Revising Greenblatt's Magic Formula approach

I've been spending alot of time working on Joel Greenblatt's Magic Formula approach to stock selection, and thought I'd refine an initial model that I'd put together to help select investments. Over the past week I've spent almost every free waking hour away from work working on this new model (I say waking hour because I'd often fall asleep at my PC at night while building out the calcs). The new model was designed to take into account several more factors than the one he describes in his book. I added several variations of return on capital, a few more measurements on valuation, added an entire section on growth measurements, and added a couple of components on balance sheet strength and price strength.

Well, the new model is pretty much completed and I downloaded new data, sorted things, and .... despite all the changes I pretty much end up with the same list of stocks that the initial simplified model supplied. Maybe in slightly different order, but they're all there. Greenblatt's initial criteria of Ranking stocks based on Return on Capital and EBIT earnings yield seems to be pretty robust, even when supplemented with quite a bit of other types of similar desirable measurements.

Stocks at the top of the interest list are same as before: ARO Aeropostale, WDC Western Digital, TJX TJ Maxx, MRK Merck, BBY Best Buy, GD General Dynamics, and so forth for the bigger names.

I'm still digging around with some of the smaller company names, but overall I'm being presented with a similar set of companies as before.

I guess the next step in working with this is figuring out buy/sell points to make this an ongoing/workable approach.