We are almost through the regular season, and it is time to start thinking through who will be the top teams in post season and how one's own team is likely to match up.
If you happen to be a fan of one of the top teams, history suggests that the first two rounds are likely to be wins. The real nail biting starts when the field is winnowed down to sixteen.
When you look at the top sixteen - for the sake of using a concrete measure, I'll look at the top sixteen teams as measured by RPI - you think about how you think you can do against each of them. If you've already played them once, or have them yet to play on your schedule, you have an insight into what can happen post-season. Neither a win nor a loss guarantees the same result in the next meeting, but an actual game provides a rich source of information for discussion.
When the team isn't on your schedule, you have to look at less direct information. You can look at the rankings, which is to say, you can see what other knowledgeable people think about the relative strength. They can also look at other measures, such as the RPI, or the Team Rankings measures or Sagarin rankings. Interesting numbers, but that's for the casual fan. The more serious fan wants to analyze, not just to accept the conclusions of others.
If you want to look at something more than other people's opinions, and you don't have head to head results, looking at common opponents is the next best thing. Of course, it isn't as good as analyzing a head to head game, but it is second best.
If you are a UConn fan, you don't have to resort to second best very often. UConn has played or will play ten of the current top sixteen teams, leaving only five teams not on their schedule. No other team in the country can boast more of the Top 16 on their schedule. Based on the RPI top 16 (as of February 1), UConn plays the toughest schedule in the country.
The six teams not on UConn's schedule, in order by RPI, are:
Let's look at each, starting at the end.
California
California has two common oppoents with Connecticut, for a total of four games. All have been played except the final CT-RU game. California lost all three games against common opponents, as the following table shows:
California did manage to play strong against Stanford at home, and against RU, but came up empty. The resutls are also shown int his graph:
Texas A&M
Texas A&M has no
common opponents with UConn. This means fans cannot look at head-to-head
results or common opponents, and will have to resort to something else. Let's
leave that exercise for anther time.
Oklahoma
Oklahoma and Connecticut have exactly one common opponent,
South Carolina. Oklahoma beat South Carolina by 32, while UConn won with a 58
point margin. No, I wouldn't conclude that UConn is 26 points better than Oklahoma.
Let's face it; both teams are far better than South Carolina, and both won by
a considerable margin. Having said that, I can't imagine that Oklahoma fans
would complain if the results were reversed. While it isn't clear how much
stronger UConn is than Oklahoma, the common opponent's margin goes to strengthen
the belief that UConn does deserve to be favored should they match up post-season.
Baylor
Baylor has two common opponents with UConn. Both have played
Stanford, and St. John's. The Baylor-St. John's resulted in a
23 point victory by Baylor; while UConn beats St. John's by 57. A common theme
in this anlysis of common opponents is that one cannot draw much of an inference
from a single game - it would be foolhardy to conclude that UConn is24 poitns
better than Baylor based upon this single observation, but when one can look
at many games, a picture emerges.
Baylor lost to Stanford by 24, while UConn beat Stanford by 12. Again, this doesn't automatically translate into a predicted UConn margin of 36 over Baylor, but it certainly suggests that UConn fans feel far better than if, say Baylor had beaten Stanford by 15. Some people who saw the Baylor-Stanford game might feel it wasn't a good measure of the relative strength, but as a person who did see the UConn-Stanford game - it wasn't that close.
The graph shows how each did against the two common opponents:
Happily, when we get to the top two remaining teams, there are a lot more common opponents to analyze.
Maryland
Maryland and Connecticut have eight teams in common, playing
a total of nine games (Maryland plays Duke twice). The following table shows
the results so far, using only common games (that
is, LSU is not in the average).
As before, the results aren't definitive. UConn fans will be happy to see that they've prevailed by eleven more points, on average, compared to Maryland, and will look forward to seeing how the other games turn out. In addition, it is comforting to see that in most games, the exception being LSU, UConn has won by more than Maryland, although the relative margin against Pittsburgh was very close. Both won by substantial margins.
The column graph shows the same information.
Tennessee
Finally, UConn and Tennessee have nine teams in common,
possibly the most number of teams in common for two teams not in the same conference.
It's probably not reasonable to conclude that Connecticut is 9 points better than Tennessee. Some of the larger Connecticut margins happened before they lost starters Kalana Greene and Mel Thomas. While they haven't lost a game since losing those players, there is no doubt their contributions are missed. Some analysts might want to look at home versus away games, but with home advantage worth between 3 and 4 points per game, that isn't going to change the results very much. Tennessee fans may hope that the Notre Dame game is representative, as that is one game in which Tennessee won by a larger margin. Those fans may wish to ignore the fact that UConn was up by 25 with two minutes to go, when the deep bench gave up a chunk of the lead against a mixture of starters and bench for Notre Dame. They may also want to believe that DePaul is representative. Tennessee certainly won handily, and UConn did not look like the number one team in the country just managing a last minute win. In the end though, UConn won more games, and by larger margins against common opponents. None of which guarantees anything should they meet in post-season, but a UConn fan can look to the common opponent results and look forward to such a meeting.
Tennessee also "won" their game against Rutgers by a bigger margin that Connecticut, but consider that the Connecticut game was at Rutgers, while the Tennessee game was at Tennessee, and it won't give much comfort to those tryng to use margin of victory against common opponents to argue that Tennessee is the stronger team.
It is fair to say that one cannot conclude much by looking at, say, the ODU game, where UConn won by 43 versus "only" 32 by Tennessee. Both of those games were clear evidence that the winning team was a far better team, and it is torturing the data to insist that a larger winning margin is evidence that the team with the larger margin is better. Yet, we aren't forced to look at a single common opponent, we have seven data points so far, with two more to come. The pattern is clear, and while this analysis does not tell us with any certainty that UConn can beat Tennessee, a neutral fan would have no trouble preferring the team with the 24 points margin to the team that wins by 12.
This page last updated on 1 March 2008.