I don't expect Jeopardy contestants to be game theory experts, but I expect them to at least consider where they and their opponents will land on incorrect responses. It doesn't take much for Lanny to see that a wager to cover Ray by $1 will drop her to $1,599 if she's wrong, and it shouldn't take much for her to also see that Liz's wager to cover her by $1 will drop her to the same amount. Then a little playing of the what-if game should solidify the tie offer as a valid strategy.omgwheelhouse wrote:So is there any chance a small wager (playing for the triple stumper) was the best play (for both 2nd and 3rd)? In my opinion, it's a pretty tough ask to expect lanny to bet exactly 5200. The contestants aren't game theory experts, after all.
Tuesday, October 18, 2011 Game Recap & Discussion [SPOILERS]
Moderators: alietr, trainman, econgator, dhkendall
-
- Undefeated in Reruns
- Posts: 8937
- Joined: Fri Jul 01, 2011 9:31 am
Re: Tuesday, October 18, 2011 Game Recap & Discussion (SPOIL
- TheConfessor
- Jeopardy! Champion
- Posts: 53
- Joined: Fri Jul 22, 2011 1:27 pm
Re: Tuesday, October 18, 2011 Game Recap & Discussion (SPOIL
I wonder what would have happened if someone answered "bicycling" instead of "speed skating" on that one question. Both Heiden and Blair were serious competitive cyclists, which is a common cross-training sport for speed skaters.
-
- Contributor
- Posts: 13
- Joined: Wed Oct 12, 2011 7:38 pm
Re: Tuesday, October 18, 2011 Game Recap & Discussion (SPOIL
If only you were still on the show, Joon, you could have showed 'em!jpahk wrote:ouch, ugly game, culminating in a real s***show of a FJ by all three players, who came up with a total of 0 foreign-born inventors among them (let alone the correct one). and to compound matters, all three miswagered.
the problem is that with evenly spaced scores, nobody can afford to add the extra dollar, so they both messed up. it's a complicated situation, but it basically boils down to this: liz needs to bet at least $6000 to cover lanny doubling up. lanny could bet it all to try for a tie on a double-get, but she has a much better shot to win by betting exactly $5200: that would cover ray if they both get it right, but not drop her below liz if both ladies miss. ray, therefore, should bet it all and try to tie lanny on the double-get. as for liz, she can't afford the extra dollar to beat lanny outright, because that would cause her to drop below lanny's optimum bet of $5200 on a double-miss. the wagering calculator agrees with me, without going into as much detail.dhkendall wrote:So, not being an expert, or even an adequate novice, at wagering theory, who forgot to add for the dollar to prevent the tie?
ray was pretty handily the best player in this game, and was even in the lead late in DJ after dropping to zero on his DD. i guess it's only fitting that he won the most money.
-
- Valued Contributor
- Posts: 71
- Joined: Mon Jul 11, 2011 10:33 am
Re: Tuesday, October 18, 2011 Game Recap & Discussion (SPOIL
Precalled Tesla and stuck with him on FJ without considering Bell. I didn't know he was Scottish but he should at least have been in consideration. oh well.
34,800 Coryat which is my highest in recent memory, though.
I knew the actor they were going for for Anita Van Buren, but my thought process was "I know! that's S. Epatha Merk... -elson? -inson? -ison? which is it?" and then time ran out.
34,800 Coryat which is my highest in recent memory, though.
I knew the actor they were going for for Anita Van Buren, but my thought process was "I know! that's S. Epatha Merk... -elson? -inson? -ison? which is it?" and then time ran out.
