Monday, March 24, 2014 Game Recap & Discussion [SPOILERS]

[Fishercat]

Infuriating wagers from the two guys on FJ. A rational wager from either of them wins the game.

Very tough FJ. I went Groucho, not even the right group.

I agree with Bamaman, Joe's bet is defensible; Derek's is not.

Nancy Akerman: $13,200-$10,401=$2,799
Derek Arnold: $11,800-$11,800=$0
Joe Morse: $6,400-$6,400=$0

Derek needs to be at least $1,000 to cover Joe, so he might as well bet at least $1,400 to cover a possible $0 wager from the leader, therefore on a miss he is likely to drop to $10,400 or below. If Joe considers this, then he needs to bet at least $4,000 to give himself a chance to win on a rational wager by Derek. Betting $4,000 and missing drops him to $2,400, and since Nancy's MSBIW score is $2,799, I have no issue with Joe just betting it all here to maximize his chances of winning on a get.

Now sure, if Joe has watched as many games as many of us, he could have foreseen the suicidal all-in wager from second, and bet something between $0 and $3,600, leading to victory. I certainly don't blame a player for not making this assumption. Joe's big mistake was going to the top of the last category, when he should have been trying to find the last daily double, which turned out to be in the $1200 box. After that mistake, he was pretty fortunate just to have a chance to win in FJ, due to Derek missing the DD and Joe picking up the final $2000 clue.

Certainly a fair argument. It'd be great if we had stats on how often the suicide wager occurs. I wonder how often Derek (hypothetical Derek) does make the rational wager though. What's more common: the triple stumper or single-get with a suicide wager or a single-get (from third, albeit he was a strong third place player considering he lost 7k on a daily double) with a rational wager?

[PeteMoss]

Several problems with tonight's game's clues.

1. Haiti is not an island.
2. 1:2,500 does not mean 1 inch equals 2500 miles.
3. A heart is always used to mean 'love', and "Faith, Hope and Love" is a very common expression (found in a very popular passage in the bible), so HUGE negbait.

I think there were others too.

Brian

1. Since they accepted Haiti as a legitimate answer, I wonder if they would have accepted Dominican Republic?

[This Is Kirk!]

"The orangutan is indigenous to these two islands."

"What is Indonesia?"

"OK, we'll accept that."

[John Boy]

All I could think for FJ was Yellowhair.

Makes two of us.

Me three.

I knew the "Faith, Hope and [Agape]" clue would be a trap. Nearly every translation other than the KJV renders the third term as love rather than charity. That ambiguity knocks the clue down to below the usual J! standard.

Me four. I had heard of Custer being called "Yellowhair," never Goldilocks or Blondie or Curly. This whole clue sounds like b***s*** to me.

I couldn't imagine a Marx Brothers name for Custer; said "Bud" (as in Abbott) only because it was better than no answer at all.

I agree with several comments on this board:

1. "Haiti" is not an island and the answer should have been counted wrong.

2. "Love" instead of charity should have been accepted, as the trio is almost always rendered that way in translation and in common usage.

3. Derek's wager? What else can be said? That's the seventh time this season that an all-in or nearly-all-in FJ wager lost a game that would have been won by a strategically sound one. Pathetic.

[cheezguyty]

1. Haiti is not an island.

Should he have received credit for Haiti when they wanted Hispaniola?

Absolutely not! That kind of stunned me when Alex made that call.

1. "Haiti" is not an island and the answer should have been counted wrong.

"Haiti" was the U.S. government's official name for the island of Hispaniola until the 1930s.

[Magna]

The heck do bears have to do with Switzerland?

I think they played an exhibition game there once.

[davey]

The first biography of Custer that I found on Google Books, by Jay Monaghan, calls him “Old Curly” several times, and suggests it was a popular nickname.
However, the only Curlys I can find (using Amazon) in Evan Connell’s Son of the Morning Star are Crazy Horse – Connell says “Curly” was his name before he earned his warrior name – and a Crow Indian scout for Custer by that name.

Like others, I don’t think of the Stooges as a ‘30s team, but still there are a finite number of movie teams that come to mind, and I know Custer’s flowing locks, so I put Curly down when I could think of nothing better …

[opusthepenguin]

1. "Haiti" is not an island and the answer should have been counted wrong.

