D&D.Sci December 2022: The Boojumologist

This is an entry in the ‘Dungeons & Data Science’ series, a set of puzzles where players are given a dataset to analyze and an objective to pursue using information from that dataset.

STORY (Skippable)

You were saddled with debt, and despair, and regret;

But you left it behind to embark,

With a visiting ship who were planning a trip,

Hunting some strange sea-beasts they call . . . “Snark”?

(After climbing aboard and departing the shore,

Your life is if anything worse.

The grog makes you groggy; the sea makes you soggy;

The songs leave you thinking in verse.)

Snark-hunting, you find, is a peaceful pastime.

By now, every crew knows the way,

To—with ease! - guarantee their success and safety,

As they seek, and they lure, and they slay.

A single exception proves the above rule:

While with most Snarks, Snark-hunting’s a breeze,

If your Snark is a Boojum, it hunts you right back,

And slays you with similar ease.

When you learn the detail that a Snark-hunt can fail,

You speak with the crew and insist,

On wielding Science skills to discern which Snarks kill,

And becoming the ship’s Analyst.

(The Butcher and Baker and Barnacle-Scraper,

All tell you that simply won’t do.

“You’ll be telling us what is a Boojum, what’s not;

Thus, you’re our Boojumologist; true?”

When you meekly protest you’ve not earned that address,

They rebut: “As you clearly percieve,

The Bellman can’t ring, the Belter can’t sing,

And the Beaver knows not how to Beave.”)

Notwithstanding contention viz naming conventions,

Your task is as clear as could be:

Using data from trips made by similar ships,

Choose which Snarks to fight, which to flee.

DATA & OBJECTIVES

  • You have a dataset of Snark sightings. This tells you whether a sighted Snark was hunted and whether – if hunted – it turned out to be a Boojum. (To clarify, the order of operations is “someone sights a Snark” “they take notes on its behavior and relay this information to the nearest Snark-hunting ship” “the ship’s crew decide whether to hunt it”).

  • You have a list of Snarks your ship can choose to go after. The ship must hunt six of them for the venture to be profitable; the crew won’t accept a plan targeting fewer than six. If your ship hunts a Boojum, everyone aboard vanishes, including you.

  • Your objective can be to choose the six Snarks which maximize probability of survival, maximize the EV of the number of Snark bodies in your possession at the end of the trip (i.e. max([Snarks hunted] * P(Survival))), or do anything inbetween.

  • (You could also optimize for killing the crew as quickly and reliably as possible, if you like. Or you could ignore the ‘challenge’ part of the challenge entirely, and just try to characterize the dataset. I’m not the boss of you.)

BONUS TASK

If you post a selection of Snarks before the deadline, and your selection is on the efficient frontier – that is, if you’re the first to submit that selection, and if no other submission has both better EV and better survival probability – you will be entitled to receive [UNSPECIFIED BENEFIT (UNDERWHELMING)] at [UNSPECIFIED TIME (DISTANT)]. Every player can make one submission; if you post multiple plans, please specify clearly which of these is for the Bonus Task.


I’ll post an interactive you can use to test your choices, along with an explanation of how I generated the dataset, sometime on Monday the 12th. I’m giving you nine days, but the task shouldn’t take more than an evening or two; use Excel, R, Python, recollections from a previous timeloop, or whatever other tools you think are appropriate. Let me know in the comments if you have any questions about the scenario.

If you want to investigate collaboratively and/​or call your decisions in advance, feel free to do so in the comments; however, please use spoiler tags or rot13 when sharing inferences/​strategies/​decisions, so people intending to fly solo can look for clarifications without being spoiled.