Canadian Basketball: the REAL March Madness

cis_champs_2012

Canadian basketball fans are probably tired of seeing this.

 

While most sports fans are getting ready for the NCAA’s yearly March Madness basketball tournament, March is also the time for another, much less widely known basketball tournament. From March 8th to March 10th, the Canadian Interuniversity Sport (CIS) Men’s basketball Finals will be taking place at Scotiabank Place in Ottawa, Ontario. While most of Canada’s top young players play basketball in the NCAA, this tournament is the pinnacle of the Canadian post-secondary basketball season; many of the players participating will go on to play in professional leagues around the world. Here are the teams that are participating this year:

Rank Team Conference
1  Carleton OUA
2  Cape Breton AUS
3  Ottawa OUA
4  UBC CWUAA
5  Acadia AUS
6  McGill RSEQ
7  Lakehead OUA
8  Victoria CWUAA

Of course, the big question is: who will win? Using point differential, we can come up with an idea of the strength of each team. The following numbers include both regular season and playoff results:

Rank Team W/L Record PD/Game
1  Carleton 22-1 27.4
2  Cape Breton 21-1 13.2
3  Ottawa 17-6 10.8
4  UBC 22-5 11.0
5  Acadia 16-6 10.9
6  McGill 13-5 5.9
7  Lakehead 16-8 9.0
8  Victoria 19-7 6.0

Unsurprisingly (to those who follow CIS Men’s basketball), Carleton is the clear favourite this year (and every year since at least 2003). Their win-loss record is an impressive 22-1, but it’s their point differential that makes them stand out from everyone else. At 27.4, this means that — on average — every game that the team plays, it wins by over 27 points. Cape Breton, with a nearly identical 21-1 record, has “only” managed an average point differential of 13.2 — and they posted the next best average point differential!

Beyond ranking the strength of each team, we can also look at each matchup and identify the likely victor. Here are Friday’s games:

  • Quarter-final #1: Cape Breton (13.2) vs Lakehead (9.0).
  • Quarter-final #2: Ottawa (10.8) vs McGill (5.9).
  • Quarter-final #3: UBC (11.0) vs Acadia (10.9).
  • Quarter-final #4: Carleton (27.4) vs Victoria (6.0).

Carleton and Ottawa should win their first games rather comfortably. Cape Breton should beat Lakehead, although the margin will probably be smaller. And the game between UBC and Acadia is a toss-up. Given these results, the Saturday matches would be as follows:

  • Semi-final #1: Cape Breton (13.2) vs Ottawa (10.8).
  • Semi-final #2: Carleton (27.4) vs UBC (11.0)/Acadia (10.9).
  • Consolation #1: Lakehead (9.0) vs McGill (5.9).
  • Consolation #2: Victoria (6.0) vs UBC (11.0)/Acadia (10.9).

In the semi-finals, Carleton should win their game comfortably, regardless of who they end up facing. Cape Breton should be favoured to win over Ottawa, but it’s going to be close. In the consolations, either UBC or Acadia should beat Victoria. Lakehead should win against McGill, but that game will be closer. Given these results, the Sunday matches would be as follows:

  • Championships: Cape Breton (13.2) vs Carleton (27.4).
  • Bronze medal game: Ottawa (10.8) vs UBC (11.0)/Acadia (10.9).
  • 5th place game: Lakehead (9.0) vs UBC (11.0)/Acadia (10.9).

Of these matches, the only easy pick is Carleton over Cape Breton in the championship game. The bronze medal game is too close to call, and while UBC or Acadia would be the favourites over Lakehead, the difference in the teams’ point differentials isn’t very large.

Now, all these predictions must come with a caveat: these predictions are solely based on average point differential. It does not take several things into account, including strength of schedule, injuries, and random chance. Random chance is the biggest obstacle to predicting the outcome of single elimination tournaments, as even the best of teams can be derailed by a single poor performance. For example, Carleton lost to Saskatchewan in the second round of 2010 edition of the tournament, despite having a 6.8 edge in terms of average point differential.

In order to determine the better of two teams with statistical confidence, it takes more than just a single game. In fact, even if a team will beat its opponent 66% of the time, it would still require a “best of 23″ game series to avoid having the lesser team advance. If the odds of winning drop down to 55%, a series would have to be 269 games long in order to determine the better team! So there is a very good chance that at least one of my predictions will be incorrect. But as long as I give myself the best odds of being correct, over time I will make more correct selections than incorrect selections. And in a world filled with uncertainty, that’s the best I can hope for.

- Devin

Comments are closed.