-
- Jeopardy! Champion
- Posts: 127
- Joined: Fri Jul 01, 2011 3:25 pm
- Location: Chicago, IL Area
Re: Tuesday, October 18, 2011 Game Recap & Discussion (SPOIL
After my stupid wagering error in my last game (which turned out to be irrelevant, but could have cost me a winnable game), it occurred to me that the players in a contested FJ! would be well-advised to calculate not only their own wager, but also the wagers that they would make if they were in the other two positions. Seeing the possible scores that would result from what you think are three sensible wagers could be very valuable in identifying mistakes in your own thinking, like, say, neglecting to notice that the third-place player was close enough to be a factor. In situations like the one that occurred last night, this seems especially important, since I'd bet that even seaborgium would want to double-check his thinking in such an oddball circumstance. The show gives you a pen, paper, and all of the time you want; why not use it to be 100% sure that you're doing things right?seaborgium wrote:I don't expect Jeopardy contestants to be game theory experts, but I expect them to at least consider where they and their opponents will land on incorrect responses. It doesn't take much for Lanny to see that a wager to cover Ray by $1 will drop her to $1,599 if she's wrong, and it shouldn't take much for her to also see that Liz's wager to cover her by $1 will drop her to the same amount. Then a little playing of the what-if game should solidify the tie offer as a valid strategy.omgwheelhouse wrote:So is there any chance a small wager (playing for the triple stumper) was the best play (for both 2nd and 3rd)? In my opinion, it's a pretty tough ask to expect lanny to bet exactly 5200. The contestants aren't game theory experts, after all.
- mam418
- Jeopardy! Champion
- Posts: 214
- Joined: Sun Jul 03, 2011 10:55 am
Re: Tuesday, October 18, 2011 Game Recap & Discussion (SPOIL
That's interesting that he got the call in January, to tape in August. Nick (NJCondon) and I (and the rest of our group) got the call in late January to tape in late February, for games airing in June. Odd, but maybe Ray had something come up and his taping was pushed to this season?seaborgium wrote:Now that I've watched the game, I would like to say that I had nothing to do with that FJ wager. To quote a message I sent Ray in January when he got The Call ...(SNIP)
- jpahk
- Jeopardy! TOCer
- Posts: 615
- Joined: Sun Aug 28, 2011 1:16 am
Re: Tuesday, October 18, 2011 Game Recap & Discussion (SPOIL
yeah, ray's bet was the least defensible of the three. i don't have a problem with going low and trying to win a triple-stumper from 3rd place, especially if you have good reason to believe that 1st and 2nd will both bet pretty big. if you're not going to do that, though, you have to bet it all; leaving the extra $2 just opens you up to lose by $2, because you're sure as hell not going to win a triple stumper with a $2 total.jeff6286 wrote:While I agree with Joon that either or both of the ladies should have taken off the extra dollar in an effort to avoid the tie on a miss, I think Ray had by far the worst wager of the three contestants. Each of the ladies at least had a logical reason to bet the amount that they did, but Ray's wager was completely pointless. He could have been the sole winner with $1,600 if he had realized what the predicted wagers were going to be for the ladies, or even if he couldn't get that far, he could have simply bet $0 or any other small amount, realizing that he would probably need both of his opponents to miss in order to win.
I disagree that Ray should bet it all, as I think he has a better chance of winning with a bet of $4,400 or less, as it seems to me that if you are assuming that both of your opponents will miss, then it is likely that you too will miss, so you are better off playing for the triple stumper if there is an amount you can bet that leaves you in the lead on said triple stumper. Sure, there was the possibility that Lanny would bet exactly $5,200, and he could tie her at $12,000 on a double get, but if I was in his shoes I would consider it pretty unlikely.
i wasn't on the old sony boards, but welcome back. and thanks!goforthetie wrote:I'm sad to have missed out on a bunch of stuff, including the SHC and contributing yelps of joy on two math TDDs in two weeks. Well done, Joon!
sorry, i didn't mean to sound like an arrogant jerk. (but sometimes i can't help it...) i don't know what would've happened if i'd been on the show. i got FJ wrong, i'll admit that, but at least tesla's a foreign-born inventor. the DDs were brutal—i don't think i would have gone big on either 2008 or world geography, but i probably would have in fiction (luckily, that's the only one of the three i knew).nlafavore wrote:If only you were still on the show, Joon, you could have showed 'em!
-
- Contributor
- Posts: 10
- Joined: Mon Sep 19, 2011 4:20 pm
Re: Tuesday, October 18, 2011 Game Recap & Discussion (SPOIL
This is a really really interesting FJ wagering scenario to me. Liz pretty much has two options: $6001 to cover a double-up by Lanny, or $6000 to offer a tie on Lanny's double-up. A correct response will give Liz $13601 or $13600, and an incorrect response will leave her with $1599 or $1600. She'll win every time she gets FJ right, and she'll probably lose every time she gets it wrong.
