Judging the results on 2 examples over a 1-week time frame is unproductive. Most option traders have a longer market outlook than 1 week. Do you believe that a long ATM weekly GOOGL straddle can be bought for about $11.10 and exited at about $18.00 on a weekly basis? From your post you seem to think so. Over time (a few months) and a few trades both the long and short ATM straddles are dead money (zero-sum game) minus the bid/ask spread and commissions. #1 through #6 in the OP. Very active thread with lots of participants, including you.
OP, it would help if you clarified your intentions. Being wrong is ok, I (we) might still learn something yet. You have a couple of broken trades in a short period of time. Shrugging them off won't work. Try being short the spread X 100. Then tell me how you would have felt Saturday morning.
My thread originated from a post by J_Smith about predicting a stocks close on expiry based on option open interest and option volume. IMO ........ the ATM straddles are a much better indication of were a stock is most likely to close at when the options expire. Broken trades? Those aren't trades in the OP.
Ok volume and open interest. But you didn't say that. If so it's interesting. Do you want to repostulate?
Ok the short straddle closed quite a bit above its upper band. Whatever you want to call it, that trade didn't work. When questioned about it, you shrugged it off. Yes or no?
It's not a trade. I'm using the cost of the ATM straddles to get a trading range of the underlying during the life of the options AND the expected close on expiry, then base a real trade on that info. Perhaps sell an Iron Condor outside of the trading range, buy a Debit Spread, buy some ATM calls only, etc. My trading and market outlook is much longer than 1 week.
If you are saying there is a skew away from the strike, I'd love to see that over time. Put it together w volume and open interest, maybe there is something. Usually it ends up as a bell curve centered around the strike, I'm guessing. So option sellers would be favored, with the occasional blowout.
So let me ask again, in this case, what's the point of this thread then? I believe nothing other than it's relatively easy to do some basic analysis using historical data. I am reasonably sure that any such analysis will show that you're talking bollocks. Sadly, yes, including me... I'm regretting it already.