SESKStuff Everyone Should Know
SESK: Stuff Everyone Should Know · sesk.info/ai-accuracy-imbalanced-trap

If 1 percent of cases are positive, what accuracy does always guessing negative score?

99 percent, while catching zero positives.

Start playing free

  1. 99 percent ✔
  2. 50 percent
  3. 1 percent

Why: Accuracy counts all correct calls. When positives are rare, a useless model that always says no is right 99 times in 100. That is why accuracy misleads on imbalanced data.

Huh, didn't know that: Google's course advises: avoid accuracy for imbalanced datasets and use another metric.

The dad joke
My fraud detector is 99 percent accurate. It just says "no fraud" every time.
Like this card?
Play the game

References

Answer: 99 percent. Accuracy counts all correct calls. When positives are rare, a useless model that always says no is right 99 times in 100. That is why accuracy misleads on imbalanced data.

  1. Google ML Crash Course: Accuracy, recall, precision developers.google.com/machine-learning/crash-course/classification/accuracy-precision-recall
    a model that predicts negative 100% of the time would score 99% on accuracy
    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 IntelligenceWhich metric answers: of all the real positives, how many did the model catch?Artificial IntelligenceHow do AI, machine learning and deep learning fit together?Artificial IntelligenceWhat did Alan Turing call his 1950 test of machine intelligence?

That's stuff everyone should know.