"Why can't they use real people and real numbers?"
I would suggest that simple models can help clarify things, in a way real situations do not. People use simple models in math, physics, medicine, sociology, statistics, economics, finance, chemistry, ecology, climatology, astronomy, etc. In almost every field of inquiry, people attempt to build simple models to try to get at the underlying truth of some situation. Real events are full of complex interactions, which often involve more variables than can be easily understood. The human brain is not set up to understand real events in which 200 variables all have some influence.
Issac Newton was offering a simple model when he said all matter attracts matter, proportional to its mass and inversely proportional the square of the distance - his model does not explain everything in the Solar system, it especially failed to describe the orbit of Mercury, but it did offer wonderful insight all the same.
Lister and Pasteur were both offering a simplified model when they advanced the germ theory of illness, and yet their model offers wonderful insights into disease.
Gauss offered simple models for more things than I could possibly list here - some have argued that he was the most important mathematician of all time. He contributed something to group theory, geometry, analysis and he largely invented statistics. And what is statistics, but a set of methods for quantifying simple models or sorting relevant data from the irrelevant?
With only slight exaggeration, I could make the argument that the progress of science is the progress of simplifying models.
That is the argument for not using "real people and real numbers".
"Why can't they use real people and real numbers?"
I would suggest that simple models can help clarify things, in a way real situations do not. People use simple models in math, physics, medicine, sociology, statistics, economics, finance, chemistry, ecology, climatology, astronomy, etc. In almost every field of inquiry, people attempt to build simple models to try to get at the underlying truth of some situation. Real events are full of complex interactions, which often involve more variables than can be easily understood. The human brain is not set up to understand real events in which 200 variables all have some influence.
Issac Newton was offering a simple model when he said all matter attracts matter, proportional to its mass and inversely proportional the square of the distance - his model does not explain everything in the Solar system, it especially failed to describe the orbit of Mercury, but it did offer wonderful insight all the same.
Lister and Pasteur were both offering a simplified model when they advanced the germ theory of illness, and yet their model offers wonderful insights into disease.
Gauss offered simple models for more things than I could possibly list here - some have argued that he was the most important mathematician of all time. He contributed something to group theory, geometry, analysis and he largely invented statistics. And what is statistics, but a set of methods for quantifying simple models or sorting relevant data from the irrelevant?
With only slight exaggeration, I could make the argument that the progress of science is the progress of simplifying models.
That is the argument for not using "real people and real numbers".