To improve, calibrate. To calibrate, employ expected value calculations (which is more fun than it sounds).

Two years ago, a friend passed on the opportunity to invest in Anthropic. He now regrets passing.

This friend prioritizes calibration. Yesterday I learned why he generally avoids risks. When viewing a risk, he considers only the magnitude, failing to also include the risk’s probability.

Expected value is one of game theory’s most powerful calculations: 

• Probability × Magnitude = Expected Value. 

My friend is not unique: people often drop one of the two terms. Include both in your simple math; your outcomes will be better calibrated. (And it only takes a second!)

Magnitude-Only: Frozen From Fear

Considering only the downside magnitude over-biases against action. This heuristic will keep you alive but not enable you to thrive.

This example occurs whenever somebody says, “But what if it fails?” as a conversation ender. Slap a probability on that bad boy and you’ve got an EV.

Probability-Only: Missing the Magnitude

A friend once described why they don’t vote: “the likelihood I’ll have an impact is near-zero.” Ah, but tomes have been written about the enormous magnitude.

When someone says, “It probably won’t happen,” or “it’s a drop in a bucket,” check they’re also considering the size of the bucket. If the impact is large enough, the low likelihood that your single drop makes it overflow may be enough.

EV → More Fun

During my all-time favorite date, Partner and I hooted & hollered at an alligator.

Twenty minutes earlier, we strolled down a path through a Louisiana swamp. Partner meandered toward some tall grass. I said, “I wouldn’t do that.” Partner was surprised because her self-preservation instinct usually eclipses mine. I mentioned alligators, and said: in this circumstance, the magnitude is very bad (being eaten by a gator) and the probability much higher than usual (Louisiana swamps contain alligators, and Pokémon has taught me to beware tall grass).

Compare that moment to later: on a bridge 10 feet above an alligator. The magnitude is still very bad, but the likelihood de minimis. (Teehee, alligators can’t jump 10 feet!) 

Since we knew ourselves to be safe, we could poke the gator with our metaphorical stick.

When something seems scary, investigate the probability. E.g. “I can’t miss work to see my child’s school play: I could lose my job.” 

This concern is well-founded, but misses two elements: the likelihood of losing your job (probability) and the fact that you can influence that likelihood via other actions (malleable probability). (I’ll write more on malleable probability another time, especially as it relates to luck.)

Expanding your Vocabulary

Expected Value is a core step, but it is only the first step.

Error bars can make this topic tougher. The question “How likely is extra-terrestrial life to exist in our universe?” prompts wildly different outcomes whether you include the error bars.

So even if you do the EV calculation, you’ll still be wrong sometimes, you’ll still lose sometimes, and there’s still more calculation to do. (For a fun example, see Pascal’s Mugging.)