Think Clearly:

I bought the book "Think Clearly - Eight Simple Rules to Succeed in the Data Age". It's a light read.

A review will follow soon.

-----

1. Accept How Complex the World Is

"The world isn’t simple. Things interact; many variables; unintended consequences. Don’t expect neat causality or one-size-fits-all explanations.

We tend to oversimplify because it’s easier. That risks misinterpretation or being misled.

2. Think in Numbers

Use quantitative thinking: look at rates, proportions; compare reasonably. Rough back-of-envelope estimates are useful.

Relying solely on stories or intuition can hide scale issues or exaggerate rare events.

3. Protect your Samples from Biases

Be aware of how data is collected; sample bias, selection effects, survivorship bias etc.

Many conclusions fail because the sample isn’t representative.

4. Accept that Causation is Challenging

Correlation ≠ causation. Need to think about confounders, experiments vs observational studies, natural experiments, etc.

It’s easy to see patterns and assume causation (politics, media, health, etc.)

5. Don’t Underestimate the Power of Randomness

Random variation is everywhere. What looks like pattern may be just noise. Also rare events do happen.

Overfitting, seeing meaning when none is there; neglecting probabilistic thinking.

6. Predict Without Ignoring Uncertainty

Make predictions or decisions acknowledging what you don’t know; use confidence intervals or ranges; expect surprises.

Overconfidence and ignoring what might go wrong tends to cause big mistakes.

7. Accept the Trade-offs

Every choice has costs and benefits; you can’t optimize everything. You often need to balance competing values (speed vs accuracy; fairness vs efficiency etc.).

People or organizations try to maximize one dimension and ignore others, which causes unintended consequences.

8. Don’t Trust Your Intuition

Intuitions are useful in some domains, but they are also often biased. It’s better to test them, use data, and be aware of when intuition misleads.

False confidence; heuristics that are useful sometimes but misleading in others."