"Haiti" was the U.S. government's official name for the island of Hispaniola until the 1930s.

What's your source on that? My Google-fu has deserted me. Assuming it's correct, I withdraw my objection.

[opusthepenguin]

I wonder how often Derek (hypothetical Derek) does make the rational wager though.

I've started to wonder how often hypothetical Derek realizes he threw the game away as he's trying to get to sleep that night. I'd bet for a lot of contestants, that's the first time they ever think through wagering scenarios.

[lieph82]

I wonder how often Derek (hypothetical Derek) does make the rational wager though.

I've started to wonder how often hypothetical Derek realizes he threw the game away as he's trying to get to sleep that night. I'd bet for a lot of contestants, that's the first time they ever think through wagering scenarios.

I'd bet many contestants don't think about it then, either.

[aeq5006]

They specified that they were talking about Forest Whitaker in the top clue, but in all the other clues, they just said "Forest". Anyone else find any irony in the fact that if the contestants had jumped around in the "Forest" category, rather than going from top to bottom, they might have ended up confused?

Sigh. Get's depressing after a while.

I disagree! A get is never depressing!

If you don't want the marriage to end, it is.

[El Jefe]

I wonder how often Derek (hypothetical Derek) does make the rational wager though.

I've started to wonder how often hypothetical Derek realizes he threw the game away as he's trying to get to sleep that night. I'd bet for a lot of contestants, that's the first time they ever think through wagering scenarios.

Wonder not! (Our sister community?) The Jeopardy Fan features evidence of his stewing:

http://www.thejeopardyfan.com/2014/03/w ... r-chu.html

ETA: Never mind, it was Joe's Reddit AMA regrets I was remembering...

[Mathew5000]

It'd be great if we had stats on how often the suicide wager occurs. I wonder how often Derek (hypothetical Derek) does make the rational wager though. What's more common: the triple stumper or single-get with a suicide wager or a single-get (from third, albeit he was a strong third place player considering he lost 7k on a daily double) with a rational wager?

The past few seasons, the frequency of a single-get from third is about 10%, although in the current season it's only 5.6% (7/126). The frequency of a triple-stumper is more than twice that of a single get from third.

Probability of a suicide wager from second in a game like this? It depends how we define "suicide wager from second". You can define it as any wager greater than 3B-2A (i.e. he probably finishes behind first when they are both wrong), or you can define it as any wager greater than B-C (i.e. if wrong he finishes behind third's pre-FJ score).

Scores are Nancy 13200 ; Derek 11800 ; Joe 6400.

Here's how I would do the math from Joe's position: estimate there's a 60% probability that Derek will wager greater than 3B-2A=9000 and an 80% probability that Derek will wager greater than B-C=5400. Also estimate that 15% of the time Nancy will make a small bet, let's say smaller than A-C=6800.

So based on all those estimates (and a few more that I won't state explicitly), what is Joe's probability of winning (I.) if he wagers 0? (II.) if he wagers 6400? and (III.) if he wagers 3599?

I. Joe wagers 0. Wins if Nancy wagers ≥ 6800 and Derek wagers ≥ 5400 and they are both wrong:
0.85 * 0.8 * 0.3 = 0.2

II. Joe wagers 6400. Wins on a sole get if Nancy wagers ≥ 400. Also wins on a double get by Derek and Joe if Derek wagers ≤ 1000.
(0.1 * 0.98) + (0.07 * 0.1) = 0.1

III. Joe wagers 3599. Wins on a triple stumper if Nancy wagers ≥ 10400 and Derek wagers ≥ 8999. Also wins on a sole get if Nancy wagers ≥ 3201 and Derek wagers ≥ 1801.
(0.2 * 0.8 * 0.6) + (0.1 * 0.9 * 0.9) = 0.18

Based on my analysis, Joe should wager 0, giving him about a 20% chance of winning.

[JeopRDFan]

Also missed FJ, going with 'Harpo' just to have something. I never thought of The 3 Stooges as being a 30's comedy act as others have stated -- most of their best stuff was done in the 40's and 50's, imo. "Why Soitenly!"

