# What is a “free parameter” in a computational model?

In many articles regarding computational models of some particular phenomenon, there seems to be a consensus: "the smaller the number of 'free parameters' in the model, the better". So, what is meant by "free parameter", and why is it less desirable to model something with such a parameter?

Thanks!

-

A free parameter is one that can be adjusted to make the model fit the data. If I make a model that says $A$ is proportional to $B$, there is one free parameter, the proportionality constant. If my model has a specific value of the proportionality constant, there are no free parameters.If I say that $A$ is a quadratic function of $B$, there are three free parameters, $a,b,c$ in $A=aB^2+bB+c$. That makes it easier to fit the data, even if my model is not correct, so it is a less impressive.