Hacker News new | comments | show | ask | jobs | submit login
The Resulting Fallacy Ruins Decisions: An Interview with Annie Duke (nautil.us)
57 points by acconrad 3 months ago | hide | past | web | favorite | 17 comments

Looks basically like black swan theory. Looking for explanations as if someone could have seen this coming.

No, it's simple misunderstanding of randomness. Black swans are unpredicatable unmodelable events.

"Resulting" just means misjudging well-modeledrandom events. It's thinking that if you win a small prize on a lottery ticket, that buying the ticket was a good choice _because you won_. While in reality, it was more likely that you'd lose.

"Think about the 2015 Super Bowl. The Seahawks are on the 1-yard line, they’re down by four, there’s 26 seconds left in the game, Pete Carroll has Russell Wilson throw and it’s intercepted. Do you remember what the headlines looked like the next day? “Worst play in Super Bowl history,” “What was he thinking?” “Idiot.” That kind of thing. But imagine it was caught—what do you think the headlines would have looked like then? The outcome was irrelevant to the decision quality."

no that was a pretty bad play choice.

As I recall the time on the clock allowed for only 1 running play or 1 quick pass play plus a running play. They made the best choice given the information they had.

Would you be saying that if it had been caught?


You've been posting mostly uncivil and/or unsubstantive comments. Could you please read the guidelines and start following them?


I’m skeptical.

The QB's completion percentage is ~57%, meanwhile seatle had one of the best running backs in the league with 4.25 yards per carry in the game.

wilson doesnt throw many interceptions, but Lynch almost never fumbles, Wilson was 4 times more likely to turn over the ball than Lynch was. write that down, if they turn the ball over they loose the game. the play they choose increased their risk of loosing over 400%.

wilson scores 28% of the time from the 10 (dont have 1 or 5 yard stats), lynch scores 42% of the time from the 5. he even scores more than wilson from 10 yards out (37%), which is not even a valid comparison because they only need 1 yard.

There is just no way to cut it, it was a bad decision.

Exactly. But it's hard to make a living criticising bad play calls that go well. That's why I think there are so many bandwagon analysts.

Same with Atlanta passing while in field goal range last Super Bowl and getting a holding penalty and a sack that forced them to punt and gave the Patriots a chance to tie.

How do the stats change when the opposing team knows what you know about these stats? That's a very relevant dampening factor.

From memory, the explanation for the play was that everyone expected them to go with Lynch, so they tried the pass for the surprise factor, and still had the fourth down to go with Lynch if the pass failed. So yes, very much affected by knowing that the opponent knows your stats.

You've done a good job convincing me that it was a bad play. But I'm still not convinced that people would be calling it a bad play if it had worked.

I couldn't read past the first, fawning paragraph about a figure in the poker world who likely stole millions of dollars from online players[1] directly, and at a minimum profited from turning a blind eye to her associates' theft[2].

[1] http://pokernewsboy.com/poker-player-news/secret-ub-tapes-an...

[2] https://en.wikipedia.org/wiki/Cereus_Poker_Network

Annie is well known as an awful person who hasn't been a top tier player for over a decade. The best players in today's game would love to have her in their games.

But she has taken in over 4 million in lifetime earnings... from FTP players.

supporting evidence for being an awful person >> in the interview she gives partial credit for her poker success to her brother while revealing that he lost his college fund to it...

Applications are open for YC Summer 2018

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact