Superb if all are "re-accommodated"

The more you stay quiet and just "accept it", the better.  Domestic airlines scramble to maximize their opportunity to increase profits off the backs of everyday American passengers.  Foreign air carriers are no better.  The goal is to find a way to jam in as much revenue as possible, and at all costs.  Even if that means purposefully double or triple booking a seat, and then letting the ground staff sort it out once passengers arrive.  Even if it means flying your own personnel around at the last minute, but not without first calling in the “re-accommodation” militia to physically manhandle paid customers taking up seats you suddenly need for something else.  There are nearly a dozen moving variables in an over-booking statistical model, and it no longer attempts to populate the “correct” number of seats that will be used.  These models are pushing the envelope (and in turn the probability errors) that will occur when they overplay the oversales in specific situations where it is not meaningful to do such.  They also don't know exactly or ever how many volunteers they can jerk over with token monopoly money (turns out all carriers have far too high a number of people who simply go this route), and they certainly just hope they could contain any fall-out from the PR damage of the eventual involuntary bumps or police-enforced ejections.  This is the disturbing state of affairs that we are in.  It is legal (yet distasteful) and they can pad up thin margins at all cost (perhaps in the future they will charge a re-accommodation fee because why not?)  So game on; and poor statistical models that no one audits have ensued.  Curiously, some airlines are far better than others at controlling only parts of their oversales’ models, but this is not a customer service success.  A deep dive into the data shows that even these best-in-class regular-fare carriers are operating in the same mechanism that leaves them vulnerable to turns in local weather or economics or other surprises that temporarily (at best) thoroughly throw-off their overbooking projections.  A system that ignores consumer wishes, and unfortunately increasingly rely on overbooking bait-and-switch, is a problem that continues to simmer.  And on the other extreme, despite minimal complaints from frequently over-booked and involuntarily bumped low-fare airline passengers, those carriers simply take an even more extreme leap of faith to plug holes and use overbookings to smooth out passenger (revenue) into less desirable seats during the week that the customer never desired in advance.  Their probability models are hoaxed into trying too hard to place emphasis on these dozen or so random variables, and it results in large uncertainty -relative to seat capacity- on an on-going basis.  United Airlines is actually a better airline of this bad bunch, though of course the Dr. Dao incident this week evidenced that Americans want much better than to be taken for granted and dragged into evening TV just for peacefully standing their ground.  United Airlines is also not forgotten by the public community.  Its share price is still down a whopping 3% since the incident (while the broader S&P is down only 1% and United Airlines' competitors have barely budged.)  See these charts below that summarize the findings from a decade worth of scraped monthly airline and consumer survey operating data.

Lower capacity usage wastes money, and so out of greed we see those carriers push the envelope on over-bookings (and hence higher missteps there):

These low-fare carriers take their customers much more for granted, and those same customers rarely complain to their mistreatment: