In more detail
A mark of 85 with mean 70 and SD 10 has z = 1.5, about the 93rd percentile of a normal distribution. A positive z is above the mean, a negative one below, and z = 0 is exactly average. Z-scores let you compare values from different scales, because they replace units with "standard deviations from the mean". Converting z to a percentile uses the Normal distribution table or function, and is only accurate when the data are roughly bell-shaped.
Why it matters
Z-scores standardise test results, growth charts, credit and quality metrics, and are a simple outlier screen: under a normal distribution only 0.27% of values lie beyond ±3. They need the mean and standard deviation of the right reference group; comparing against the wrong population gives a precise-looking but meaningless number.
Example
Which is the stronger result: 1300 on Test A (mean 1050, SD 200) or 28 on Test B (mean 21, SD 5)?
Test A z = (1300 − 1050) / 200 = 1.25 → 89.4th percentile
Test B z = (28 − 21) / 5 = 1.40 → 91.9th percentile
Test B is slightly better relative to its group. A mark of 55 with mean 70 and SD 10 has z = −1.5, the 6.7th percentile.
Common confusion
A z-score is not a probability. z = 1.5 means 1.5 standard deviations above the mean; the matching percentile (93.3%) or p-value has to be looked up from the distribution.