Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Tools for decision making (Brian Haff) ProductCamp Boston 2014


Published on

We've all made rational decisions and forecasts based on individually analyzing the best available data. But there are many other aspects of decision making. This session will examine some of those. When can groups of non-expert individuals beat some of the best experts? What are some of the common biases that cause ordinary people to make decisions differently from those that they "should" make. Can you take advantage of the ways other makes decisions or is this unwarranted manipulation?

Published in: Business, Technology, Sports
  • Be the first to comment

Tools for decision making (Brian Haff) ProductCamp Boston 2014

  1. 1. 1 Tools for Decision Making Gordon Haff Cloud Product Strategy Red Hat 3 May 2014
  2. 2. 2 About Me •  Red Hat Cloud Product Strategy •  Twitter: @ghaff •  Google+: Gordon Haff •  Email: •  Blog: •  Flickr: •  Formerly: Illuminata (industry analyst), Data General (minicomputers/Unix/NUMA/etc.), shareware developer
  3. 3. 3 Some Decision-Making Tools and Approaches •  Data vs. models •  Complex models vs. heuristics •  Data-driven vs. hunches •  Individual vs. group •  “Crowd wisdom” ç •  Behavioral decision theory ç
  4. 4. 4 The Wisdom of Crowds
  5. 5. 5 “The average competitor was probably as well fitted for a just estimate of the dressed weight of an ox, as an average voter is of judging the merits of most political issues on which he votes.” Francis Galton
  6. 6. 6 The Crowd’s Result? Within one pound
  7. 7. 7 An Oscar Contest Example: Consensus* vs. Experts *read “consensus” as modal group choice for each decision Best picture Best actress Phil Eve Sue Tom Harry ✔ Eve Tom Harry Phil Sue ✔
  8. 8. 8 Oscar contest: Consensus vs. Winners by Year
  9. 9. 9 Is Averaging Experts Better? Best > 20 year contestant: 28.5 (vs. 30.5) Best 5 year contestant: (31 (vs. 33 during same period)
  10. 10. 10 When do wise crowds work? •  Diversity of opinion •  Information •  Independence •  Decentralization •  Aggregation
  11. 11. 11 Decreasing Advantage = Decreasing Independence?
  12. 12. 12 Nate Silver: Aggregating Models
  13. 13. 13 Prediction Markets are an Implementation •  Regulatory hurdles when real money involved •  Have been experimented with for internal use by companies such as Google •  Ultimately, observing that aggregating many models/ decisions may trump individual “expertise” may be most important lesson
  14. 14. 14 Why are Individual Decisions Flawed? •  You’re stupid •  You lack data •  You lack domain expertise •  You lack time •  You’re being influenced by persistent and common biases and therefore making non-optimal decisions
  15. 15. 15 Behavioral Decision Theory •  Humans vs. homo economicus •  1979: Kahneman and Tversky: Prospect theory: An Analysis of Decision Under Risk •  Choice architectures •  Anchors •  Frames •  Sunk costs •  Probabilities •  Availability bias
  16. 16. 16 Organ Donation Rates in Europe What’s different between the gold and blue countries?
  17. 17. 17 Choice Architectures are Unavoidable •  At best, people depend on simplifying rules •  Can nudge in positive directions while maintaining freedom to decide differently •  But is a very powerful tool o  Default savings o  Automatic renewals o  Default options in software, etc.
  18. 18. 18 The Offered Choices Matter •  Caveat: Choice need not create complexity
  19. 19. 19
  20. 20. 20 Source: Wikimedia
  21. 21. 21 Anchoring: We Don’t Adjust Enough •  Bias can appear even when you know the anchor is random •  Leads to incrementalist approach (e.g. in policy-making) •  Common in negotiations •  Also related to setting confidence limits on forecasts and other predictions too narrowly
  22. 22. 22 •  Choose between: o  A sure gain of $240 o  25% chance to win $1,000 and a 75% chance to gain nothing
  23. 23. 23 •  Choose between: o  A sure loss of $750 o  75% chance to lose $1000 and a 25% chance to lose nothing
  24. 24. 24 Framing •  Risk aversion •  Risk seeking in losses •  Discounts vs. surcharges •  More broadly: o  25% meat vs. 25% fat o  “Pro-choice” o  Survey language
  25. 25. 25 Sunk costs The basketball game is still on in spite of the blizzard. Do you still go? •  You won the ticket from the local radio station •  You bought the ticket for $10 •  You bought the ticket for $100
  26. 26. 26 Sunk Costs •  Sunk cost avoidance is pervasively seen in experiments •  Possibility that spending more money may make the whole thing worthwhile consistent with prospect theory •  Also cognitive dissonance •  With respect to using sunk costs, foot-in-the-door, bait- and-switch, low-ball, skin-in-the-game are related •  Can make it sensible to focus on variable costs (temporarily) •  Always a fallacy?
  27. 27. 27 People Suck at Statistics (aka Gambler’s Fallcy) We haven’t had much snow for the past three years. The big one is due! Buy any model of Acme snowblower by September 30 and take 20 percent off. Be ready for the coming blizzard!
  28. 28. 28 Availability Bias •  Which is more common cause of death (US)? o  Fires? o  Parkinson’s Disease? •  Which is more common cause of death (US)? o  Road accidents? o  Flu and pneumonia? •  Overreact to new or familiar information
  29. 29. 29 Summary •  Groups can be powerful but composition and organization matter o  Wisdom of crowds vs. group-think o  Governance and culture also matter for joint and community undertakings (e.g. open source) •  Individual decision making subject to many biases o  Be aware!
  30. 30. 30 QUESTIONS & LEARN MORE MY INFO Twitter: @ghaff Google+: Gordon Haff Email: Blog: