Ken Jennings, the 74-day Jeopardy grandmaster champion talks about Arthur Chu who is “gaming” the trivia contest by seemingly randomly hunting for the Daily Doubles.

In any game of Jeopardy!, three clues have been secretly earmarked as Daily Doubles. The player who finds each one can bet any or all of her winnings on responding to it correctly. By and large, Jeopardy! players are a risk-averse bunch. Unless a player is in need of a big comeback, the Daily Double wager is usually a smallish one.

Strategically, this is crazy. Like a poker player trying to increase the size of the pot when he has a good hand, Jeopardy! contestants should maximize their upside when the odds are in their favor. Historically, the odds of getting a Daily Double correct are very good: Between 65 and 70 percent. Too many players instead let games come down to Final Jeopardy, where conversion is much less predictable. (Fewer than half of all Final Jeopardy responses are correct.) Finding the Daily Doubles becomes more important the stronger a player you are, since it lowers the influence of chance on the outcome. Crunching some numbers, I see that my own Daily Double conversion during my Jeopardy! run was about 83 percent. In hindsight, my wagers were almost always too small.

Jennings goes on to look back at his own Daily Double stats and talks about the probability of finding one based on location. Personally, I’d love to see some big data research applied to every one of the games played in its 30 year history. I’d be interested to know what squares are most likely to be answered correctly, what subjects are most answered correctly, the average Daily Double wage, the average Final Jeopardy wage/success rate is, etc.

Going back on 30 years of data could prove interesting to future players and might actually give us some insight on how we behave.

Slate via Matt Jacobs