Josh Lyman: (sighs) Yeah. What happened to the good old days when...a couple of hacks with cigars chose the nominee in a smoke-filled back room?
Leo McGarry: They didn't do so bad, did they? Roosevelt, Truman, Eisenhower...
Josh Lyman: We need a back room.
Averting descent into the abyss of metaphysics, consider the other point raised. That point is that somehow the pairwise rating regime is just because it is "open" and "transparent." A system does not yield better results just because it is done in open daylight. This reality is starker when it involves semi-complex statistical models that most fans find confusing.
A mathematical model is not self-legitimizing. A model is only as good as its parameters. Fans and media seem to forget this reality. Those in that camp would do themselves well to consult South Park's "Margaritaville." The episode concerns worshiping the economy as a natural phenomenon of immutable power and utility. Fans and media view the pairwise in much the same way. Both are equally comical.
This is not a call to dismantle the system. It is one to re-evaluate its results. Consider this an invitation for conversation.
Firstly, let's get one piece of rhetoric out of the way. Fairness is treating constituents in the same manner. Fairness is cold and indifferent. Justice, now, is a concept that cracks the door for subjectivity.
The selection process for the national tournament in women's ice hockey is fair. In theory, each team is subjected to the same model and the best eight teams in the nation emerge. This does not shield it from being unjust in its results.
Tradition has become the final refuge of defending the pairwise. It is how the national-tournament field has been decided for several seasons, so it must be how it will be decided evermore. The collegiate sporting landscape bears the fruit of a great thought experiment in another sport. Hockey, for many reasons, does not take its cues from football, but consider the first College Football Playoff and the advent of a four-team national playoff field.
A four-team field for the College Football Playoff generates no less controversy than the eight-team field for the NCAA tournament in women's ice hockey. Let the debates surrounding that postseason's selection guide this thought experiment.
Who did not make the final four of the College Football Playoff that many thought should have? Baylor University and Texas Christian University are the consensus answers on that point. The majority agrees that only one of those two should have made it. Why did neither make it? The Big 12 does not have a championship game.
The committee for the College Football Playoff refused to reward a conference's choice to have co-champions and avoid a deciding clash. The lesson of this expands beyond football. Playoffs matter. Conference playoffs matter on the national stage. They vet out the unhoned as undeserving of a chance to play on the national stage. If a hypothetical postseason for the Big 12, a conference founded in 1994, would have mattered to the College Football Playoff committee, then across sports, the conference tournament of ECAC Hockey, a women's ice hockey conference one decade older than the Big 12, should matter greatly.
Reader, indulge this writer and continue down the line of this thought experiment. If the College Football Playoff expanded to eight teams, which teams would be most likely to have been added next? Baylor and TCU would have their advocates. However, how would the remainder be filled? A large contingent would want runners-up Arizona, Georgia Tech, Missouri, and Wisconsin included. They would be legitimized in that belief. Otherwise, the conference playoffs do not matter.
The final College Football Playoff ranking does not include in the top eight any teams that finished below an effective third place in their conference's respective postseason. That reality is telling. The postseason matters. Winning a conference championship matters. A close second is competing for one zealously.
College football at the highest level is in its nascent stages of developing postseason play. College hockey is not. College football has at most one round of postseason play at the conference level. College hockey on the women's ice hockey side has three rounds in each conference. A larger sample size, including first-round series, ensures that aberrations do not sway which teams advance through the conference-level postseason. The playoffs in each conference siphon off the weaker teams and prove which teams are in the best postseason form in a way unequaled in college football.
Why then is it in college hockey that teams that finished below second and third places in their conference's postseason found their way into the national-tournament field? Oh, yeah, "math." Why are the parameters of the model legitimate? Oh, yeah, because that is how it has been done for years. It is a perversion that teams that could not prove they deserved to compete for being the best in their conference are afforded the opportunity to compete to prove they are the best in the nation.
Clarkson, Minnesota, and Quinnipiac did not qualify for their conference's championship game. Quinnipiac finished third. Clarkson and Minnesota finished fourth in the postseason. Runners-up Bemidji State, Cornell, and Syracuse were excluded. Did their exclusion make the coming tournament better or worse?
The purpose of the committee in crafting a national tournament is twofold: select the most deserving teams or give fans the best tournament possible. A general assumption deferring to conference tournaments accomplishes both. How can a team prove its deservingness by dropping out of its conference's first or second round of the playoffs to an opponent regarded as inferior? It cannot. It is the upsetter, not the upsettee, that proved its deservingness. The best tournament is created when conference champions and runners-up are admitted because teams that have made weeks-deep playoff runs have proven their postseason form.
Giving second life to teams that exited at the midpoint of conference tournaments achieves neither goal. Despite the points made so far, this is not an argument that the system should admit immediately conference runners-up. It is a plea for subjectivity. The same subjectivity that wafted through the smoke-filled rooms that Josh and Leo reference in the opening quote. Canvassers at caucuses did not choose candidates for running a good campaign over the long haul. Instead, they elected those who were resonating with the right chords at the right moment and a general subjective appeal to the timber of which they knew a candidate was, often without empirical evidence to support it.
The College Football Playoff reinvigorated a similar reasoned committee-based approach to collegiate sports. It would serve well college hockey, especially women's ice hockey with a field so small, to follow suit. College hockey once had the smoke-filled room. Some of the greatest programs rose and prospered during that era. Michigan would lack many of its nine national championships without the dense nicotine-laden fog of a conference hall. What would this national tournament look like if the majority of the committee did not feel imprisoned by the pairwise?
Why is Minnesota in the tournament? No, that is not rhetorical. Why are the Gophers in this tournament? Minnesota has had a fine season. It did not have a fine 2015 WCHA tournament. Minnesota, the first seed, fell to fifth-seed Bemidji State in the WCHA semifinals. Oh, yeah, it was a fluke, right? Nope. The Beavers downed Minnesota once in the regular season too.
