The entire group you want conclusions about — every case, not just the ones you measured.
Visual intuition
same population
+2 cases
Absolute risk keeps the whole group in view
The population is the whole crowd — every case counts, including the ones your survey never reaches.
Example
If you want the commute of a city's 500,000 workers, those 500,000 are the population — even if you only survey 1,000 of them.
How It Works
Every secretly names a population: all German adults, all bulbs off one production line, every match a team ever played. You rarely get to measure them all, so you study a and generalize — which only works if you stay clear about which whole your handful came from. A survey of gym members says something about gym members, not about everyone.