Least squares regression is suitable for a continuous dependent (response) variable IFF the residuals (error terms) are approximately normal.

For a dependent variable based on count data, Poisson regression is a first choice.

For a dependent variable based on binomial data (yes/no), logistic regression is a first choice.

For a dependent variable based on data where value ranges from 0 to 1 (baseball scores), beta regression is a first choice.

For a continuous dependent variable that does not follow any known parametric distribution, robust regression (non-parametric) is a first choice

Simple Regression is the most basic concept of fitting a line to univariate data.

It is better to learn linear regression, scatter plot, test for regression coefficients, model test for linear model after go higher.