I think some of you will find the attached article of interest. I'm not sure I believe the results claimed for this approach, and the study was done in a different environment than today's. Still there may be the germ of something worthwhile here.
It's not arbing. You're just looking for options to buy that are "cheap" in the sense that their IV is less than their SV. If you're long IV, you want to lessen the chances that IV will go lower. You want it to increase after you go long. I look for options where the IV/SV ratio is < 1.0 all the time. Think about what happens to a stock and its options right after an earnings release. In anticipation of the release, the IV of the options gets pumped up due to speculative demand (which is why straddles suck). At the same time, the stock's price generally goes into a low volatility range because everyone's waiting to see whether to buy or sell. IV/SV ratio at this point is going to be much > 1.0 usually. Right after earnings, the stock may gap one way or the other, producing a sudden increase in SV. The IV in the options immediately decreases because the cat's out of the bag. At this point, the IV/SV ratio is likely to be < 1.0.