March is the best month of the year for basketball fans. But this week, when so much of the news and chatter is dominated by complaints about the bracket, is the worst week of March.

I maintain that selecting and seeding a 64-team bracket is an indeterminate and often arbitrary endeavor. I would have given Tennessee a #1 seed over OSU, and if I did give the Vols a #2, I wouldn't have paired them with UNC. But I don't think the Committee's decisions to the contrary are unreasonable (much less conspiratorial).

Pat Summitt, however, says she got rooked. The most revered figure in the history of women's basketball has made national headlines this week with her bitter complaints about the bracket.

If Coach Summitt honestly believes that there is something wrong with the process or criteria used by the Selection Committee, then perhaps this offseason she should use some of her considerable insight and influence to work for a better system. So should any other coach who believes that her team has been unfairly slighted.

I make this challenge in all seriousness. The current system is showing cracks. Maybe it's time for a change.

The most important story about the bracket this year is the Committee's willingness to ignore the wacky RPI. The Committee picked Cal, e.g., over Western Kentucky even even though the Committee's own RPI ranking showed the Bears ranked 64th and the Hilltoppers ranked 20th. I don't care about the specific result — I have no idea whether Cal deserved to be in over Western Kentucky — but I am thilled to see the RPI kicked to the curb like that.

Consider the implications. In picking Cal over WKU, the Committee necessarily concluded that the RPI was off — not just by a little, but way, way off. And this begs the subversive question: If the RPI can be so wrong, why do we use it at all?

The Committee might respond:

We don't put that much weight on the RPI alone (and we've been saying so for years). We use it as an organizational tool, not a final ranking. In other words, we just use it as a preliminary ranking to help us quantify the second-order criteria, which are the important criteria.

Ok... but that argument dissolves almost instantaneously. If "wins over top-50 teams" is an important criterion for selecting and seeding the bracket, then we need to have some reasonably good system for deciding who the top 50 teams are in the first place. Even if the RPI is only used as an "organizational tool" and not as a final ranking, it's still an important tool. If it's not a good tool, we should find another.

More pointedly: If other ranking systems are better, then we should dump the RPI for one of the alternatives.

The RPI is an old and simple formula. It was designed at a time before fast, cheap computers made more complicated ranking systems practical. Those limitations are now gone. Maybe it's time for the RPI to go with them.

Maybe it's time for coaches to stand up and fix the system rather than just griping about it for one week each year.