Thursday, December 8, 2011 Game Recap & Discussion [SPOILERS]

[slam]

The risk (gambit) of such a small wager is, of course, that it leaves the leader vulnerable to losing on a double get and a big "irrational" bet from second place. Part of the problem with Tony's bet is that it wasn't big enough to take the lead if William had actually tried "Pawson's Gambit" (another argument for going all in if you're betting on yourself to win) So, in this case, I think Pawson's Gambit of "assuming the trailer will bet rationally, or only slightly irrationally" would have been the optimal wager, given that Tony did the latter.

I disagree with your characterization of Tony as betting as "only slighly irrationally". I think that 2nd place going all in (or close to it) is betting "only slightly irrationally". With both those wagers Tony would need to get FJ and have the leader miss it to win. At least with the bigger wager, his payoff is greater if he actually does win.

It seems that Tony did a slight variation of what we've seen many 2nd place contestants do, bet to exceed the leader's pre-FJ! score by $1. Tony's wager was a bit better than that since he avoided the leader wagering something like $2 to counter that common wager (though this strategy is most commonly seen in the QFs of tournaments when wildcards are available).

[marpocky]

Same way, but my problem is that it's the same word! For that matter, "orange" and "orange" are homophones (giving what I thought was an almost unrhymable word a rhyming pair). It's bad enough when songs do it, they shouldn't be called homophones if they're pronounced the same way and spelled the same way. In my book, that's the same word.

I also thought the words had to be spelled differently to be considered homophones, but we're wrong, at least according to dictionary.com:

1. Phonetics . a word pronounced the same as another but differing in meaning, whether spelled the same way or not, as heir and air.

Yep. From what I can tell, homophones only require identical pronunciation, and are irrespective of spelling.
Homonyms, on the other hand, require both identical pronunciation and spelling (and as such, are a subclass of homophones).

[Turd Ferguson]

The risk (gambit) of such a small wager is, of course, that it leaves the leader vulnerable to losing on a double get and a big "irrational" bet from second place. Part of the problem with Tony's bet is that it wasn't big enough to take the lead if William had actually tried "Pawson's Gambit" (another argument for going all in if you're betting on yourself to win) So, in this case, I think Pawson's Gambit of "assuming the trailer will bet rationally, or only slightly irrationally" would have been the optimal wager, given that Tony did the latter.

I disagree with your characterization of Tony as betting as "only slighly irrationally". I think that 2nd place going all in (or close to it) is betting "only slightly irrationally". With both those wagers Tony would need to get FJ and have the leader miss it to win. At least with the bigger wager, his payoff is greater if he actually does win.

It seems that Tony did a slight variation of what we've seen many 2nd place contestants do, bet to exceed the leader's pre-FJ! score by $1. Tony's wager was a bit better than that since he avoided the leader wagering something like $2 to counter that common wager (though this strategy is most commonly seen in the QFs of tournaments when wildcards are available).

What I meant is that a $3K or less bet would be "rational", and anything more than that would be "irrational", with "slightly" irrational being closer to $3K than all in. As I'd said earlier, "very irrational" is a better bet than "slightly irrational" in this case... I was just trying to say that trying Pawson's Gambit is the best bet to make for all rational and some irrational second place bets (like this one)

Betting to pass the leader's pre-FJ score is indeed common and is a defensible move, as we've seen the leader bet $0 before. In a 3/4 game, it also happens to be a rational bet... it's just the "quirk" of this being a 2/3-but-not-3/4 game that makes it a particularly bad wager.

[cinemaniax7]

Pennsylvania and Rhode Island were the only possible alternatives.

Yes, well, I took a left turn at Poughkeepsie, and ended up in Rhode Island.

[Mark B]

I was feeling pretty good about 27/1, 23/2 with all 3 dd's until missing FJ. Definitely not a wheelhouse cat. for me, natch.

I missed tau/tao too -- totally not pronounced the same way (wouldn't have rung in, but briefly thought chi/chi).

I got Beta, but there is a point here. Beta the tape is named after beta the letter: the tape looked like the letter. Is 'snake' the plumber's tool a homophone of 'snake' the creature?

Psi, Pi, Nu, Mu, Rho. There, was that so hard?

[John Boy]

No one getting Sri Lanka was the other surprise in the game.

Really? I thought that clue was pretty tough; I've never heard of the "Teardrop Island" moniker before (I think that's what it was). I got it, but only because of the mention of Trincomalee, which I had thought was fairly obscure.

