Inversion and Probabilistic Thinking

More mental models coming your way to expanding your arsenal of frameworks in decision making - Inversion and Probabilistic Thinking.

Inversion

Inversion is backward thinking - instead of looking at how to achieve success, we assume failure and identify the factors for the same.

Two approaches to applying inversion in our life.

• We start with a hypothesis and assume it is true or false. Given that the belief is true or false, we figure out what else would have to be true to support our claims.

• We think in terms of obstacles instead of goals - and think deeply about how we want to avoid them.

The mathematician Carl Jacobi used this model a lot and believed that the road to solving challenging problems is thinking backwards. Here is an example. We wish to improve innovation in the early days of the company.

• Forward Thinking - Think of initiatives to improve and support innovation. Implement these initiatives.

• Inversion - Think of the obstacles to innovation. What are the factors that would stifle innovation in the company? Remove those roadblocks.

Forward-thinking tends to increase the risk factor while inversion reduces the risk overall. Inverting the problem isn’t always about solving the problem, but it is about keeping trouble at bay.

Here are the steps in an inversion workflow:

• Take a core project and fast forward six months or a year and assume it failed.

• Do a thought retro. What are the core mistakes and factors that would have caused us to go wrong?

• Avoid these missteps in the actual implementation.

• To improve general productivity as an early employee

• Ask if we wanted to decrease our focus, what do we do to get distracted? The answers are the frequent interruptions to avoid.

Probabilistic Thinking

Probabilistic thinking use the tools of math and logic to better estimate the likelihood of any expected outcome.

Here are three tools for probabilistic thinking

• Bayesian thinking - use as much relevant information as possible and build context, in the presence of prior context, we try to estimate the outcome.

• Long-tailed curves - For ‘normal’ curves, the probability of the average case is high, and the number of outliers is limited. As the number of extreme events increases, the curve gets flatter. In this case, the probability of an average scenario would be lower, and unprecedented events happen.

• Metaprobability - Sometimes we overestimate our capacity to make predictions. Think of all the times; the election analysts got it wrong. Due to this, we must be cautious of the projections we make.

This model is all about analysing the world and making the right calls. Very few people have honed this skill well. The best we can do is to become ‘antifragile’. Antifragility means to set ourselves up for victory in volatile or unpredictable conditions. Two tools for getting into the antifragile mode:

• Be in situations with significant upside but zero downsides. For example, networking with people on Twitter. At max, people don’t respond. On the other hand, great opportunities can arise.

• No failure should pass by without us learning something valuable. Don’t go for a risk that can make or break. We must be resilient to failure and keep moving ahead.

References:

The Value of Probabilistic Thinking, Farnam Street Blog, May 2018 and Antifragile by Nassim Taleb

Inversion and the Power of Avoiding Stupidity, Farnam Street Blog, Oct 2013