[Vanya]

Also missed FJ, going with 'Harpo' just to have something. I never thought of The 3 Stooges as being a 30's comedy act as others have stated -- most of their best stuff was done in the 40's and 50's, imo. "Why Soitenly!"

You mean they weren't done in a TV studio in the 60s? Another childhood illusion shattered.

[opusthepenguin]

I wonder how often Derek (hypothetical Derek) does make the rational wager though.

I've started to wonder how often hypothetical Derek realizes he threw the game away as he's trying to get to sleep that night. I'd bet for a lot of contestants, that's the first time they ever think through wagering scenarios.

I'd bet many contestants don't think about it then, either.

I definitely wouldn't take the other side of that bet. I wouldn't even bet that the the majority think about it. But I think a lot do.

Anecdotally, I recall one third-place-before-FJ contestant who checked in here. He said he lay in bed that night musing over his own loss when it hit him that the second-place contestant had over-wagered and cost herself the game. It seemed pretty clear this was a new line of thought for him. Given a single game with numbers he knew by heart and an outcome he cared about, he was able to work it out. He had trouble getting to sleep that night. He was content with his own loss because it couldn't have been prevented. But he felt terrible for the other losing contestant.

I suspect other contestants come to a depressing conclusion when they try to justify their wagers--either in their heads or to friends and family. Many more, I'm sure, do not.

[Vanya]

mbclev does it for all of them anyway.

[Stanislaus Jacob]

I instantly thought "Curly" for "Custer nickname" and then was almost thrown by the rest of the clue. (Groucho? Harpo? Oh, wait, Curly of the Three Stooges!) So whether the nickname is actually accurate or not, it has been disseminated widely enough to have burrowed into my head at some point.

[Woof]

I got stuck on "Son of the Morning Star" for Custer and then struggled with the comedy team TOM. Got to the Marx Brothers and briefly considered the Stooges, but never got to Curly, sadly. Not the J! writers' finest moment IMO.

[chalupabatman786]

It'd be great if we had stats on how often the suicide wager occurs. I wonder how often Derek (hypothetical Derek) does make the rational wager though. What's more common: the triple stumper or single-get with a suicide wager or a single-get (from third, albeit he was a strong third place player considering he lost 7k on a daily double) with a rational wager?

The past few seasons, the frequency of a single-get from third is about 10%, although in the current season it's only 5.6% (7/126). The frequency of a triple-stumper is more than twice that of a single get from third.

Probability of a suicide wager from second in a game like this? It depends how we define "suicide wager from second". You can define it as any wager greater than 3B-2A (i.e. he probably finishes behind first when they are both wrong), or you can define it as any wager greater than B-C (i.e. if wrong he finishes behind third's pre-FJ score).

Scores are Nancy 13200 ; Derek 11800 ; Joe 6400.

Here's how I would do the math from Joe's position: estimate there's a 60% probability that Derek will wager greater than 3B-2A=9000 and an 80% probability that Derek will wager greater than B-C=5400. Also estimate that 15% of the time Nancy will make a small bet, let's say smaller than A-C=6800.

So based on all those estimates (and a few more that I won't state explicitly), what is Joe's probability of winning (I.) if he wagers 0? (II.) if he wagers 6400? and (III.) if he wagers 3599?

I. Joe wagers 0. Wins if Nancy wagers ≥ 6800 and Derek wagers ≥ 5400 and they are both wrong:
0.85 * 0.8 * 0.3 = 0.2

II. Joe wagers 6400. Wins on a sole get if Nancy wagers ≥ 400. Also wins on a double get by Derek and Joe if Derek wagers ≤ 1000.
(0.1 * 0.98) + (0.07 * 0.1) = 0.1

III. Joe wagers 3599. Wins on a triple stumper if Nancy wagers ≥ 10400 and Derek wagers ≥ 8999. Also wins on a sole get if Nancy wagers ≥ 3201 and Derek wagers ≥ 1801.
(0.2 * 0.8 * 0.6) + (0.1 * 0.9 * 0.9) = 0.18

Based on my analysis, Joe should wager 0, giving him about a 20% chance of winning.

I really like your thinking, but your conclusion is iffy at best. The difference between 0.2 and 0.18 is minuscule, and if you fiddled with your initial estimates even a tiny bit you would come to the opposite conclusion.