What do you think of this betting "rule"? (June 1 SPOILERS)

[jgroub]

Today's DD bet by Joe is an example of a betting "rule" I came up with; I'd love to hear what you all think about it.

If you are significantly behind in the middle or late in DJ, it means two things - 1) that you've been regularly getting beat, and at this rate, the trend will most likely continue, and you'll continue to get beat through to the end of the game; and 2) that there isn't much time (money) left to make up for the lead.

Therefore, in the situation that Joe was in, with only 10 other clues left on the board after the DD, he's gotta bet BIG. Not just to catch up, but to take a nice-sized lead, too. Because the trend that that other player (in this case Coby) will continue to beat you on the rest of the clues is likely to continue, and you always want to go into FJ with the lead to control your destiny.

In other words, Joe's bet was really, REALLY good for that situation, especially in a topic that seemed pretty easy. Maybe not $9400, though; maybe like $8800 would have been enough, but now I'm just quibbling.

What do you think?

[AndyTheQuizzer]

FYI: I added a spoiler tag in case people click on this without realizing it's for tonight's game.

[Wheatley]

I think it depends on your confidence level of the category and of the ones left - I can name an infinite number of categories where I'd make a DD wager of $5 no matter what

[jgroub]

I think it depends on your confidence level of the category and of the ones left - I can name an infinite number of categories where I'd make a DD wager of $5 no matter what

Heh, heh, Opera and Ballet fulfilling that role for me!

What I'm saying here though, how many times have we all seen someone in Joe's position make a $2000 wager? Or even a $4000 wager? And I think - subject to your caveat, of course - that you've gotta go as big as Joe went.

[jeff6286]

You're not wrong, but you're sort of overlooking something big, saying you wouldn't go as high as he did. If he goes anywhere near $9400, then there is zero reason not to bet it all, because he's going to be out of the game on a miss anyway. When I saw the thread I thought the point you were going to be making was that it was foolish to leave $1200 on the table, since Coby still had plenty of time to come back with the amount of money still left on the board, and could have taken a slim lead, meaning that the missing $1200 Joe didn't wager could easily cost him the game.

[Ursa Minor]

Here's how I think about this, vis à vis what I did on the program back in 1987:
If the probability of getting a DD right -- p(DD) < 50% -- bet the five bucks, because P(FJ) for me is about 80%, so the two combines are less than 40%.
If p(DD) > 50%, bet enough to take the lead going into FJ.

And be prepared to spend 30 years of remorse. With less than a minute to go, my DD category was mountaineering, about which I know almost nothing. Pitons and carinbiners and such. It was the $1,000 (this was then) question. I figured my chances were 10% at best, so I bet small--enough to go into FJ with a mathematical chance. And to my everlasting annoyance, I got the dambthing right. What does USGS stand for on a map? FJ was a no-brainer, and that was the end of my Jeopardy! career.

[This Is Kirk!]

You're not wrong, but you're sort of overlooking something big, saying you wouldn't go as high as he did. If he goes anywhere near $9400, then there is zero reason not to bet it all, because he's going to be out of the game on a miss anyway. When I saw the thread I thought the point you were going to be making was that it was foolish to leave $1200 on the table, since Coby still had plenty of time to come back with the amount of money still left on the board, and could have taken a slim lead, meaning that the missing $1200 Joe didn't wager could easily cost him the game.

I think you'd also need to consider where the third player was in this situation. If they were in a situation where the leader may have to bet big on FJ you might want to leave a little back in hopes of "playing spoiler" if you miss the DD.

[Volante]

You're not wrong, but you're sort of overlooking something big, saying you wouldn't go as high as he did. If he goes anywhere near $9400, then there is zero reason not to bet it all, because he's going to be out of the game on a miss anyway. When I saw the thread I thought the point you were going to be making was that it was foolish to leave $1200 on the table, since Coby still had plenty of time to come back with the amount of money still left on the board, and could have taken a slim lead, meaning that the missing $1200 Joe didn't wager could easily cost him the game.

EV calculations be damned, I would not wager myself out of FJ that close to the end.

[Anachronism]

Could be one of those unwritten rules that you shouldn't leave just $5 on the board if you miss a late DD. Bet it all or don't go that close. I'd be tempted to leave the $5 and stand there with the mask of shame waiting for FJ (then bet $4 on FJ and have Alex make some sort of remark about "strategy"). Yeah, I honestly see no strong reason to leave anything at all on the table if you are down to the last handful of clues and missing means no chance at winning. Go down swinging, even if the category is Bible quotes about ballet danced to '60s music.

[jgroub]

I think you'd also need to consider where the third player was in this situation. If they were in a situation where the leader may have to bet big on FJ you might want to leave a little back in hopes of "playing spoiler" if you miss the DD.

Yes, that's exactly why I was saying not to bet it all - you might still have a shot in FJ if the other two miss and you get it right.

[slam]

It's a little more complicated than you're describing depending on what you mean by "significantly behind". Let's assume for the moment that you're in 2nd place and 3rd place is effectively out of the game with a very low total. You need to take into consideration the 2/3 level of leader's score.

For example, if leader has $18,000 and you have $13,000 and there's only a $400 clue on the board, you're probably better off $700 or less. That way even if the leader gets the final clue, you'll be above 2/3 and just need the leader to miss FJ (assuming he/she makes the normal shutout bet). You're usually better off relying on the leader to miss FJ than on you getting both the DD and FJ (though you're estimate of your chances on the DD with known category and FJ with unknown category are relevant). The multiplicative property of probabilities is very powerful.

I'm just playing devil's advocate here. If you're further behind than 2/3, you very likely do want to follow your proposed rule. There are so many times that I've seen the trailer get the DD and make a bet that is WAY too small. So, in principle at least, I agree with your rule.

[Wheatley]

Could be one of those unwritten rules that you shouldn't leave just $5 on the board if you miss a late DD. Bet it all or don't go that close. I'd be tempted to leave the $5 and stand there with the mask of shame waiting for FJ (then bet $4 on FJ and have Alex make some sort of remark about "strategy"). Yeah, I honestly see no strong reason to leave anything at all on the table if you are down to the last handful of clues and missing means no chance at winning. Go down swinging, even if the category is Bible quotes about ballet danced to '60s music.

It's the human factor. I'd feel like my Jeopardy experience was incomplete if I didn't get to play an FJ. Plus, I could end on a high note by showing I knew the answer and would have won the game if I got that DD right.