Ray has the next easiest decision to make. If Lanny bets to cover his double-up, she'll wager $5201 and end up with $1599 on an incorrect response. Or, she might wager $5200, offering him a tie on a double-get, and beating out or tying Liz on a TS. OR, she might wager it all. Ray can't be certain that Lanny will offer him a tie, but he can be much more certain that he can win on a TS. I think a wager of $4399 is best, as it maximizes his winning while winning on a TS, given probable wagers by Liz and Lanny. I don't have solid numbers on this, but I think a TS among evenly matched opponents is more likely than a double get, so he can't wager everything in hopes that Lanny offers him the tie.
Lanny is in a tough spot. Wagering everything will only result in a win if she gets it right and Liz either negs or offers the tie. Wagering $5201 to cover Ray's double-up on a double get will leave her with $1599, tying her with Liz if they're both incorrect. But Ray might bet to win on a TS, so she shouldn't bet $5200 to offer him the tie on a double-get... that dollar margin she'd have over Liz on a TS might not matter. If she feels that Ray (who was probably the strongest player in the game) will wager big, she's left with either offering him a tie on the double-get, or offering Liz the tie on a TS... but will Liz bet to cover, or will she offer the tie on a double get?
I think it's reasonable to assume that all three players won't end up at the optimal betting strategy suggested by the j-archive. The player in the lead usually wagers to cover a double-up by the player in 2nd. In this case, it makes the most sense for Lanny to wager $5201 to cover Ray and tie Liz in the event that there's TS and Ray bets big. Ray should bet small enough to cover all likely wagers on a TS, something like $4399 or less, giving him the win on a single get and a TS.
Ray has the next easiest decision to make. If Lanny bets to cover his double-up, she'll wager $5201 and end up with $1599 on an incorrect response. Or, she might wager $5200, offering him a tie on a double-get, and beating out or tying Liz on a TS. OR, she might wager it all. Ray can't be certain that Lanny will offer him a tie, but he can be much more certain that he can win on a TS. I think a wager of $4399 is best, as it maximizes his winning while winning on a TS, given probable wagers by Liz and Lanny. I don't have solid numbers on this, but I think a TS among evenly matched opponents is more likely than a double get, so he can't wager everything in hopes that Lanny offers him the tie.
Lanny is in a tough spot. Wagering everything will only result in a win if she gets it right and Liz either negs or offers the tie. Wagering $5201 to cover Ray's double-up on a double get will leave her with $1599, tying her with Liz if they're both incorrect. But Ray might bet to win on a TS, so she shouldn't bet $5200 to offer him the tie on a double-get... that dollar margin she'd have over Liz on a TS might not matter. If she feels that Ray (who was probably the strongest player in the game) will wager big, she's left with either offering him a tie on the double-get, or offering Liz the tie on a TS... but will Liz bet to cover, or will she offer the tie on a double get?
I think it's reasonable to assume that all three players won't end up at the optimal betting strategy suggested by the j-archive. The player in the lead usually wagers to cover a double-up by the player in 2nd. In this case, it makes the most sense for Lanny to wager $5201 to cover Ray and tie Liz in the event that there's TS and Ray bets big. Ray should bet small enough to cover all likely wagers on a TS, something like $4399 or less, giving him the win on a single get and a TS.
- earendel
- Watches Jeopardy! Way Too Much
- Posts: 767
- Joined: Fri Jul 01, 2011 3:22 pm
- Location: mired in the bureaucracy
Re: Tuesday, October 18, 2011 Game Recap & Discussion (SPOIL
I was surprised to see the tie, because I didn't think the wagers would go like that.
For FJ! I came up with Bell but it was after the think music ended; prior to that I came up with Tesla but the time didn't sound right.
For FJ! I came up with Bell but it was after the think music ended; prior to that I came up with Tesla but the time didn't sound right.
"Elen sila lumenn omentielvo...A star shines on the hour of our meeting."
-
- Jeopardy! Contestant
- Posts: 198
- Joined: Fri Jul 01, 2011 11:50 am
Re: Tuesday, October 18, 2011 Game Recap & Discussion (SPOIL
I came up with Bell, but the clue was kinda crappy. As for the playing...no comment.
