Ken Jennings

Blog

May 26, 2016

A reader named Anthony writes:

I hope this message makes it to you. My friend and I are debating how one might estimate the number of trivia questions that exist. Or more precisely, how many distinct questions could be asked to the American participants on Jeopardy or in any trivia game, where the reaction is neither “that’s too easy to be considered trivia” or “no one except an academic in that field should know X”.

We tried several different statistical arguments like how often questions repeat themselves on Jeopardy and how many things a person learns in a day versus what % of trivia questions do they know. However, neither of us are really trivia experts, and we found ourselves challenged making a reasonable back of the envelope estimate. My best guess is half a million distinct trivia questions, but I could be off by orders of magnitude.

If you find this question interesting, I’d love to hear your take on it.

Are you kidding, Anthony? I’m probably one of the six or seven people in America who would find this question interesting!

Even boiling down your question to “askable on Jeopardy!” doesn’t quite solve your definitional problems here. Jeopardy! tends to keep its material fresh by adding cosmetic second clues. Are these two the same “trivia question” or not?

The oldest subway system in continental Europe is found in this Hungarian capital city.

In 1242, this city replaced Esztergom as the capital of Hungary.

I would say yes, for these purposes: they are both restatements of “What is the capital of Hungary?” But that underlying uber-question might not always be obvious to the layperson. And there are probably much trickier edge cases.

There’s also the problem of Jeopardy! clues where the response isn’t really something that anyone KNOWS-knows, but smart players should be able to intuit. These are especially common as Daily Doubles or in Final Jeopardy…take a clue like

He was born in 1728 in Yorkshire, England and died in a skirmish February 14, 1779 in Kealakekua Bay, Hawaii.

Now, the birth and death dates of [this historical figure] aren’t really knowable facts. It wouldn’t be fair game to ask “What were the birth and death dates of [Historical Figure]?” But the writers have carefully put in just enough information to trigger a well-informed guess from a smart player. I suppose you could restate all questions of this form as “What 18th-century Englishman was killed in Hawaii?” but there are endless variations, even here. Subtract the dates and add a quote from his diary. Withhold the place names but add more details about the cause of death. Are those really tests of the same fact? [Historical Figure]’s death, which took less than a minute of real time, can spawn almost endless trivia material in subtle shades of difficulty and variety.

Problems of definition aside, what is the order of magnitude here? Are there tens of thousands of facts that Jeopardy! can ask? Hundreds of thousands? Is it over a million? Thoughts?

Posted by Ken at 3:10 pm