I can definitely understand brain farting that question, I didn't come up with it either. I was like January, no, February, no, March, no, oh someone rang in, wrong, what, where was I, oh, someone rang in again, wrong again, ok, January, February, March, April, May, June, July, August, times up.

Jesus ****ing Christ at that last DD wager.

*headdesk*

I am so tilted right now

If i cost myself $28,000 on national TV like that, I'd never leave the house again.

So, I wrote a markov chain monte carlo sim of Jeopardy to analyze DDs over the weekend, and this is as good a time as any to show it itt.



Would have been better off betting a thousand, lolololololol.

What am I looking at here? Fourth-year probability was a long time ago, lol.

Winning percentage on the y axis, wager on x. This situation is sort of boring, it's pretty obvious to everyone but the guy who matters that you should bet for the lead, but it's sort of funny that he didn't even choose the second most optimal wager, he chose the third.

Dudd-the small uptick around 1,000 has to be getting into the 3rd place situation where 2nd can longer bet to match 1'st and still cover 3rd betting 0? (12,700 for 2nd and 15,200 for 1st). Basically getting in position to win a triple stumper as well as being on the only player to answer FJ correctly (WWR).

The 1000 uptick is because it's a big enough bet to move you into position to win on a triple stumper if you get it right, but it's small enough that you can still win on a sole get in FJ if you get it wrong. Every other wager is either too small to get you into triple stumper territory if you get it right, or it's so big that if you get it wrong, you're locked out from catching second place. It's really not all that interesting a situation, shoving is still obviously correct, but I can simulate non-last question DDs with this too. Which, is actually pretty boring, Jeopardy strategy pretty much comes down to make every wager a true DD unless you have a big lead late in FJ. The way to maximize your win equity is to get to FJ with it locked, and making your DD's as big as possible are the best way to do that. I'm 99% sure Roger Craig did this same thing and figured it out, hence why he bets it all every time.

Chris, so sad. Funny that Jeopardy felt it needed to change FCC head's quote from "My day is going to suck tomorrow" to "My day is going to" be lousy "tomorrow."

grunching as i just got home.
just watched the episode where GM was the FJ answer.

i thought that was insanely easy FJ question and was SHOCKED that both 1st n 2nd got it wrong.
of course that could be cuz i had no idea that sears declared bankruptcy (if they even did) i just remember hearing so much about the GM deal that i couldn't imagine anyone guessing anything else.

i did call that lady doing the true daily double as soon as she got the DD in DJ.
guess i can read tells on people's face

wait WTF HOLY **** did he really just risk 7K???????
omg what in the **** ???

Yea

What was wrong with that wager? He seemed to know the category well and left himself a small chunk for FJ in case he got it wrong, which he didn't.

He could've bet enough to be in 1st place if right and still have been around for FJ if wrong.

If you're not trolling then you are really missing something.

As long as he bets big enough to get near the lead he might as well go all in or damn close. If he ends up with any small number his chance to win is 0.

If he gets to first his chance to win is =>50% (more correct answers than misses on FJ).

Much less than 50% from second.

Even the asian guys (not named Arthur Chu) on Jeopardy suck at math, lol at shoving from third

He can't win by standing pat, therefore shoving gives him the best chance to win.

I must have not have been paying attention well enough, thought the scores were such that he could have won with a zero wager.

Ugh, so if I did things correctly when

1. 1st place miss on a cover bet is less than 3rds current score
2. 2nd place miss on a bet to tie 1st is less than 3rds current score


3rd overbets (i.e. loses a triple stumper if 1st and 2nd wager as listed above) when 3rd could have bet 0 or small at 87%

Example 13K-10K-8K and 3rd bets >1K

3rd rarely goes for the triple stumper.

Won't 2nd almost always bet to cover third in such a scenario? 1st's minimum is 6k, hence 3rd can bet up to 2k (this also protects against a pussy 0 wager by 2nd).

I think all the bets were fine today, although arguably second should have wagered more if she assumes that third is shoving.

Probably not the best example as there are other options as you note.

Basically I calculated 1st protect bet and assume 1st missed.
Then I calculated 2nd's tie bet and assume 2nd missed.
The I calculated 3rds actual score based on their actual bet and missing.

I checked quite a few examples, I didnt see any where 3rd did the intermediate you suggest, but know I've seen it, typical play is for 3rd to shove. Too tired to go back and filter out those cases right now.

In any case, 3rd going for WWW is a better play than hoping for WWR.
WWW happens 19%
WWR happens 9%

2nd also overbets a signficant portion which further allows 3rd back to win on WWW when 2nd can bet small enough to on WWW but shoves instead.

Agree, tried to calc the general case without watching this episode, 1st had 3rd locked from a triple miss, so 3rd might as well shove.

But in general:
Amazing how timid everyone is a bets small when they shouldn't on DDs and then they go over-aggro on FJ when they should bet small.