:
- When thinking of making a "luck" decision (not based on skill) we should base our decision on the base rate and not a single case
- Untangling skill and luck helps make better predictions
- People base predictions on how well they fit a story. Fail to see how reliable story is or how that story actually played out
- Luck is what is left when you remove all sources of skill
- 3 steps of skill building
- Cognitive
- Associative
- Autonomous
- Understanding difference between experience and expertise is crucial
- Dubois equation - Be very aware of sample size. Some patterns of luck/skill will not become apparent until you have an appropriate sample size. May win one hand against a pro poker player, but as you play more hands his advantage will appear
- More luck in activity leads to a quicker reversion to the mean. If above average results and it's a lucky activity, quickly revert to average results
- Possible to over-test and the findings or relationship between two things is truly just luck. So many inputs that one of them is bound to show show some relation to the other
- The closer people's skill at a certain task is, the more luck effects the outcome
- Less skilled players want to introduce more luck into an activity
- Fluid intelligence - ability to solve problems never seen before
- Crystallized intelligence - ability to use knowledge gained through learning
- Checklists and deliberate practice can help in pretty much every situation
- Disruptors often start at the low end of their market and slowly improve their product and steal market share from the dominant players. (Toyota, mini steel mills)
- Nassim Nicholas Taleb - the author of the Black Swan and this book speaks to predictable and unpredictable events. 4th quadrant events are huge disastrous events that nobody saw coming such as the financial crisis in 2008/09 and argues that having no model to deal with these situations is preferable to having a set model which is wrong
- 10 things to improve guesswork
- Understand where you are on luck/skill continuum. Provides guideline for how much to expect reversion to mean (luck - a lot, skill - less so. We fail to regress as much as we should)
- Assess sample size and swans. Appropriate sample size hugely important. Luck needs larger sample size. 4th quadrant Swans have complex patterns and huge payoffs. Make a decent return for a while and then lose a huge fraction of it in a swan event
- Always consider a null hypothesis. What are you comparing to?
- Think carefully about feedback and rewards. Deliberate practice. With lucky activities, the link between a proper process and desired outcome is not as solid as it is for skill activities
- Make use of counterfactuals. Come up with a story to explain the past when in fact luck played a huge role. Must consider how things could have turned out differently
- Develop aides to develop and improve skill. Humans tend to be overly confident in their predictions. Have a process on place and that way, even if turn out wrong, you know your process was correct and long term you should win. Checklists very important and helps keep you on track in stressful situations. Must measure how your decisions turn out. Keep a decision journal - what decision was made, why, what you expect to happen and then later the actual result
- Have a plan for strategic interactions. If stronger player, simplify to stress stronger skill. If underdog, complicate to introduce more luck. Disrupters often start at the bottom of the market an slowly improve their product and start stealing market share from the dominant players
- Make reversion of the mean work for you. Humans innately try to find causation even if none exists. Understand correlation - 0 correlation and best guess is whatever the average is, 1 means there is no reversion to mean
- Develop useful statistics. What gets measured gets managed
- Know your limitations. Goal of untangling skill and luck is to help us make better guesses
What I got out of it:
- A must read for anybody interested in better understanding themselves and others in order to make better decisions. Better understanding the role that luck and skill plays in a particular area and where you fall in that spectrum is extremely useful regardless of the types of decisions you are making or the game you are playing.