- Volante
- Harbinger of the Doomed Lemur
- Posts: 9254
- Joined: Thu Jul 14, 2011 11:42 pm
Re: Tuesday, October 18, 2011 Game Recap & Discussion (SPOIL
I called Bell as soon as I saw the category. Briefly thought of Edison, but was fully confident he was USA born and bred.
I'm also absolutely, positively 100% sure that this piece I read last month played no role what so ever in the confidence I had in answering Bell:
http://arstechnica.com/tech-policy/news ... entury.ars
I'm also absolutely, positively 100% sure that this piece I read last month played no role what so ever in the confidence I had in answering Bell:
http://arstechnica.com/tech-policy/news ... entury.ars
The best thing that Neil Armstrong ever did, was to let us all imagine we were him.
Latest movies (1-10): Everything Everywhere All at Once (10), Ruby Gillman: Teenage Kraken (6), Black Sunday /1960/ (6), Marcel the Shell with Shoes On (7)
Latest movies (1-10): Everything Everywhere All at Once (10), Ruby Gillman: Teenage Kraken (6), Black Sunday /1960/ (6), Marcel the Shell with Shoes On (7)
-
- Contributor
- Posts: 34
- Joined: Sat Jul 02, 2011 9:22 am
Re: Tuesday, October 18, 2011 Game Recap & Discussion (SPOIL
As to why the wagering calculator suggests that 1st and 2nd wager for the tie, I think it stems from the fact that cover bets if they both miss will drop them to the same point. Since your opponent may choose to bet for the tie instead, you should do the same so you won't lose by a dollar in that case. This "evenly spaced scores" scenario is essentially the break point between 2nd being and not being in Stratton's Dilemma (when A-B is more than B-C in a 2/3+ game SD applies). As for how 3rd should wager, as someone suggested in a case like this game instead of counting on 2nd's bet to tie going for a TS win might be the better route (on the other hand when 3rd doesn't have enough for that the best move is to go all-in with the possibility of a tie).
- MarkBarrett
- Watches Jeopardy! Way Too Much
- Posts: 16467
- Joined: Fri Jul 01, 2011 10:37 am
- Location: San Francisco
Re: Tuesday, October 18, 2011 Game Recap & Discussion (SPOIL
I had a similar reaction playing along at home.jpahk wrote: the DDs were brutal—i don't think i would have gone big on either 2008 or world geography, but i probably would have in fiction (luckily, that's the only one of the three i knew).
My wagers were:
1400 out of 11400 on Remember 2008? (Even though I had the four previous in the category I didn't trust myself to get the run.)
2400 out of 12400 in World Geography (I had the three leading up to the DD and got the 2000 clue after, but I still consider myself only average in that category)
6400 out of 16400 in Fiction (I had the four leading up to it and felt like I had a decent chance to get the run)
- MarkBarrett
- Watches Jeopardy! Way Too Much
- Posts: 16467
- Joined: Fri Jul 01, 2011 10:37 am
- Location: San Francisco
Re: Tuesday, October 18, 2011 Game Recap & Discussion (SPOIL
I could not find the tie game in my tape archive, but I did find the game after. Jimmie Bucci was at the middle slot as a champ with $100. Stephen Lebowitz was a 4x champ with $50,900.seaborgium wrote:
By the way, according to J! Archive, ToCer Stephen Lebowitz won his fourth game in a tie at $100.
The third player was Steve Herron and went into FJ with a lock:
Steve 9000 + 1000 = 10000
Jimmie 1900-200=1700
Stephen 1000+898=1898
While cuing up the game I also found Michael Rankins debut.
Champ Jack 5400 + 5400 = 10800
Michael 5700 + 5699 = 11399
Tom 7300 - 4101 = 3199
Yes, luck can play a huge part in someone's success on the show.
