Polls & Ratings & Comparisons, Oh My! Making Sense of College Hockey’s Rankings

Nearly every Monday morning from late September through mid-April, USCHO.com and USA Hockey Magazine post their weekly college hockey media polls. And nearly every Monday morning from late September through mid-April, College Hockey Twitter (it’s a thing! Come join us and get mad at everything!) goes into an uproar over the polls, regardless of the rankings and who is and isn’t ranked.

Some of the various opinions you’ll see on a weekly basis can be summarized as follows:

“Polls are irrelevant!”

“The Pairwise is the only thing that matters!”

“Why do the polls even exist?!”

“Abolish USCHO.”

Ok, fine, the last one was a joke but the question at the heart of every poll-related opinion is this: Do the polls really matter at all?

Ignoring context and nuance, they most certainly do not.

The reality of Division I College Hockey is, unlike football, soccer, basketball, volleyball, and most every other collegiate sport, the annual national tournament is not selected by humans. Rather, college hockey’s NCAA Tournament field is selected by a mathematical formula – the Pairwise Rankings (College Hockey News has an excellent primer on the system if you’re new to college hockey or need a refresher on how it all works) – and is not subject to the whims of east coast college hockey journalists who refuse to stay awake long enough to take in NCHC action (it’s like Pac-12 After Dark but on ice). For all of the valid criticism of the Pairwise system, the reality is, it has done us all a favor and reduced the weekly polls to nothing more than glorified popularity contests and excellent marketing for the sport.

To say it another way, these weekly polls have no effect whatsoever on the outcome of any season.

So then why do we, as a community, place so much importance on them and bellyache when a team deserving of that #1 ranking gets passed over for another equally deserving team?

The answer probably lies in human psychology and the need for external validation but it’s best to leave that to those better qualified to diagnose such oddities about the human condition (Mike Leach is so greatly missed). But the reality remains that we, again as a community, are not placing these weekly rankings in their proper context. That is, we’re failing to remember that such whims of the 50 writers chosen to fill out their USCHO ballots every week have no bearing on the season.

But that still begs the question – if they’re irrelevant, why do they exist at all?

There are a variety of legitimate answers to this question, none of which will satisfy the Pairwise Truthers and Disciples of KRACH.

For one, the Pairwise isn’t generally a very reliable measuring stick until midseason – usually late November at the very earliest – and that doesn’t satiate most sports fans’ constant hunger to compare their favorite teams to others for the first two or three months of the season. We could pretend that comparing all 61 teams doesn’t matter in October and November, but we would be lying to ourselves and ever-important content engagement numbers would suffer.

Recruiting, especially for teams that don’t annually contend for the NCAA Tournament, also relies heavily on the polls. All too often, the college hockey community at large looks at the rankings through the lens of the top contenders, but where the Pairwise Rankings are relevant only for Tournament contenders, the weekly polls are extremely valuable recruiting tools for teams that aren’t always in the Tournament conversation. Telling a potential recruit that you’re a team that regularly gets mentioned in the top 20 of the USCHO poll is a hell of a lot more valuable than it is telling them you’re regularly in the top 20 of the Pairwise.

Probably the best answer, though, is that the USCHO and USA Hockey Magazine polls are just easy to understand. The easier something is to understand, the more widely accessible it is. Even if, in any given week, you disagree with the results of the polls, at least you fully understand how they were put together. It doesn’t take advanced brain power to figure out that Minnesota is #1 this week because they earned 38 #1 votes from the 50 USCHO voters.

The Pairwise and KRACH (now known as the College Hockey News Power Ratings), on the other hand, are based solely on complicated, though not proprietary, mathematical formulas. No matter the sport, casual fans, which dominate viewership across the board, don’t care about the formulas. They want easy-to-digest explanations and rankings and no matter how hard we try to “dumb down” the formulas for better accessibility, weekly polls more easily and quickly satisfy those casual fans.

For comparison’s sake, consider men’s college basketball. If you’re a casual fan of that sport, what ranking first comes to mind? The AP Top 25 is likely the answer and that’s undoubtedly where you’re going to go to find out who the #1 team in the country is. However, if you’re a more avid fan who pays attention to the details of the sport on a daily basis, chances are, KenPom – the sport’s leading objective rating system that ranks all 363 Division I men’s basketball programs – came to mind.

Though the comparison isn’t perfect (the NCAA still uses a committee to select the 68-team March Madness field and the committee leans heavily on RPI, a formula even simpler than the Pairwise), it still does a good job of helping to explain college hockey’s system. The USCHO and USA Hockey Magazine polls are college hockey’s version of the AP Top 25 while the Pairwise and the CHN Power Ratings/KRACH are akin to KenPom.

The difference, of course, is that the Pairwise is the sole criteria used to select the NCAA Tournament’s at-large bids. And that lies right at the heart of the issue. For the Pairwise and KRACH truthers, this renders the polls useless…and the thing is, they’re not totally wrong!

Where their argument falls apart, though, is their common refrain that the polls and formulas can’t coexist, that the only ranking that should exist is the Pairwise. The polls’ inherent value, however, lies in their simplicity and, maybe more importantly, their popularity. Are NHL media outlets going to stop putting out their weekly power rankings pieces because they have no bearing on the outcome of the season? Are you nuts?? Those pieces are routinely the most clicked-on and best engagement-driving pieces every year!

The bottom line, though, is this: like it or not, you don’t cater to a sport’s most hard-core, avid demographic when you’re trying to grow its popularity. The college hockey establishment understands that. If you want to make the sport more popular, you must capture the attention of as many casual fans as possible. The best way to do that, as engagement numbers and overall clicks on weekly rankings, power or otherwise, show, is via the polls.

It might be tough for the hardest-core college hockey fans to stomach but there is a difference between what they want in their ideal world (no polls, only math) and the real world where we actually reside. In the real world, there is plenty of value in the polls – they are extremely useful recruiting tools and incredible marketing material that helps attract casual fans to a traditionally niche sport even though they don’t affect the outcome of the season.

Would it be nice if the college hockey establishment treated the Pairwise with the same reverence as the polls on a weekly basis? Yes!

Would it be wonderful if teams posted their Pairwise ranking alongside their position in the polls in their weekly graphics? Absolutely!

But make no mistake – the polls are not irrelevant. The Pairwise is not the only ranking that matters. When you take them both in their proper contexts – polls are marketing, Pairwise sets the tournament field – they both provide great value to a sport desperate for attention and growth. You just have to understand their context and appreciate nuance to see it.

5 thoughts on “Polls & Ratings & Comparisons, Oh My! Making Sense of College Hockey’s Rankings”

  1. Excellent article. Personally, I have no problems with any of them, and any publicity is good publicity.

    My personal favorite is the annual preseason LetsGODU Super Poll ™, in which Denver is always placed first, CC is always placed last, the rest of the teams are placed where they are reasonably expected to finish at the end of the year.

  2. Great explanation Nick. Thanks.
    I believe you are spot-on as to why both rankings are valuable, but for the different reasons noted.

  3. Dunker likes this
    needed article. The polls are needed for bragging rights which fans love. Also, maybe they are needed to get a feel for the best teams early on before the PairWise is settled in some which may not be till around January 1.

  4. Polls are also good for attendance. For example, “we are playing the 6th ranked team in the country! Let’s go to the game”. It creates ‘value’ for an in-season game and provides relevance to a match-up.

Leave a Reply