i am trying to develop an lc-ms-method for 50 components. when trying to compare different columns/elluents i want to compare the ellution of these 50 components. when i make an overlayed chromatogram of the 50 chromatograms i get a real mess an i am not able to see which conditions give the best seperation. so wat i need is an calculated number that will represent how good the components are spread over the run-time of the chromatogram. anybody an idea which number i shoud calculate and how?

The classical way of looking at this is the minimum resolution. However, I was never completely happy with this approach.
I just got an article published in the Handbook of Analytical Separations Vol 4, pp 185-214, that describes a criterion that is suitable for judging the overall quality of a separation. It is 0, if there is overlap, and 1, if all peaks are cleanly separated. If you give me your e-mail, I'll send you a copy.

You need only be concerned with separating peaks that have the same spectra--in theory. The biggest advantage in using MS is because it isn't necessary to separate everything with the LC. So the resolution of those peaks are the key numbers.

To measure analysis speed, peaks/minute or total run time are useful.