This is covered in either Fooled by Randomness or The Black Swan, and is pretty basic probability, actually. The Big Short has a section on a fund that made a crapload betting on situations w/options that were basically yes/no, that is, the underlying was going way up if things worked out one way, and way down otherwise, whereas options are priced as if there's a range of probabilities across strikes. Interesting stuff, and it really almost amounts to a manual on how to trade options successfully by assuming Black Scholes is indeed BS and given you've researched the underlying properly. Anyway, the thing I find weird is the use of weeklies. In the short time I spent researching IC's, I very quickly found the weeklies to be extremely useful for hedging against losses in IC's, but found the probabilities were way against you if you were net short them. That's what's getting me about this whole thing.
If you're talking about the weeklies comment, it was in relation to you repeatedly saying you find that you can make incremental money on them. I always closed out before the last week of expiration, myself, when I was fiddling with IC's a few months ago. When weeklies came out, I experimented with them on the short side a bit, but then decided they were good for, to quote myself, "a cheap hit of high gamma", and are far more effectively, not to mention safely, used for that purpose. High theta and high gamma go together. I think if you asked options coach, Maverick 74, and atticus, among perhaps others, what they find wrong with this strategy, most of it is probably covered by that sentence. Most people get theta and delta pretty quickly, but have trouble with vega and gamma. I had trouble with vega, but gamma was something I always understood. As I've noted before, if I did a collar on a stock that had an earnings report coming up, I was always careful to use the front month for the put side, because they had the highest gamma. If you're hedging against disaster, gamma is your friend. But being short high gamma options isn't really a good thing.
In general, by the way, Black Scholes (Bachelier-Thorp, if you're Taleb) is fiction. A useful fiction for quick and dirty risk profiling, which according to Taleb is why the formula was derived in the first place, but a fiction nevertheless. IC's depend on it being true. That's another big knock against this strategy.
I have done weeklies consistently since 10/18/10. One of the losses was an unforced error. I failed to properly set up the circuit breaker. PHP: Credit Spreads Results View Winners 34 Average win 5.2% Lossers 3 Average loss 21.5% Expected value 3.1% Average life 4.1 days The majority of spreads were paired with companions to create Iron Condors so the return was better than is indicated by looking at them just as spreads alone. PHP: Iron Condors Results View Winners 10 Average win 10.9% Losers 2 Average loss 13.4% Expected value 6.9% Average life 5.0 days I have found weeklies no more difficult to manage that the last week of a monthly. The last week is riskier for the reasons you mention. Because they are riskier I trade them in small money mode. Nonetheless, they have contributed significantly to the monthly returns.
Mathematically that's highly improbable. Definitely not something I'd do at this point. If you haven't read it, pick up The Big Short and read it. The part about the guys who made money on options should be something of an eye-opener on the whole subject of the real probabilities in real life of options. In my case, the most money I ever made in a single day in my entire life was the Monday after Bear was sold to JP Morgan for 2 simoleons. I had a strangle on Lehman because they were going to report earnings that Tuesday, and wound up winning the lottery completely by accident. An object lesson in gamma, vega, and, if you were to reverse what happened that day, the dangers of selling options, naked or not.
My largest win ever in %/$ was in short g/v. I have nothing against selling gamma, but take issue with passive strategies involving selling garbage. WINGS ARE MEANT TO BE BOUGHT All of the convexity on greeks that blows people up involves the deep otm junk. The difference between a short straddle(strangle) and a butterfly(condor) are the LONG WINGS. The OP is far better off selling atm gamma naked or covered. Naked is safer simply due to the fact that the haircut is stiff enough to limit size. These terrible r/r spreads are costly due to the aforementioned(r/r). Unfortunately, when you trade outside your strikes you have nothing left or reload with another dozen buy-ins like some degen NLHE player. Sell the body, buy the wings. Options 101.