So let's pretend that we're playing with toy blocks. We have some red blocks, which stand for "X" values, and some blue blocks, which stand for "Y" values. We're trying to figure out how many blue blocks go with each red block.
When we try to match up the blocks, it's not going to be perfect. There might be a little bit of space between the red block and the blue block. Those little gaps are called "residuals."
The residual sum of squares is just a fancy way of saying "add up all the little gaps between the red and blue blocks." We want to make that number as small as possible, because that means we have the best possible match between the red and blue blocks.
So basically, it's just a way of measuring how closely two sets of data points fit together. The smaller the residual sum of squares, the better the fit.