Surely they've worked out (with far more information than we have) what they think is the optimum price on the supply-demand curve to maximize profits.
You're likely correct, they probably did some a/b-like testing to see how high they could raise concession prices before profits from concessions began to drop.
What I think they failed to look at was the long-term effects overpriced concessions had on the ticket sales. Concession prices have absolutely contributed to the long-term decline in ticket sales. I know lots of people that stopped going to movies because they knew their kids/SO would want that $15 drink and popcorn.
I only started going back when a theater that offered real food opened up. I won't pay $10 for a bucket of popcorn, but I'll definitely spend $15 on a decent burger.
If there are N people who would still buy at BIG_PRICE and M people who would buy if it's LESSER_PRICE and below, even if M > N, you would still be losing money if NBIG_PRICE < MLESSER_PRICE.
There's a demand curve, and if you lower below it you don't get more profit, you just sell to more people -- search for product elasticity curves, maximum willingness to pay, etc., for more info.
Vendors, and vendors like cinemas even more so, have experimented to find the price that maximizes those curves.