Weighing Chance Against Cost
A probability becomes useful when you multiply it by what is at stake.
In this lesson
- It starts with a question. Mere sells mats at an open market. On days like today there is a 30% chance of rain, and rain ruins $50 of mats. A tarpaulin costs $10 to hire for the day. Thinking about many days like this, is the tarpaulin worth hiring?
- The idea you will find. Expected loss is the chance multiplied by the cost. — You found $150 over ten days, or $15 a day, which is 30% of $50. That average is called the expected loss. No single day costs $15: Mere loses either nothing or $50. But over many days the average is what her choices add up to, so it is the fair number to set against the price of protection. One caution: an average does not capture a loss too big to recover from, and that is a separate question to ask.
- How you learn it. By doing: a slider, sorting.
- Length. 6 short steps. Progress saves on this phone.
Part of Judging Forecasts
What "70% chance of rain" actually promises, and how to judge whether a forecaster is any good.