Why can standard scores (z-scores) be used to compare scores across different samples?

 Why can standard scores (z-scores) be used to compare scores across different samples?