The video introduces the challenge of comparing seemingly disparate medical measurements, such as blood pressure (140) and cholesterol (220), to determine which one is more concerning. These measurements use different units and scales, making direct comparison difficult.

Many biological measurements follow a predictable pattern called the normal distribution, or bell curve. This curve indicates that most values cluster around the average. The 68-95-99.7 rule explains that approximately 68% of data falls within one standard deviation of the mean, 95% within two standard deviations, and 99.7% within three, highlighting the rarity of extreme values.

To compare different bell curves, a 'universal yardstick' is needed. This is the standard normal distribution, a special bell curve centered on zero with a standard deviation of one. It acts as a universal translator, allowing any bell-shaped data to be converted to a common scale.

A z-score is a new unit of measurement on the standard normal distribution. It quantifies how many standard deviations a value is from the average. It's calculated by subtracting the group's average from the individual value and then dividing by the group's standard deviation. A positive z-score means above average, negative means below, and zero means exactly on average.

The video applies the z-score calculation to the initial medical examples. A blood pressure of 140 (average 120, standard deviation 10) results in a z-score of +2.0, indicating it's two standard steps above the average and higher than 97.5% of the population. A cholesterol of 220, if calculated, would yield a z-score of +1.33, showing it's less extreme than the blood pressure. This demonstrates the power of z-scores in providing meaningful comparisons.

Z-scores are a versatile tool used in various fields, including public health (tracking children's growth), medical labs (interpreting blood tests), and research (comparing drug efficacy). They provide a universal language for understanding how normal or rare a measurement is, without being constrained by original units.

The video concludes by posing a thought-provoking question about where else this universal yardstick for comparing data, such as z-scores, could be applied. Examples include comparing student test scores, athletic performance, and economic data, highlighting its broad potential to change how we interpret information across different domains.