NCAA Tournament Analysis: Comparing At-Large Candidates By How a Typical Tournament Team Would Do Against Their Schedule

Select members of the media did their bracket exercise at the end of last week. I was not one of them. Seth Davis has often asked for people to submit their brackets if they don’t like what the committee does. What follows is my attempt to do that, first by laying out a way to compare teams, and then later, to use it to build a bracket in part II.

If you have read my bracketology projections every Friday, you know that I have been projecting how I think the field will turn out. That, though, is trying to guess what the committee will do, as well as projecting how the rest of the regular season will play out. This will be an exercise trying to measure results only to date.

Last week, I talked about New Mexico and Pittsburgh and how the RPI treats strength of schedule. I also talked about how the committee should try to objectively measure how many wins above or below an average at-large candidate each team is. So today, I did just that.

Using Ken Pomeroy’s ratings, I tried to assign a win likelihood for each game on every tournament contender’s schedule, then looked at actual results to see if they had more or less wins. To calculate, though, I used what the 25th best team would have been expected to do against the same schedule, on the idea that represents about the average at-large team.

This method will give a little more certainty than things like “Top 50 wins” (where wins over Indiana and Florida are waaaaaaay better than beating a bubble team), “bad losses” (which are overused, referrring to both competitive conference road games against NIT type teams and sub-200 teams), as well as accounting for the site of the game.

So, basically, I’m looking at how an average at-large team would have been expected to do against each team’s schedule to date, with average luck, and then comparing it to the actual win total. No margin of victory, so overtime wins and losses count the same as blowouts. This is all about what a team has “accomplished” but tries to put it in some context of how others should have fared across the same schedule. How do the teams stack up, and how do you compare, say, Iowa to Villanova to Indiana State? Here you go. This is every team that shows up in the Top 60 of either the RPI, Ken Pomeroy, or both, as well as a few other select teams that are on bracket lists.

Not only can this type of analysis be done on a full schedule, it could be useful for objectively viewing things like how Duke has done without Ryan Kelly compared to a typical tourney team, how others with early injuries have done since returns, and how Kentucky will fare looking at their closing schedule.

[photo via USA Today Sports Images]