subtraction

The passage around this formula

The two quantities point in opposite directions as sampling increases. With per-attempt success probability p and independent attempts, compare $1 - (1-p)^k \qquad \text{against} \qquad p^k$. The first rises toward one; the second falls toward zero. They are computed from the same runs and answer entirely different questions, and reporting the first while the deployment depends on the second is the most common quiet misrepresentation in agent evaluation.

Read this part in the article →