Flashing back to 2010...when there was a large, vocal distaste for Pay the Rent.
It's the only pricing game I can think of where deception was a principal part of the game's design and very knowingly exploited for effect.
There's a big difference in a game being setup for a loss and the type of deception historically involved in Pay the Rent. A car with a price of $19,191 in Lucky Seven is definitely put there to be lost with the way most contestants play the game, but it's not as if a contestant applying a little logic and skill (or just outright studying the cars on the show to know the different price combinations with different options by checking against the manufacturer's suggested retail prices) cannot win the game; or viewed another way, it's not as if the game is usually going to be lost right out of the gate on the first action by the contestant.
Contrast that with Pay the Rent, where the perception si "put the least expensive item in the mailbox", and while I suppose you can engineer a setup where this can be true, the fact is that it so often wasn't true (and a contestant could be smart enough to do the math to figure this out, but let's face it: Most of Price's contestants since the '90s haven't been the brightest bulbs); and the moment the contestant made his first pick, the game was already over. Golden Road or 3 Strikes may be lost more often than won, but at least there is (usually) some element of buildup in excitement rather than being brought to a losing conclusion immediately. Eventually the powers that be realized that having a game that has no random element to it, such as the bouncing of a Plinko chip, never being won makes for poor television and started to at least engineer some possible wins.
A $60 Raspberry Pi could be programmed in 10 minutes to operate Magic # if that were the true problem.
Considering all the other stunts they have pulled, including "superfans", why not put it to the "superfans" to fix Magic #? Make a contest out of it to get around union rules (though if it really is that simple, then you shouldn't need a publicity stunt to fix a pricing game).
I thought it was retired because because 4 digit prizes were getting to be quite a bit above the $1000 to $2000 for the lower priced prize and contestants weren't going high enough to get into the magic range (even when the range between the two prizes came to be absurd). That and/or it took too long with nothing happening while someone held the up lever.
Then start the game at $1,000 or whatever opening bid you want a la Card Game. Really, if contestants are not going to apply a little thinking and a little patience to wait for the number to climb and get it somewhere above the price of the lower-priced prize, then you can't hope for them to win. It always amazed me how many contestants barely pushed the number past $1,200. It seemed to me contestants seemed to think they were supposed to set the magic number as
the difference between the prices of the two prizes (a tall order, which really wouldn't make sense). There's more "Range Game syndrome" happening here than there really ought to be.