Normal Distribution

A symmetric, bell-shaped probability distribution where most outcomes cluster around the mean and become increasingly rare further away. About 68% of outcomes fall within one standard deviation.

Like the distribution of adult heights: most people are near average, with fewer and fewer at extreme tall or short ends.

Why it matters

Many player stat distributions approximate a normal curve, which lets us calculate the probability of clearing a given prop line using well-known math.

How DMP uses this

DMP models most counting stats (points, rebounds, assists) as normally distributed to compute P(over) — the probability a player exceeds a given line value.

Common mistake

Assuming every stat follows a normal distribution — some stats like three-pointers made are better modeled by other distributions.