Tuesday, May 19, 2015
Andrew Haringer has won three games, all from behind. What will happen in his fourth?![]()
He’s gonna need to do it again.
|
Andrew Haringer |
Todd Grabarsky |
Liz Rapelye |
| 13,200 | 13,600 | 7,400 |
Tonight’s Final Jeopardy! category is:
| EUROPEAN CITIES |
Remember that yesterday, Andrew played it safe and won by a dollar.
First-order calculations
Second doubles up
If Andrew doubles his score, he’ll have 26,400.
To cover this all-in wager, Todd will need to wager 12,800.
| Andrew |
Todd |
Liz | |||
| 13,200 | 13,600 | 7,400 | |||
| 13,200 | 12,800 | ||||
| 26,400 | 26,400 | ||||
| min | max | min | max | min | max |
| 12,800 | |||||
An incorrect response with that wager will leave Todd with 800.
To stay above his total, Andrew can wager up to 12,400.
Liz can wager up to 6,600.
| Andrew |
Todd |
Liz | |||
| 13,200 | 13,600 | 7,400 | |||
| 12,400 | 12,800 | 6,600 | |||
| 800 | 800 | 800 | |||
| min | max | min | max | min | max |
| 12,400 | 12,800 | 6,600 | |||
Third doubles up
A successful doubling will put Liz at 14,800.
To cover this, Andrew should wager at least 1,600.
Todd would need to wager at least 1,200.
| Andrew |
Todd |
Liz | |||
| 13,200 | 13,600 | 7,400 | |||
| 1,600 | 1,200 | 7,400 | |||
| 14,800 | 14,800 | 14,800 | |||
| min | max | min | max | min | max |
| 1,600 | 12,400 | 12,800 | 6,600 | ||
| 1,200 | |||||
Liz would need to wager at least 5,000 to get above Todd if he makes that play.
| Andrew |
Todd |
Liz | |||
| 13,200 | 13,600 | 7,400 | |||
| 1,600 | 1,200 | 5,000 | |||
| 11,600 | 12,400 | 12,400 | |||
| min | max | min | max | min | max |
| 1,600 | 12,400 | 12,800 | 5,000 | 6,600 | |
| 1,200 | |||||
Here’s where things stand after the first order:
| Andrew |
Todd |
Liz | |||
| 13,200 | 13,600 | 7,400 | |||
| FIRST-ORDER | SECOND-ORDER | COVER ZERO | |||
| min | max | min | max | min | max |
| 1,600 | 12,400 | 12,800 | 5,000 | 6,600 | |
| 1,200 | |||||
Second-order calculations
Mind games don’t make a ton of sense here. Todd and Andrew are too close in score for Todd to go for less than the lockout wager; there’s also no downside to Andrew covering a double-up by Liz rather than, say, wager 800 to cover her maximum against Todd.
Zero wagers
Score differences:
Todd and Andrew: 400
Todd and Liz: 6,200
Andrew and Liz: 5,800
| Andrew |
Todd |
Liz | |||
| 13,200 | 13,600 | 7,400 | |||
| FIRST-ORDER | SECOND-ORDER | COVER ZERO | |||
| min | max | min | max | min | max |
| 1,600 | 5,800 | 12,800 | 6,200 | 6,600 | |
| 11,600 | |||||
| 12,400 | 1,200 | 6,200 | 5,000 | ||
Then we add a dollar to the minimum wagers, and subtract a dollar from the maximum wagers.
What actually happened
| Andrew | Todd | Liz | |||
| 13,200 | 13,600 | 7,400 | |||
| 13,200 | 12,801 | 1,600 | |||
| 26,400 | 799 | 5,800 | |||
| min | max | min | max | min | max |
| 1,601 | 5,799 | 12,801 | 6,201 | 6,599 | |
| 11,599 | 1,201 | 6,199 | |||
A few questions about Andrew:
Why? That’s such a completely reckless wager, when holding back even $800 would be far better.
How? He’s picked his spots perfectly: all-in both times he got Final right, and conservatively both times he got Final wrong.
What? … the hell is going on? (Note that this question is distinct from WHAAAAAT?)
Andrew has now won 4 games and $72,399 – all from second, the first time that’s happened to my knowledge. Using the same baseline assumptions that I did for Kerry Greene’s run, the odds of him pulling this off given the actual wagers are:
Game 1: 3/8 (WRx, WWW)
Game 2: 1/4 (WRx)
Game 3: 1/2 (Wxx)
Game 4: 1/4 (WRx)
Overall: 3/256 = 1.17%
The Final Jeopardy! clue
| EUROPEAN CITIES |
| THE EUROPEAN CENTRAL BANK IS BASED IN THIS CITY SOMETIMES CALLED “MAINHATTAN” |
[spoiler]What is Frankfurt?[/spoiler]
[php snippet=12]

Andrew seems to be one of those players who wager based on category rather than game theory, even though he can wager pretty well from second when he wants to (game 3).
Dont know if I’d go Red for Liz here, 2nd over wagers a big enough percent of the time to make it a viable wager (note I’m not arguing its good game theory, but ignoring observed betting patterns isnt wise either). Maybe a light pink 🙂
Since Liz can both get into range if Andrew wagers the minimum to cover her while at the same time remaining above Todd’s MSBIW that’s probably her best range (and since she can additionally cover both opponents’ zero wagers that narrows it further). If the two ranges were mutually exclusive then her choice would have more justification, but since she doesn’t have to choose her wager gets the “bad but not terrible” grade IMO.