SESKStuff Everyone Should Know
SESK: Stuff Everyone Should Know · sesk.info/ai-precision-formula

A model flags 200 emails as spam; 150 really are. What is its precision?

0.75, which is 150 out of 200.

Start playing free

  1. 0.25
  2. 0.75 ✔
  3. 1.33

Why: Precision asks: when the model said positive, how often was it right? True positives divided by all positive predictions: 150 over 200 is 0.75.

Huh, didn't know that: The 50 wrong flags are false positives: real email sent to the spam folder.

The dad joke
My spam filter has 75 percent precision. The other 25 percent was my grandma's emails.
Like this card?
Play the game

References

Answer: 0.75. Precision asks: when the model said positive, how often was it right? True positives divided by all positive predictions: 150 over 200 is 0.75.

  1. Google Machine Learning Glossary: precision developers.google.com/machine-learning/glossary
    When the model predicted the positive class, what percentage of the predictions were correct?
    Checked 2026-10-10.

Think this answer is wrong? Tap "Challenge this answer" on the card in the game and tell us why.

Test yourself in the game

Artificial IntelligenceWhat makes a regression model different from a classification model?Artificial IntelligenceWhat does a confusion matrix show for a classifier?Artificial IntelligenceHow does k-nearest neighbors classify a new point?

That's stuff everyone should know.