Meet the Real Masterminds Behind Jeopardy's Game Theory Tactics

This article is from the archive of our partner .

Arthur Chu, Jeopardy's latest and greatest star, has been everywhere since we covered his unorthodox tactics late last week. Between his out-of-order Daily Double hunting, his all-or-nothing approach to wagering on those Daily Doubles, and his Final Jeopardy strategy of going for the tie, Chu has become hero to some, and a villain to others with his four-day winnings of $102,000. With appearances on Mental FlossCNN, and Good Morning America, it's safe to say America is in full Arthur-mania. Most of the responses have been that Arthur "figured out" Jeopardy, or found a "hack" of the game.

Though he has become the face of game show strategy and game theory, Arthur readily admits he is not the mastermind behind those ideas. "I don't know that much about game theory myself," he said in an interview with The Wire. "I took one undergrad course on it." So where did these strategies come from? "I didn’t make them up. I liberally stole them."

Keith Williams. (Final Wager via Youtube)

Instead, meet the true mastermind of the tie wager: Keith Williams, a former Jeopardy champion who now runs The Final Wager blog. Every day since the beginning of this Jeopardy season in mid-September, Williams has taken to his blog to analyze the wagering strategies of contestants in Final Jeopardy and promote the optimal method of betting for the tie. Arthur mentioned Williams' videos specifically in explaining his preparation for handling Final Jeopardy wagers.

Recommended Reading

The impetus for Williams' blog and its wagering tutorials was mostly based on his general dissatisfaction with players' bets, he told The Wire. "I started it in part because I was getting frustrated about some of the wagers," he said. That frustration became more pressing when a friend from college was accepted to the show last February. To help her, Williams churned out a 20-page "manifesto," as he calls it, of the optimal method. "I tried to make it as simple as possible with addition and subtraction, and not focusing on any of these algebraic formulas that a lot of Jeopardy fans like to geek out over. [I] just made it very accessible for people," he said. Though Williams' friend still lost to a stronger player, she did make the correct wager amount in Final Jeopardy. 

That manifesto soon turned into his The Final Wager blog, filling a void for explaining Jeopardy's game theory strategies in simple, understandable ways. "All Jeopardy players who go on the show are very, very good at trivia," Williams explained. "They're not there because they're good at math, necessarily."

There was one major change between the manifesto and the blog, though. Williams' original idea encouraged the leader to bet twice as much as the trailing contestant, plus an extra $1 as a wager. That way, a correct answer in Final Jeopardy will lock out the trailer, clinching a win for the pre-final round leader with a correct answer. But upon more consideration, Williams decided that adding that $1 was not the optimal play, largely because a tie still counts as a win in Jeopardy rules. He cited the idea of "dominance" in game theory, that failing to add $1 would, at minimum, cause no harm. Tying for the win still sends you through to the next round. Saving that extra dollar has meant the difference between a win and loss in several cases, he said. And there is also a key psychological aspect to the strategy that increases the leader's odds of winning. (**For a more full explanation, see below.)

Though Arthur is getting most of the credit, Williams is more than happy to see his ideas in action. "I think it's awesome. That was my intent for people to come on and be prepared to wager properly," Williams said. "It's really amazing to see how pop culture has picked up on Arthur's story and by extension my story." Arthur-mania is easy to see in Williams' page views. Most of his video breakdowns get between 50-100 views; the video that explains why Arthur went for the tie on Wednesday is pushing past 36,000. Similarly, The Final Wager blog itself saw 30-times its normal traffic on Monday, Williams explained in an email.

Naturally, Williams is convinced of the importance of the Final Jeopardy wage value, one of the areas of the game where the contestant is totally in control. "A lot of people think it's a bigger part of the game than the trivia itself is, and I'm inclined to agree with them on that," he said.

Though he was one of the show's early stars, Jeopardy still
can't spell Chuck Forrest (Play Jeopardy via Youtube).

But Williams isn't the only Jeopardy mastermind that Arthur's strategy comes from. So, too, Arthur's in-game strategy — bouncing from category to category, searching for Daily Doubles in the more expensive questions — was borrowed from an early Jeopardy star. That strategy, called the "Forrest Bounce," is named in honor of former winner Chuck Forrest, who first employed the technique back in 1985. Forrest utilized that disorienting play to become the "Alexander the Great of 'Jeopardy!' players," the Los Angeles Times wrote in 1989, as he rattled off five straight wins, which was then the maximum allowed. Forrest claimed in his book Secrets of the Jeopardy Champions that the bounce idea was first proposed by his law school roommate. Still, the "Forrest Bounce" it remains.

By fortune's favor, Forrest actually appeared on Monday night's episode as part of Jeopardy's "Battle of the Decades." There, he of course employed his patented bounce, and right on cue, found the game's first Daily Double. Though he entered Final Jeopardy in second place, a correct answer coupled with an incorrect from the leader gave Forrest the big win. 

So while Arthur may be the face of the game show theory movement, he is certainly not the inventor of these ideas. But hey, he clearly chooses his advisors and strategy role models well.

**Here's a more detailed explanation, with help from Williams. Wagering with that extra $1 encourages the trailing contestant to wager an amount so that, if the leader guesses incorrectly, the trailer can win. Consider a case where the leader has $10,000 and the trailer has $8,500. Most often, the leader will bet $7,001 to get to a total of $17,001, or one more than double the trailer. The trailer, since he can't win if the leader gets it right, should then bet up to $5,500 to win if the leader guesses wrong. A correct answer from the leader will win; a wrong one gives the game to the trailer. This is how most Final Jeopardy wagering goes.


But, as Williams suggests, if the trailer knows the leader tends to wager to tie, the trailer might feel compelled to risk it all and tie for a win if correct. That then gives the leader a chance to win even if he gets the question wrong, if both the leader and the trailer answer incorrectly. In the case above, for example, a wrong answer from both would leave the leader with $3,000 and the trailer with $0. That extra possibility of winning is a big advantage for the leading wagerer to have, and can only happen with a wager-to-tie strategy.

This article is from the archive of our partner The Wire.