In this chapter, Greenblatt speaks to the skeptic who doesn't believe the formula described in the last chapter. To that end, Greenblatt identifies possible pitfalls of using the formula, and attempts to disprove them.
First, he uses the past data to ascertain whether these results could indeed have been realized. For example, if the formula only showed great results because the simulation model used prices of small, illiquid stocks, then it may not be very useful. But Greenblatt ran the formula against markets of all sizes, and found that it beats the markets handily even among large caps.
Second, Greenblatt discusses whether the formula could have been lucky. After all, one could use historical data to come up with several winning strategies, but that doesn't mean they will work going forward, as they could be the result of coincidence. Considering the size of the sample Greenblatt employed (number of stocks over number of years), he does not believe the formula to work as a result of luck.
Finally, another possibility is that the market could wise up going forward, and no longer offer the top 30 stocks (as ranked by the formula) at such a large discount. If those 30 are no longer available, is the formula useless? To counter this theory, Greenblatt divided the stock universe (in his study) into deciles. He found that the deciles outperformed each other exactly as expected. In other words, the 4th ranked decile outperformed the 5th ranked decile, the 5th ranked decile outperformed the 6th ranked decile etc. This suggests the formula works on the general market, and does not require a special group of 30 mispriced securities (which is considered the 1st decile).