Plattski Saves Halloween

We’re rapidly approaching Halloween, the second most popular holiday in the US after Christmas. But in this time of Covid, how can kids reasonably trick-or-treat? (For those of you unfamiliar with that American custom, kids dress in costumes and go door-to-door in their neighborhoods, extorting candy by threatening vandalism.)  Never fear, Plattski’s got it all figured out.

I don’t want to get close to a bunch of rug rat strangers, walking petri dishes in the best of times. Covid doesn’t appear to hurt them very much very often, but I sure don’t want it for myself.  I considered installing an Amazon Blink video doorbell by my driveway, so I could pre-screen the kids.  I’d allow only those wearing safe costumes to approach my house – masked surgeons, astronauts with sealed helmets, maybe deep sea divers.  Shouting cheerleaders or serenading opera singers would have to stay out by the street, and I’d deliver their treats via slingshot. It sounded like a whole lot of work.

And then it came to me – everything is virtual these days. Why not a virtual Halloween? I’d write an Azure web app called “Halloween Saver”. The kids would specify their home addresses. The app would show them a map, and they would choose their routes. Information about the kids’ age and physical condition, pulled from their social media posts, would determine the distance they’d be allowed to virtually travel.

Meanwhile, the hosts would specify the types of treats they’re dispensing – Snickers bars, Three Musketeers, that awful Good and Plenty, or (even worse) healthy celery sticks – and kids could see these while choosing their routes. I initially worried that such transparency might lead to gaming the system, concentrating on the houses with the best treats, but my daughter reports that kids already share this intelligence over their phones. Maybe a smart geeky kid could figure out an algorithm to maximize each kid’s utility, given their preference for specific types of candy and the availability within their traveling distance. That’s the classic traveling salesman problem, a great project to put on a college application.

Hosts could specify their spending limits. But if they run out, they could instantly sign up for more, and avoid unpleasant tricks. Of course, that means they wouldn’t have huge bags of leftovers to enjoy afterwards. Bug? Feature? You tell me.

After the kids specify their routes, and the hosts their bribes, a central processor will tally the haul for each kid – two regular-size Snickers and a fun size, four Reese’s cups, and so on. Amazon will do the  fulfillment.  The kids could choose either home delivery (free overnight for Prime members), or specify an Amazon storage locker where they could pick it up after midnight.

Halloween Saver would also support automated trading of candy.  I’ve always hated licorice Good and Plenty, but love fruit-flavored Mike and Ike. My friend Russell had the opposite preferences, so we’d swap, making both of us happier in the classic free market way. Each kid could specify what they’d be willing to give up, and what they wanted in return. Halloween Saver would use AI to implement a matching algorithm between bids and offers, like the spot market for electric power (see https://www.sciencedirect.com/science/article/pii/S1110016816303192). We’d establish a market clearing price for each type of candy (possibly through a universal reserve currency such as jelly beans), settle all the trades, and the results would appear in the final tally.  Easy-Peasy.

And that’s it, my friends. Our fall ritual proceeds unimpeded, perhaps even enhanced. Now if I can just figure out what to do about Thanksgiving …

Join the Conversation

1 Comment

Leave a comment

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: