Ontario University Athletics, of course, tweaked the basketball RPI-offs, so- called, to make all games count in the standings. That could be read as a response to what happened on the men's side last season. The RPI only included games among playoff teams, remember. Ryerson grabbed first place since its loss to Guelph did not count because those Gryphons, with nine wins in the conference, were on the sideline whilst Laurentian (6-13, third in the North behind Ottawa and Carleton) and Toronto (5-14, third in the East led by Ryerson) participated. Unintended consequences come in threes: the bracket had to be revised York, which had initially qualified, forfeited games for using an ineligible player.
The logic of the change sounds like, make all games matter again. Another complaint that was addressed was that coaches wanted a better idea about potential playoff opponents, so the one-through-17 table became a season-long reference point. It is admirable that this is a level of sport that sees the playoff format as a living document that can be updated quickly and promptly.
While it's an at-large playoff format (rather than divisional), using RPI tethers to the lodestar(s) in its division. Who knows, perhaps Nipissing and Laurentian should get a few bonus percentage points as a trade-off for being tag-teamed twice a year by Carleton and Ottawa.
Here's how the men's RPI and women's RPI standings were distributed division by division, prior to Friday's action:
Without making direct comment on the quality of teams, it looks like the men's standings are skewed since the two best teams, or at least two of the best three, are Carleton and Ottawa, both in the North. One tries not to hung up on reputations, but Nipissing and Laurentian are fifth and seventh, sandwiched around McMaster.
It must seem confusing that McMaster is immediately below the team they defeated rather soundly on Jan. 6 and immediately above one who defeated them on Jan. 7, but that always happens.
Each of those northern teams, who are travel partners, has signalled they are headed in a positive direction. The larger contributing factor to the Lakers' and Voyageurs' places in the RPI is having had eight of 13 games already against the rest of the top eight in OUA RPI: Carleton and Ottawa twice apiece, plus Brock, McMaster, Ryerson and Toronto. That strength-of-schedule factor will be diluted over these next three weeks, but Laurentian and Nipissing are always going to get a little extra credit from who is in their division.
A few percentage points in either direction can move a team up or down two spots in the standings.
The "tethering" is further illustrated by the lack of a strong team in the West. Windsor (13th in RPI with an 8-6 conference record) is retooling and Laurier (9th at 7-8) is going through a culture change under first-year coach Justin Serresse.
What prompted this post idea was last weekend's results. Algoma picked up a fairly landmark victory by defeating Windsor 79-78, with their undersized guard Sean Clendinning hooping 24 points. It was quite a bounce-back after losing by 42 points to the same Lancers on Friday. The alright-Algoma automatic response quickly faded, through. Windsor has run the OUA West for a few years, but isn't strong statistically this season. The net result of Algoma's big win was dropping from 13th to 14th, illustrating I do better math in my gut than on paper.
The effects from the extra division games will be lessened by the end of the regular season, but it seems reasonable to think the home-and-away series have too much effect on the standings.
RPI was created to rank 300-plus teams in NCAA Division I, not leagues, where everybody plays everyone once and plays a handful of teams twice. The main critique is that it can often reward a team as much for who it played as how well it played.
It is understandable why it has been embraced in OUA and Canada West. It's part of the lexicon of college basketball, which is product many people in university sports consume enthusiastically. But to strain another tortured tech analogy, RPI is like the app that ranks every restaurant in a large city, when what you really want is one that will help you order food without waiting in line.
It is best to keep this open-ended and express hope the format gets discussed more after the season and gets farther than the inert "we can't change the format again so soon" response. What is in place may be an improvement, but it's not yet close to perfect.
What one would like to see happen:
- At the end of the season, determine how the RPI standings would have been affected if divisional games were half-weighted. In essence, make each opponent 1/16th of the grade.
- Also run the results through Simple Ranking System, which takes both margin of victory (not factored into RPI) and strength of schedule into account. See if it turns out differently.
- Since using winning margins often makes administrators paranoid about coaching running up scores, calculate it with margin of victory capped. Try it with a 25-point cap, or 30, or 35. Also try only capping margin of victory against lesser teams, arguably defined as those which win fewer than 30 per cent of their games. It's one thing to shelter the fledgling teams, but the Ottawa Gee-Gees should accept consequences for losing to Carleton by 41 points.
- Semi-related, all of those ideas allow for getting back to a 20-game schedule. It is really odd that Carleton and Ottawa only come to Ryerson every second season. It would be great to see an à la carte approach where the other 12 schools could line up an extra game.
Who knows, maybe there is a case for the status quo. It would not be bad if Nipissing got a home playoff date out of this deal.
Collectively, OUA has been very good at keeping the league fresh: single-site final fours, divisional re-alignment, RPI. Being progressive means owning the missteps, too.