This is NOT options question. I asked myself the question and did a quick Google search. Found below answer which I liked. Curious if others think about it in same terms given that mean reverting strategies are very popular.
"Mean reversion" implies a "Law-of-Large-Numbers" central tendency, no matter the particulars -- that means you have a primary measure of rate-of-change ("delta") and a rate-of-change on that primary ("gamma"). An interesting conflation. Works fine, though.
I like to keep things simple. Long options are long gamma and short options are short gamma. If I buy a straddle I'm long gamma. IMO, mean reversion has nothing to do with gamma, and the question "Are mean reversion strategies short gamma" are linking the two in a way that implies that somehow gamma has some unique relation with mean reversion - it has no more a relation than delta, theta, vega. To me, mean reversion applies mainly to volatility. And I trade it when the conditions are right. When VIX was above 20 in Aug, I traded mean reversion in many different ways - some of them involving going long, and some short. I sold credit call spreads; I bought debit put spreads; I did ratios; I even did a couple of calendars. All these strategies have opposing Greeks - some long theta some short, some long delta some short, and some long gamma some..... There is more than one way to trade mean reversion and selling straddles is just one such way. Hence, I disagree with Mr Vladimir Novakovski.