The Pete Wells Thread in New York Posted September 9, 2016 It's not immoral to apply Bayes's theorem. Like, yeah, the situation isn't great, but it's probably a valid inference that restaurants that don't recognize Wells are more likely to be bad than restaurants that do recognize Wells. Yes, exactly. Although inelegantly phrased, that's what Gordinier is likely getting at. Keller himself has a similar comment in the T&C piece linked above: One red flag was that his staff hadn't recognized Wells, who, unlike some past Times critics, isn't known for wearing disguises. "I told them, 'How can you miss a diner who comes in multiple times, even if they're not a critic?' " Keller says. "Somebody should catch that, because it's our job to track what people like if they come in more than one time. You have to make a connection." A good restaurant staff should recognize any diner who makes repeat visits, regardless of who that diner is.