- whoisalexjacob
- 2015 TOC'er
- Posts: 563
- Joined: Fri Jul 15, 2011 12:19 am
Re: Tuesday, October 18, 2011 Game Recap & Discussion (SPOIL
So you think the joon/calculator strategies are the best plays, right? What percent of Jeopardy contestants make the 5200 wager in that situation? My assumption is that it's pretty low, so I don't know if I'd plan my strategy around my opponent betting exactly 5200. Not trying to say the consensus is wrong; I haven't done a lot of work on this, so I'm not speaking authoritatively at all.seaborgium wrote:I don't expect Jeopardy contestants to be game theory experts, but I expect them to at least consider where they and their opponents will land on incorrect responses. It doesn't take much for Lanny to see that a wager to cover Ray by $1 will drop her to $1,599 if she's wrong, and it shouldn't take much for her to also see that Liz's wager to cover her by $1 will drop her to the same amount. Then a little playing of the what-if game should solidify the tie offer as a valid strategy.omgwheelhouse wrote:So is there any chance a small wager (playing for the triple stumper) was the best play (for both 2nd and 3rd)? In my opinion, it's a pretty tough ask to expect lanny to bet exactly 5200. The contestants aren't game theory experts, after all.
-
- Undefeated in Reruns
- Posts: 8937
- Joined: Fri Jul 01, 2011 9:31 am
Re: Tuesday, October 18, 2011 Game Recap & Discussion (SPOIL
Ray's a Southern Californian. In January he got the call to be an alternate, and 16 hours before this taping he got called to replace someone Irene-bound in New York. (He tells me that the suddenness of that caused all the preparation advice I gave him to go out the window.)mam418 wrote:That's interesting that he got the call in January, to tape in August. Nick (NJCondon) and I (and the rest of our group) got the call in late January to tape in late February, for games airing in June. Odd, but maybe Ray had something come up and his taping was pushed to this season?seaborgium wrote:Now that I've watched the game, I would like to say that I had nothing to do with that FJ wager. To quote a message I sent Ray in January when he got The Call ...(SNIP)
How does one dollar's difference in Lanny's wager change your strategy? If I'm Liz and I think Lanny will bet $5,201, I bet $6,000 to beat her by $1 on a double stumper; if I think she'll bet $5,200, I bet $6,000 to tie her on a double stumper. If I'm Ray, I'm leaving at least $1,601 on the table no matter what I think Liz and Lanny will do. It's not planning a strategy around a $5,200 wager by Lanny; it's planning a strategy that is merely compatible with it.omgwheelhouse wrote: So you think the joon/calculator strategies are the best plays, right? What percent of Jeopardy contestants make the 5200 wager in that situation? My assumption is that it's pretty low, so I don't know if I'd plan my strategy around my opponent betting exactly 5200. Not trying to say the consensus is wrong; I haven't done a lot of work on this, so I'm not speaking authoritatively at all.
- hbomb1947
- Still hoping to get on Jeopardy! while my age is in double digits
- Posts: 2434
- Joined: Fri Jul 01, 2011 8:31 am
Re: Tuesday, October 18, 2011 Game Recap & Discussion (SPOIL
All those episodes with no female champions, and now we get two in one day.
Like some others here, I precalled Marconi off the category and then wavered between him and Bell before deciding that in 1922, the telephone probably would have been considered more valuable than radio (also, I wasn't sure that Bell was young enough in 1876 to have kicked the bucket in 1922, but given what I knew about Marconi I felt that the wireless dude had hung around a little longer than that).
I found the boards tougher than average, and yes, the DD's were brutal. I went 0/3 on those, the first time all season that I got fewer than 2 DD's correct. Antwerp for $1,600 seemed particularly brutal.
I precalled Liz's neg of "United States Geological Service."
In the earthquake category, I was disappointed that there weren't any really challenging clues (like maybe asking about the 1960 Valdivia temblor, or the Alaskan "Good Friday" quake of 1964).
And what's with all these "ripped from the headlines clues" lately? Yesterday they asked about Irene, and today a reference to the D.C. quake that also occurred just days before this week's episodes were taped.
Like some others here, I precalled Marconi off the category and then wavered between him and Bell before deciding that in 1922, the telephone probably would have been considered more valuable than radio (also, I wasn't sure that Bell was young enough in 1876 to have kicked the bucket in 1922, but given what I knew about Marconi I felt that the wireless dude had hung around a little longer than that).
I found the boards tougher than average, and yes, the DD's were brutal. I went 0/3 on those, the first time all season that I got fewer than 2 DD's correct. Antwerp for $1,600 seemed particularly brutal.
I precalled Liz's neg of "United States Geological Service."
In the earthquake category, I was disappointed that there weren't any really challenging clues (like maybe asking about the 1960 Valdivia temblor, or the Alaskan "Good Friday" quake of 1964).