You don't need to have heard of Trincomalee (I certainly never had) or even the "Teardrop Island" nickname. All you had to do was ask what large island (did they say "island nation?") is shaped like a teardrop. Sri Lanka was instaget....


...as was Maryland. I agree that both were rather disappointing TSs.

[John Boy]

I also was surprised that no player got FJ. I did think tony's FJ wager was perfectly reasonable absent not getting it right- which is always the ex-post facto criticism.

Uh, Tony's "perfectly reasonable" FJ wager cost him a game he should have won. (Other than that it was perfect.)

Don't you realistically have to assume that the leader will make the $1-lockout wager? Really? William has to bet at least $7801 to cover a doubling bet by Tony. Which he did, adding a few bucks probably for the simplicity of math. If he makes the standard wager and answers correctly, Tony loses, period.

Ergo (and for the umpteenth time on this show), Tony's only realistic hope is that William misses FJ. If William wagers correctly and misses, he drops to $9,599 (or in this actual case, to $9,400). With that in mind, Tony has to bet to cover $9,600, i.e. to wager no more than $3,000 (bearing in mind that in this game Dan is no factor). If Tony does that he climbs to $15,600 with a correct answer, or drops to $9,600 with a miss, and in either case wins the game.

Who knows what he was thinking? Perhaps he just chose a nice, round number with no strategic thinking attached to it. Whatever he did, it was a bet that lost the game that he would have won with a correct wager. Just sad.

[Volante]

I was unaware of the colonial Maryland = Catholic link before (picked Massachusetts due to its current Catholic-ness), that "Mary-land" mnemonic is a good way of remembering it, thanks! I just wish you told me that 10 minutes ago.

Same trap. I thought Catholic was meant to lead you to the Kennedys, so went with Mass. as well. I doubt I would have picked up on 'Mary-land' as a reference to Catholicism; until five minutes ago (read: Wikied it) I though it was a reference to Queen Mary (of W&M fame) which (Thanks, Wiki) is wrong on at least two levels: Maryland established in the early-mid 1600s while William and Mary were 1690s and the other level, the two were Protestants.

Finally, judges, for the clue that had "Oregon State" as a correct response, I just said "Oregon" (I often forget your states usually have two universities with the state name: University of X and X State) - I know the short form is considered short for one of those, but which one? (ie is "Oregon" short for "Oregon State" or "University of Oregon"?)

University is the word that you can omit, so "Oregon" is wrong in this case.

[seaborgium]

I brainstormed the thirteen colonies, and I think Maryland was around the fourth I came up with (never got around to all 13). I missed the cue in its name, but when I considered it, I got the vague sense that the oldest Catholic cathedral was in Baltimore. Without any inklings in other directions, I stuck with Maryland. Not only was I right on FJ; I was right about the cathedral (though it was built decades later)!

[Turd Ferguson]

I find it incredible that three J! class players miss this FJ. I thought this was college or even teen level, taught in every American History class.

Not in mine; I've only vaguely heard of the Calverts.

I just did a quick j-archive search, and it seems that Calvert comes up often. I did remember the DD from this game - http://www.j-archive.com/showgame.php?game_id=3564 (speaking of odd FJ! wagers... oh, Molly) where I too had no idea about the Calvert-Baltimore connection. I guess the Calvert-MD connection is something future contestants might want to be aware of.

(Speaking of that earlier game, the archive says -- IBM Challenge mini-documentary: "What is a Grand Challenge?" -- Does anyone know what that refers to?)

[Volante]

I find it incredible that three J! class players miss this FJ. I thought this was college or even teen level, taught in every American History class.

Not in mine; I've only vaguely heard of the Calverts.

I just did a quick j-archive search, and it seems that Calvert comes up often. I did remember the DD from this game - http://www.j-archive.com/showgame.php?game_id=3564 (speaking of odd FJ! wagers... oh, Molly) where I too had no idea about the Calvert-Baltimore connection. I guess the Calvert-MD connection is something future contestants might want to be aware of.

(Speaking of that earlier game, the archive says -- IBM Challenge mini-documentary: "What is a Grand Challenge?" -- Does anyone know what that refers to?)

I can do Calvert - Baltimore - Maryland just fine. I've never had 'Catholic' injected in there at any time.
Another tidbit: Annapolis is named after Calvert's wife. So, basically, Maryland, et. al., is not named after English Royalty. That should help with the changing synapses...

[slam]

I find it incredible that three J! class players miss this FJ. I thought this was college or even teen level, taught in every American History class.

