It has been hours and one Bayesian Man is still trying to make up his mind. After a bad suggestion from one shop worker at Tallahassee Ice Cream shop Scooper Dooper, it is taking this Bayesian man forever to decide. “It’s hot and a busy Saturday night, why can’t he just decide,” stated one Tallahassee dad Steve Germaine with four loud children in tow. “We always go get ice cream after the movies, wanted to try out this new place. We’ve been waiting in line for hours! It’s almost too late and we’ll just have to pickup a gallon of Blue Bell at the Publix if this line doesn’t budge soon.”
Scooper Dooper has naively not instituted a rule for the maximum number of samples per customer. This controversial store policy has attracted sample abusers far and wide. While some Ice Cream consumers abuse the sample policy for nefariously greedy purposes, some like Bayesian Man use it to update their prior distribution for ice cream flavor most likely to purchase.
This isn’t the first time this Tallahassee local has taken so long. Weeks ago, after this Bayesian Man was taking a break from defining his posterior at the squat rack, he decided to celebrate with Ice Cream at the recently opened store. Scooper Dooper prides themselves in dozens of entirely unique flavors. With so many to choose from, the Bayesian man entered with no idea what he would choose.
“He said something like ‘I have a maximum entropy distribution of your ice cream flavors’ whatever the hell that means and just told him he could sample whatever he wanted to make up his mind,” Scooper Dooper employee Chet Herbert told reporters.
“Next thing I know, he’s sampling every single type of ice cream in a grid search. I knew the policy and let him sample everything but when he tried everything and asked if he knew what, he wanted said the Pistachio-Salted Caramel Decision Threshold was too high and asked for more samples. I said ‘Look buddy I’m running out of wooden spoons! You’ve got three more before I’m going to have to open up another package so you’re going to have to lower that threshold. He tried three more flavors and bought all 36 flavors except for the cotton candy, I guess he lowered his threshold a little too low!”
Two weeks had passed since the Bayesian man had first grid search sampled Scooper Dooper’s ice cream selection. When Chet saw the same Bayesian man approaching in line he stopped what he was doing to google some articles on google scholar and took a smoke break to watch a Veritasium video. “That’s when I had the bright idea to speed up his decision making process with a better prior. I figured he seemed like a mint chocolate chip so I started talking up that flavor. He feels like a mint chocolate chip kinda guy,” Chet said while handing even more ice cream samples to the Bayesian man. Later data revealed he was catastrophically not a mint chocolate chip guy.
“Chet sounded pretty convincing at first, but the more I try it the more I realize I’m not a mint chocolate chip type of guy,” the Bayesian man commented. “It ranked pretty high last time I came to Scooper Dooper, but the ice cream samples are not backing up Chet’s recommendation. The problem is, he convinced me so much that their mint chocolate chip was the best that I really am having a hard time deciding to get something else. I really want to believe in the mint chocolate chip and it’s distracting me from other flavors.” With extremely problematic acceptance ratios in the Bayesian Man’s Metropolis-Hastings Ice Cream Flavor model, he still had no idea what to buy even after a hundred samples.
Chet has been getting more and more nervous as the line behind the Bayesian man has grown longer. Many would have thought that the lack of movement in the line would have scared customers away. Unfortunately for the local ice cream shop, they appear to be suffering from the night club effect and the long line has attracted even more customers. Before the Bayesian man had enough samples to narrow down his search space, the line was already reaching around the store and across the street to the CVS.
“Yeah, this place is pretty new,” customer Mary Prater said. “It’s in that place in town where restaurants always seem to fail for some reason, I think it was a Thai place four years ago, then it got bought by a vape store, then now finally this. Maybe Scooper Dooper’s will stick around. If the line is this long, it must be good.”
Even though a sale has not been made since peak ice cream began at 6pm this Saturday, Scooper Dooper management has told Chet, the sole employee tonight, to hold fast and stick to the unlimited free sample rule. “We may lose our profits tonight but we will not lose our honor,” store owner and manager stated. “Our mission statement at Scooper Dooper is Great Ice Cream, Family Fun, and Integrity. It’s framed on the wall! We just have to weather this storm. Once the Bayesian man figures out what he likes, I’m sure he won’t need nearly as many samples to updates his preference and we’ll have a life long customer, and that’s a Scooper Dooper guarantee for everyone!”
[Update 9:48pm EST] Bayesian man has decided to get a double scoop of the Chocolate Chip Cooky Dough with caramel drizzle in the waffle cone.
If you enjoyed this fake news story about a metaphorical man driven by bayesian statistics, thanks for being a nerd and please like, share, and subscribe with your email, our twitter handle (@JABDE6), our facebook group here, or the Journal of Immaterial Science Subreddit for weekly content.
If you REEEEALY love the content, for the equivalent price of a Chipotle Burrito, chips and Queso, you could buy our new book Et Al with over 20 hand picked Jabde articles for your reading pleasure, it’s the perfect Christmas/Birthday gift for confusing your unsuspecting family members! Order on amazon here: https://packt.link/at4bw Please rate and review so that you can brag to your friends about having opinions or showcase your excellent taste in reading material!
