more favorable when the P/S or EV/S is well below one. And pharmaceuticals could be A ranked based on good growth, Price to Research ratio, etc. Just a random food for thought in how you can construct a "more accurate" winning stock picker.
EDIT: Also if you would like, I would be very happy to provide more specific ideas and details to help you guys construct specialized screeners.
Customer support service by UserEcho