Had you been paying attention to NCAA selection procedures in recent years, you would have seen this coming.

Selection Misperceptions

There are a number of misperceptions about the way the NCAA ranks and selects its teams that made Harvard's selection chances look worse than they actually were.

First off, notice that the NCAA does not rank teams like the NFL tie-breaking system. The NFL teaches us that team rankings should be determined by record, then head-to-head contests, then league records, records vs. common opponents, and then strength of the schedule. The NCAA approaches a similar list of criteria in a wholly different manner.

Another common misperception is that the NCAA actually takes the time to do a pair-wise comparison of every team using human observation, matching their relative performances head-to-head, and arranging the teams in such a way as to reduce the number of contradictions between the rankings and the head-to-head results.

To see why this does not happen, I'll leave an exercise to the reader. Find the online schedules of Hartford, Connecticut, Dartmouth, Boston College, Boston University, Brown, Princeton and Harvard. Now try to rank these teams fairly based solely on head-to-head results. Are you confused yet? Mission accomplished.

A more prominent cause for confusion--one that afflicts just about everyone--is the difference between the way that polls react to losses and the way the NCAA reacts to losses.

The way polls generally work is that they change based on marginal results. In other words, your ranking only rises significantly if you beat a good team, and it only falls significantly if you lose to a bad team.