When we make decisions in the face of uncertainty, we are making a “bet” on how the future will turn out. When the future arrives, we have a chance to take stock, to assess whether this outcome can teach us anything about how we make our decisions. In some cases, the outcomes might lead us to challenge some assumptions and beliefs that influenced our decision making in the first place.
Before any of this assessment can begin, though, we need to ask some questions about the outcomes:
- If we’re honest, how much effect did our decision have on the outcome?
- How much did external factors (not considered in the decision) affect the outcome?
- Can we truly establish a causal relationship between our decision and the result?
- Were we good? Did this result from our skill in decision making?
- Or did we just get lucky?
This is how we close the loop on decision making. The learning starts with this attempt to “field” the outcome into what Annie Duke calls the “second bet”.
The first bet was the decision itself: we considered possible futures, and made a choice based on a prediction and a judgment. The second bet occurs when we close the learning loop: we decide whether we can learn anything by attributing the result to the decision made.
This second decision is whether we attribute the result to skill (in which case we have a learning opportunity), or luck (in which case we don’t).
This decision around the second bet is rarely clear cut, however. The ambiguity of the environment will leave us facing uncertainty again. But despite the challenge, it is critical that we make a choice, to close the loop. Otherwise, we suffer from what Duke calls “resulting”, where all outcomes are considered to be the direct effect of our decisions, good or bad. This hampers our learning, and amplifies the negative effects of many cognitive biases.
An additional challenge with fielding outcomes is that the answer is rarely 100% skill or 100% luck. So the question becomes, “Did my decision have enough of an effect on this result that there is something to be learned here…?”
One cognitive bias dominates the others when fielding outcomes. The “self-serving bias” will steer us to take credit for the good results (“it was my skill!”) and blame the bad results on luck, (“can you believe my bad luck?”).
Duke recommends building a new habit around fielding outcomes. Instead of allowing the self-serving bias to always make us feel good about a positive outcome (“resulting”), she suggests shifting the reward such that you feel good about making the assessment. That is, congratulate yourself for fielding the outcome! In her poker circles, Duke said the best players were always willing to sit down after a tournament and assess where their decision making went wrong, even when they won the tournament! And they learned to feel good about this effort.
And it’s easier to think objectively about this ambiguous question (of skill vs. luck) when you treat it as a second bet. Since it will generally be a little of both, you can bring the probabilistic mindset (from viewing decisions as bets against the future) to fielding outcomes, and minimize the impact of the self-serving bias.
Note that admitting that some decision you made may not have impacted this positive outcome requires considerable humility. But good leaders bring humility in spades.
And lastly, this challenge of developing a new habit of fielding outcomes gets easier when you can enlist the help of your peers. Holding decision making retrospectives with your leadership teams (and forming communities of practice around the evolution of your decision architecture) can create virtuous cycles of learning and create positive peer pressure to support this new habit.
To summarize the technique:
- Approach making decisions as a “bet” against possible future outcomes.
- Periodically review the outcomes that result from your decision making.
- Field the outcomes to select the subset that are worth studying (where decision making “skill” had an effect)
- Hold a decision making retrospective with the team or with peers to share the assessments, seek learning, and drive improvements in the decision architecture
- Share the learnings across a wider community of practice in the organization.