Note: This post represents the synthesis of the thoughts, procedures and experiences of others as represented in the articles read in advance (see previous posts) and the discussion among the students and instructor during the Advanced Analytic Techniques class at Mercyhurst University in October 2017 regarding Prediction Markets as an Analytic Technique Method specifically. This technique was evaluated based on its overall validity, simplicity, flexibility and its ability to effectively use unstructured data.
Description:
Prediction markets aggregate collective knowledge of the crowd through the use of a trading based system, allowing participants to invest a stake in a desired outcome they believe will come true. Prediction Markets, when conducted correctly, can provide a range of reasonably accurate forecasts and provide strong reliability to the answered question.
Strengths:
- Can be more efficient than some bureaucratic processes
- Able to aggregate disparate pieces of information to accurately predict resolvable questions
- The incentive to gain profits largely eliminates the occurrence of groupthink
- Works well even when people have limited knowledge about their surrounding environment and the people with whom they transact
- Can incorporate insight from experts across many different fields
- Experts not required
Weaknesses:
- No system for assessing the difficulty of the question
- Doesn’t work well with open ended questions
- Prone to question interpretation gap failures - difficult to determine if question answered actually satisfies intelligence requirement
- Can be manipulated if a speculative trade influences the beliefs of other traders, whether by playing by the rules or creating ideas that cheat the market
- Some moderate level of expertise is required to be a forecaster
- Long term estimates are at risk of forecaster apathy
- Requires a large volume of analysts to create the number of estimates needed
How-To:
- Choose a resolvable question
- Gather participants that have a general knowledge of the issue
- Stipulate incentives
- Have participants wager on the likely answer, candidate, or outcome
- Calculate market valuations
- Evaluate results
- Inform participants
Application of Technique:
The class received handouts with a list of NFL Football Teams and their standings in 2002 along with $100. They were asked to bet money on which teams they thought would win the SuperBowl in 2003 based on their final standings. After looking at the statistics, students would place bets on the top 3 teams they thought would win. After the bets were placed, we looked at the teams with the most bets and calculated the results. Without a binary choice, the class was capable of making the accurate prediction that the New England Patriots would win the 2003 Super Bowl.
For Further Information:
Posts
No comments:
Post a Comment