How do I personally do this?

When I assemble my brackets, I follow all the same rules the selection committee has to. However, much of the at-large selection and seeding process is subjective, so my process differs from the committee. While I do my best to mirror what I think the committee is going to do, I can't be in their heads. Every year, different data points seem more important, and sometimes even from team-to-team in the same year.

The main piece of information I use is simply "who have you beat?" When I rank teams before assembling any given bracket update, I go through each team's resume to see who they beat and who they lost to. I also use the quadrants for this. This doesn't mean I completely ignore the metrics, because I factor that in, but to me, it seems like basing the seeding on who a team has beat gives me the best chance to predict what the committee is going to do.