Gambling on Politics

Agenda

  • WiseQ
  • PredictIt
  • Behavioral Economics

WiseQ

  • Deep Dive into Math
  • Math Part 2
  • Math Part 3
  • WiseQ
  • Results

Math

Just kidding

Priors 

Liquidity

The Pitch:

  1. A market intrinsically provides insight

 

 

 

2.   Combinatorial 

  • ​Mitt Romney would win Ohio yet lose election.
  • Same party would win both Ohio and Pennsylvania

WiseQ

The Design:

  • ​Standard Interface
  • Fake money 
  • 400 users: 3000 trades
  • 514 distinct types bought
  • (261 by single user...)

Results

  • "Three quarters of users never sold any securities". 10% had -100% return.
  • Top performer had 1.8k profit and ~100 buys+sells
  • 11:1 buy:sell ratio, most happening immediately before election night
  • "The game was featured on Yahoo News, the Huff Post, RealClearPolitics, and NYTimes"

 

 

 

  • "The users who performed best...non-presidential races....our priors (initial prices) were least accurate for these races".
  • Evaluation based on cherry-picked election results and counterfactual analysis

PredictIt

  • Totally not gambling
  • Market
  • Inspirational stories

Totally Legal

PredictIt

The Pitch:

  • "academic research"
  • New Zealand "non-profit".  For-profit takes 10% commission.  
  • "no action letter" -> each question is limited to 5,000 traders, and there is an $850 cap on individual investments per question

The Design:

  • In 2017, PredictIt users traded more than 300 million shares
  • Thousands of comments per bet

Investigative Journalism

No Results, Just Inspirational Personal Stories

  • Variations of "I am smart so I am investing not gambling" -> proceed to win until they lose all
  • Banker wants to work from home

Text

Text

Is this any different?

Behavioral economics

  • Experiment 1
  • Experiment 2
  • Experiment 3

Which of the following would you prefer:

  1. A) A certain win of $250, versus
    B) A 25% chance to win $1000 and a 75% chance to win nothing?
  2. How about:
    C) A certain loss of $750, versus
    D) A 75% chance to lose $1000 and a 25% chance to lose nothing?

Experiment 1

Risk-averse

Certain gain is great.  Let's risk the loss.

Experiment 2

Leave a bowl of nuts/chips/guacamole/etc in front of a bunch of people

  • How much will you spend per year in retirement?
  • How much do you plan to make per year for the 30 years of your career?  
  • How long are you planning to live?

Experiment 3

Nudge

  • 401k
  • Vaccination
  • Midway Project Update
Made with Slides.com