Your college names one Athlete of the Week, and this week it’s down to two women. Leah ran the 100-meter dash in 11.45 seconds. Maya swam the 100-yard freestyle in 48.08 seconds.
They never raced each other. So the committee compares each one with the women she actually raced.
Two races, one picture
Every dot is one woman’s time, the 77 sprinters on top and the 81 swimmers below. In each row, the dashed line is that field’s mean, and the axis counts seconds from it, faster to the left.
Sprinters: mean 12.278 seconds, standard deviation 0.413 seconds.
Swimmers: mean 50.059 seconds, standard deviation 1.994 seconds.
Leah beat her field’s mean by about 0.8 second. Maya beat hers by about 2 seconds. And look how the sprinters are bunched together, while the swimmers are spread wide.
So whose performance is stronger, compared with the women each one raced?
You chose Leah’s.
That’s right. Maya’s gap is bigger in seconds. But the sprinters are so bunched that only one of them beat Leah, while 13 swimmers beat Maya.
You chose Maya’s.
Not quite. Maya’s gap is bigger in seconds. But the sprinters are so bunched that only one of them beat Leah, while 13 swimmers beat Maya.
You said you’re not sure.
That’s okay. Maya’s gap is bigger in seconds. But the sprinters are so bunched that only one of them beat Leah, while 13 swimmers beat Maya.
Count in standard deviations
Places tell you who’s ahead, not by how much. To compare the two gaps, measure each one in its own field’s standard deviations.
In seconds, a second is the same length in both rows. In standard deviations, each row is measured in its own field’s standard deviation: about 0.4 second for the sprinters, about 2 seconds for the swimmers.
Watch Leah’s and Maya’s dots, and press Count in standard deviations.
Counted in standard deviations, the sprinters spread out, the swimmers draw in, and Leah’s dot passes Maya’s. No one’s time changed. Only the ruler did.
Leah’s time is 2.00 standard deviations below her field’s mean, and Maya’s is 0.99 below hers.
The z-score of an individual data value tells how many standard deviations that value is from its population mean. Here, each field is the population.
A time below the mean gets a negative z-score: Leah’s is −2.00, and Maya’s is −0.99.
What it adds up to
So whose performance is stronger? Leah’s. In seconds, Maya beat her field’s mean by more. But the swimmers are spread much wider than the sprinters, so a second counts for less in the pool.
Counted in their own fields’ standard deviations, Leah’s time is two below her mean, and Maya’s is about one.
When you compare two events, compare z-scores: count each gap in its own event’s standard deviations.
