Jeopardy's controversial new champion is using game theory to win big

[gdahl]

Here is a better article with an interesting interview from Arthur Chu: http://mentalfloss.com/article/54853/our-interview-jeopardy-...

[syllogism]

Yes! This is exactly what I was hoping for in the comments. I found the article intriguing but could see that there was a lot more to say.

Having read the interview, what I really take away from this is a rationality success story. Chu kept a simple focus on what he was trying to achieve, and worked to maximise his chances.

[scarmig]

Of course, the real game theoretic outcome has to include the incentives of the rule maker.

This strategy generates controversy, which generates viewers, which generates revenue. But once the novelty has worn off, there's nothing prevention g the rulemakers from instituting a formal rule to replace what was once a cultural norm.

[pyduan]

Well I would say that as far as Chu is concerned, any long term consequences his actions may have on the game are likely to be mere externalities to him since these will probably not happen until he's already extracted the value he wanted out of the game.

You could also argue that the additional publicity he's enjoying thanks to his unusual strategy (eg. this article) far outweigh the costs of any such potential negative externality.

[AznHisoka]

Game theory? It's all a moot point unless you actually know most of the answers which it seems he does. Game theory is just a tactic to give him that extra edge.

[josephagoss]

It's not a moot point, assuming he knows as much as his opponent this technique is giving him an advantage. (Which is what makes him and his technique interesting.)

What surprises me is apparently he is the first person to play the game like this.

[xatnys]

Forcing a tie is an interesting approach, but daily double hunting is not new to the game at all. If you observe the way the contestants play, it becomes obvious who is playing with a strategy and who is just running down the categories. As for the ties, there's nothing wrong with it (honestly it's a good play), I only feel bad for those contestants who end up waiting longer as the queue becomes staggered.

[saraid216]

There's a difference between maximizing entertainment (which is how Jeopardy! is designed) and maximizing victory (which is how Arthur Chu is playing). In well-balanced games designed for competition, this is a non-issue. This does not describe Jeopardy! (or virtually any other game show).

[mikeleeorg]

I've always wondered why contestants don't go looking for the Daily Doubles first. It always seemed like a good strategy.

And now that his strategy has been publicized, it's only a matter of time for other contestants to try the same thing.

[simple1]

Having all players move on when there is a tie has always baffled me. If any three contestants ever agreed to always lose down to the lowest score in final Jeopardy, they could all win forever until the rules changed.

[maxerickson]

I imagine the agreement between the studio and contestant includes big loopholes giving the studio discretion to do whatever they want.

[fsk]

Offering a tie in Final Jeopardy wagering is a good idea. If you beat the opponent already, why not play him again instead of a random new opponent who might be better?

[RK]

I don't know that much about Jeopardy strategy, but it sounds like he's following Ken Jennings' suggestions.

[ilaksh]

But can he beat Watson?