Average versus median player performance

A receiver's last five appearances show 10, 20, 25, 35 and 160 receiving yards. If you only see their 50-yard average, you might picture five games around 50. Four of these five games actually finished below that number.

The average, also called the mean, includes every yard and spreads the total evenly across the appearances. The median instead asks where the middle result sits when you arrange the games from lowest to highest.

Neither is a secret forecast for the next game. They are two ways of describing what already happened. Reading them together gives you a better picture of a run with one spectacular performance.

Add the five receiving totals: 10 plus 20 plus 25 plus 35 plus 160 equals 250 yards. Divide by five appearances and the average is 50.

The 160-yard game is part of the player's output, so it belongs in this calculation. You don't have to remove it to explain that it has a large influence. The chart lets you see both facts at once.

If you compare this player with someone else's average, keep the game counts and time windows alongside the numbers. Five appearances and a full season give different amounts of evidence, even if both averages say 50 yards.

Sort the same five values: 10, 20, 25, 35, 160. The third result is 25. Two games sit below it and two above it, so 25 is the median.

With an even number of games, use the average of the two middle results. For example, the median of 10, 20, 30 and 160 is 25 because the middle pair is 20 and 30.

The median is less affected by one extreme result. That makes it useful when you want to see the middle of the run. It doesn't show how much output the huge game added, which is why it shouldn't replace the average everywhere.

Use the average when you want to summarize total production across appearances. Use the median when you want a middle-of-the-run view. Then inspect the individual results if the two numbers are far apart.

In our example, “50-yard average with a 25-yard median” quickly tells someone that the run includes an unusually high game. It doesn't establish how often the player will clear a particular threshold next time.

A longer sample, role information and the matchup can add context. The useful first step is to name the measure correctly: average and median are different summaries, not competing definitions of the same number.