And what's with all these "ripped from the headlines clues" lately? Yesterday they asked about Irene, and today a reference to the D.C. quake that also occurred just days before this week's episodes were taped.
Follow me on twitter, even though I rarely tweet! https://twitter.com/hbomb_worldwide
- whoisalexjacob
- 2015 TOC'er
- Posts: 563
- Joined: Fri Jul 15, 2011 12:19 am
Re: Tuesday, October 18, 2011 Game Recap & Discussion (SPOIL
seaborgium wrote:Ray's a Southern Californian. In January he got the call to be an alternate, and 16 hours before this taping he got called to replace someone Irene-bound in New York. (He tells me that the suddenness of that caused all the preparation advice I gave him to go out the window.)mam418 wrote:That's interesting that he got the call in January, to tape in August. Nick (NJCondon) and I (and the rest of our group) got the call in late January to tape in late February, for games airing in June. Odd, but maybe Ray had something come up and his taping was pushed to this season?seaborgium wrote:Now that I've watched the game, I would like to say that I had nothing to do with that FJ wager. To quote a message I sent Ray in January when he got The Call ...(SNIP)
How does one dollar's difference in Lanny's wager change your strategy? If I'm Liz and I think Lanny will bet $5,201, I bet $6,000 to beat her by $1 on a double stumper; if I think she'll bet $5,200, I bet $6,000 to tie her on a double stumper. If I'm Ray, I'm leaving at least $1,601 on the table no matter what I think Liz and Lanny will do. It's not planning a strategy around a $5,200 wager by Lanny; it's planning a strategy that is merely compatible with it.omgwheelhouse wrote: So you think the joon/calculator strategies are the best plays, right? What percent of Jeopardy contestants make the 5200 wager in that situation? My assumption is that it's pretty low, so I don't know if I'd plan my strategy around my opponent betting exactly 5200. Not trying to say the consensus is wrong; I haven't done a lot of work on this, so I'm not speaking authoritatively at all.
OK, well the Joon/calculator strategies say for Ray to bet it all, so I guess you may not agree with those. This is partly what I was referring to, as it was suggested that Ray should bet it all to try to tie Lanny (on the assumption she makes a 5200 wager).
-
- Undefeated in Reruns
- Posts: 8937
- Joined: Fri Jul 01, 2011 9:31 am
Re: Tuesday, October 18, 2011 Game Recap & Discussion (SPOIL
I generally don't believe in "bet everything because the person ahead might offer a tie," even if there would be some strategic logic behind the tie-offering wager. Sorry for the miscommunication. However, I do believe in wagering everything if you have to get it right for a realistic chance at victory, and I believe in wagering everything if you're going to overwager. I'll have none of that "bet it all minus a token" BS.omgwheelhouse wrote:OK, well the Joon/calculator strategies say for Ray to bet it all, so I guess you may not agree with those. This is partly what I was referring to, as it was suggested that Ray should bet it all to try to tie Lanny (on the assumption she makes a 5200 wager).
P.S. Ray says it was a "third time's the charm" attitude after his failed DDs that led to his FJ wager. Reminds me of how James Erwin lost his third game: he won his first two on triple stumpers from scores that were favorable to him in TSs, and then told himself, "I've never missed three in a row!" and bet huge. He lost from a close second on a TS.
-
- Also Receiving Votes
- Posts: 12895
- Joined: Sat Jul 02, 2011 1:39 pm
Re: Tuesday, October 18, 2011 Game Recap & Discussion (SPOIL
If Lanni and Liz had finished tied for second, would Lanni have gotten just $1,000, since she trailed Liz going into FJ? That might have been another reason to bet $5,200.
The wagering calculator often seems strange. It tells Ray to bet it all (I assume because Lanni might offer the tie to avoid a tie on a TS), but it tells Lanni to do the shutout bet.
I started with Edison on FJ, then halfway through remembered the catgory and quickly changed to Bell.
The wagering calculator often seems strange. It tells Ray to bet it all (I assume because Lanni might offer the tie to avoid a tie on a TS), but it tells Lanni to do the shutout bet.
I started with Edison on FJ, then halfway through remembered the catgory and quickly changed to Bell.