Not in mine; I've only vaguely heard of the Calverts.

I just did a quick j-archive search, and it seems that Calvert comes up often. I did remember the DD from this game - http://www.j-archive.com/showgame.php?game_id=3564 (speaking of odd FJ! wagers... oh, Molly) where I too had no idea about the Calvert-Baltimore connection. I guess the Calvert-MD connection is something future contestants might want to be aware of.

(Speaking of that earlier game, the archive says -- IBM Challenge mini-documentary: "What is a Grand Challenge?" -- Does anyone know what that refers to?)

What am I missing about Molly's wager? She had $18,800 to 2nd place's $9,800 and wagered $801, the minimum cover wager.

[Turd Ferguson]

I just did a quick j-archive search, and it seems that Calvert comes up often. I did remember the DD from this game - http://www.j-archive.com/showgame.php?game_id=3564 (speaking of odd FJ! wagers... oh, Molly) where I too had no idea about the Calvert-Baltimore connection. I guess the Calvert-MD connection is something future contestants might want to be aware of.

(Speaking of that earlier game, the archive says -- IBM Challenge mini-documentary: "What is a Grand Challenge?" -- Does anyone know what that refers to?)

What am I missing about Molly's wager? She had $18,800 to 2nd place's $9,800 and wagered $801, the minimum cover wager.

Oh, sorry, it was actually in her next game where she bet everything from the lead and lost on a triple stumper - http://www.j-archive.com/showgame.php?game_id=3565

I also think I now know the answer to my question above... the "mini-doc" was presumably a Watson preview inserted into the episode (all the episodes immediately prior to the Watson games have similar tags)

[slam]

I just did a quick j-archive search, and it seems that Calvert comes up often. I did remember the DD from this game - http://www.j-archive.com/showgame.php?game_id=3564 (speaking of odd FJ! wagers... oh, Molly) where I too had no idea about the Calvert-Baltimore connection. I guess the Calvert-MD connection is something future contestants might want to be aware of.

(Speaking of that earlier game, the archive says -- IBM Challenge mini-documentary: "What is a Grand Challenge?" -- Does anyone know what that refers to?)

What am I missing about Molly's wager? She had $18,800 to 2nd place's $9,800 and wagered $801, the minimum cover wager.

Oh, sorry, it was actually in her next game where she bet everything from the lead and lost on a triple stumper - http://www.j-archive.com/showgame.php?game_id=3565

I also think I now know the answer to my question above... the "mini-doc" was presumably a Watson preview inserted into the episode (all the episodes immediately prior to the Watson games have similar tags)

That makes more sense! The wagering situation in that game was somewhat similar to this game. The player in 2nd place in that game also bet too much, but Molly decided to give it back by wagering too much herself. So, the player who "should" have won, actually did, but with less money than expected.

[seaborgium]

I think William's situation is more interesting, though, as he had multiple options. The wagering calculator for these scores suggests making either the cover wager ($7801) or the Shoretegic wager ($4401) from that lead. What I think is missing is that I believe that (for an opponent one thinks will bet "rationally"), the best wager in this case is to try "Pawson's Gambit" (betting $1799).

Because Tony was at 2/3, but not at 3/4 of the leader's score, a get on the "smart" bet (<$3K) doesn't get Tony to William's pre-FJ score (maxing out at $15600) Thus, if William assumes that Tony will bet "rationally", he can try Pawson's Gambit and just wager less than $1800 (to stay above $15600 on a miss), which gives him the win no matter what happens in FJ!

The risk (gambit) of such a small wager is, of course, that it leaves the leader vulnerable to losing on a double get and a big "irrational" bet from second place. Part of the problem with Tony's bet is that it wasn't big enough to take the lead if William had actually tried "Pawson's Gambit" (another argument for going all in if you're betting on yourself to win) So, in this case, I think Pawson's Gambit of "assuming the trailer will bet rationally, or only slightly irrationally" would have been the optimal wager, given that Tony did the latter.

I object to your use of the term "Pawson's Gambit" for this strategy. Dan Pawson led by less than $1,000 (Jason Thweatt being well beyond the two-thirds, three-fourths, and four-fifths thresholds) and bet his lead minus a dollar. Your idea counts on a two-thirds (and no better) second place wagering an amount that doesn't catch up; Dan's wager against Jason counted merely on Jason getting FJ wrong (in fact, Jason made a proper wager that would have won against the gambit, even on a double get).

[Spaceman Spiff]

You don't need to have heard of Trincomalee (I certainly never had) or even the "Teardrop Island" nickname. All you had to do was ask what large island (did they say "island nation?") is shaped like a teardrop. Sri Lanka was instaget....


...as was Maryland. I agree that both were rather disappointing TSs.

Glad I wasn't in this game. I would have embarassed myself with Ceylon, which was an instaget. Not sure if it's acceptable (likely would have doubleclutched and corrected it before Alex had responded, but still....)

And Maryland was a quasi-instaget; I misread the clue and thought it was the lone Catholic signer, and tried to remember who from Maryland was at the convention (was thinking of Charles Carroll of Carrollton, but wasn't sure if he was it, or if one of the other Marylanders was). I was much relieved after re-reading the clue that they only wanted the state.

[Turd Ferguson]

I think William's situation is more interesting, though, as he had multiple options. The wagering calculator for these scores suggests making either the cover wager ($7801) or the Shoretegic wager ($4401) from that lead. What I think is missing is that I believe that (for an opponent one thinks will bet "rationally"), the best wager in this case is to try "Pawson's Gambit" (betting $1799).

Because Tony was at 2/3, but not at 3/4 of the leader's score, a get on the "smart" bet (<$3K) doesn't get Tony to William's pre-FJ score (maxing out at $15600) Thus, if William assumes that Tony will bet "rationally", he can try Pawson's Gambit and just wager less than $1800 (to stay above $15600 on a miss), which gives him the win no matter what happens in FJ!

The risk (gambit) of such a small wager is, of course, that it leaves the leader vulnerable to losing on a double get and a big "irrational" bet from second place. Part of the problem with Tony's bet is that it wasn't big enough to take the lead if William had actually tried "Pawson's Gambit" (another argument for going all in if you're betting on yourself to win) So, in this case, I think Pawson's Gambit of "assuming the trailer will bet rationally, or only slightly irrationally" would have been the optimal wager, given that Tony did the latter.

I object to your use of the term "Pawson's Gambit" for this strategy. Dan Pawson led by less than $1,000 (Jason Thweatt being well beyond the two-thirds threshold) and bet his lead minus a dollar. Your idea counts on a two-thirds (and no better) second place wagering an amount that doesn't catch up; Dan's wager against Jason counted merely on Jason getting FJ wrong (in fact, Jason made a proper wager that would have won against the gambit, even on a double get).

It wasn't so much what Dan actually did in his games, but what was discussed in a thread from the old board. The 2/3-but-not-3/4 situation happened in a game around Christmas last year (perhaps this one - http://www.j-archive.com/showgame.php?game_id=3534), and I made what I thought was an "original" suggestion in the game thread that the leader could have assumed the trailer would have bet "rationally", bet small and taken the "automatic win". (That happened to be the situation in my second game... my "rational" bet from second place in a 2/3s game left open the possibility of Bob betting very small and ensuring a victory even if he missed and I got FJ).

After I made the suggestion, it was mentioned that while what I was suggesting wasn't Shore's Conjecture, it was in fact a "form" of Pawson's Gambit, though not one he ever actually employed (this was referred to in another old post by another poster as the "strong form" of Pawson's Gambit)... I remember a rather long thread evolving... you were there, and Dan was there, etc.

Dan referred to an old post of his containing the results of a study he had done of archived games where betting this small was a possibility, and had found that it would have been roughly a "break-even" technique (winning less money, on average, but about the same % of games), though he referred to specific TOC players he would've considered trying it on (based on how they had bet in their regular games... since it's the trailer's betting that determines whether or not this actually works), should the situation have come up. I really hope I'm not "misremembering" or making this up or I'd have the most boring "imaginary life" possible

[Bamaman]

I ran through the colonies as fast as I could Georgia was settled by Oglethorpe's debtors. The Carolinas didn't seem right and I knew it couldn't be Virginia. I made the same mistake as someone else and thought Maryland was from William and Mary and moved on.

Pennsylvania was Quaker, Rhode Island was Roger Williams. Running out of time, I scribbled down NY to have something .I figured it was a big, important state, must have been pretty diverse and had a Catholic delegate.

Sorry, never heard the tale of Maryland being founded by Catholics, even though I have heard of the Calvert family.

[OrangeSAM]

... and tried to remember who from Maryland was at the convention (was thinking of Charles Carroll of Carrollton, but wasn't sure if he was it, or if one of the other Marylanders was).

You are correct, sir. He was the one.

[heelsrule1988]

I'm surprised some of you didn't say Massachussets for FJ... Catholics and that state go hand-in-hand.