The Gophers finished fourth in the WCHA tournament. Fourth. North Dakota, Bemidji State, and Wisconsin finished above them. Why are they in the national tournament? If it is because they are so good that they deserve it, then don't the Beavers deserve a shot at the national title for downing such a team of epic talent and undeniable potential?
Apparently not. A more just model would reward the Beavers for a game that they should not have won and punish Minnesota for a game that it should not have lost. Currently, Minnesota was unscathed (yes, the first seed overall is the definition of being unscathed) by the loss and Bemidji State gained no benefit for its victory. The WCHA tournament proved that three teams were more deserving of a national-tournament berth than was Minnesota.
Quinnipiac was another third-place finisher to make the field. A reasoned committee would not have put the Bobcats in the field. How do they make the tournament better? The answer is that they do not. Last season, yes, they would have. Quinnipiac had a phenomenal break-out season last year. It deserved more than the opportunity to begin the national tournament to see if it could seek the national championship on its home ice. The contributors of Where Angels Fear to Tread were disappointed. That narrative and their impressive play earned the Bobcats the right to seek romantic glory.
This season? The Bobcats are just another team finishing below second place that failed to win its conference championship for the second season in a row. In all honesty, their inclusion feels like a make-up choice for an egregious error last season.
Clarkson is the lone team that did not finish as a runner-up that this writer believes deserved to make the national-tournament field on the strength of its conference tournament play and other considerations. Clarkson finished fourth in the ECAC Hockey tournament. The Golden Knights and the Gophers are the lowest conference playoff finishers. Why did the former deserve the right to advance to the national tournament but the latter did not under a reasoned approach?
Clarkson is the defending national champion. The defending champion deserves an abundance of deference in its favor to attempt to defend its title. Last season's national runner-up deserves no such deference.
On the Cornell front, two ECAC Hockey teams deserved to be in the national tournament more than Cornell on a holistic, reasoned approach. Harvard, the Crimson won on the ice in the purest metric. Clarkson, the defending national champion. Were ECAC Hockey to get three teams into the national tournament, the Big Red should have been third.
A field that contains eight programs that four conferences populate with autobids is too influenced by conference postseasons to deny their role in discerning which teams are most worthy to contend for the national title. A reasoned committee would defer to the collective judgment of the results of conference tournaments. Four conferences, four tournaments, four champions and four runners-up. The means of populating the national-tournament field become self-evident.
Katey Stone, the revered coach of Harvard hockey, recently expressed a desire for the field of the national tournament to expand beyond eight teams. Her point is well-taken. Too many teams of proven playoff form are left to sit back at home. However, expanding the NCAA tournament field is not the best means to realize her goal.
An invitation to the NCAA tournament should recognize a team for its elite performance. A team that earns its way into the NCAA tournament in men's ice hockey establishes itself as roughly falling in the top quartile of all programs in the nation. The NCAA tournament for women's ice hockey, with an eight-field team and 34 eligible programs, similarly recognizes teams as being among the top quartile in the nation. Expanding the tournament to 10, 12, or 16 teams devalues the worth of the tournament and the acclaim of invitation.
To achieve Katey Stone's goal, the national tournament should subsume all conference tournaments. Every team begins the postseason in a conference tournament. Every team is immediately eligible for the national tournament. Every team that proves its timber with a decent showing, a subjective standard that a reasoned committee could discern, in the playoffs would be rewarded with an extended season to test its playoff form against the other remaining bests in the nation.
Finally, for those defending the pairwise as producing the most just and best national tournament possible until this point, consider the following. The greatest attribute of the current pairwise system is its predictability in its alleged transparency. Why then when the dust of all competition settled was Cornell tied for seventh in the pairwise until the committee "corrected" or "readjusted" the results to slide the Big Red down to a tie for ninth? Yes, a tie for seventh is no guarantee for invitation to the tournament, but if the system is transparent and not subject to the whims of the committee, why did the Red slide several rungs? Programs that are currently in the field passed Cornell from lower anticipated seeds.
Humorously, with considerations like these, tradition is now the main reason why people defend the pairwise rating system. Without the honesty to admit as much, the field was chosen under some unknown criteria that moved some programs up and others down after all games had been played. A reasoned committee would feel no need, much like the committee behind the College Football Playoff, to justify its opinions. A committee and a process that wraps itself in the emperor's clothing of fair, cold, calculated mathematical reasoning should be afforded no such comfort in the face of such inconsistencies.
The announcement of the national field was delayed 30 minutes. One can hope that the committee was doing some soul searching and realizing that its current system is failed. The committee did not give the fans the best tournament this season. How could it when one of the best playoff players in college hockey is soaking up her last few weeks on East Hill?
We live in an era when human reason is trusted more than mathematical models. Perhaps more than ever, the limitations of analytic measures are becoming obvious. Predictions of elections endure but one mere cycle and the value of players on professional rosters can be miscalculated grossly. The College Football Playoff with its committee of reasonable members manifests this trend toward favoring human judgment and justice over algorithms and fairness. It is time that the NCAA selection committee aligns with this era, especially on the side of women's ice hockey.
As someone who employs, creates, and deciphers analytics and mathematical models, it pains me on some level to argue against the pairwise rating system. It becomes easier when in the last several seasons, the committee, behind the false veil of objectivity, has "corrected" or "modified" the final rankings in women's ice hockey to justify their choices ex post.
Expose this already present subjectivity to the light of day and allow a reasoned committee to select the eight best teams in the nation based upon their overall performances with extreme prejudice in favor of deep runs into conference playoffs. Anyone who thinks this inappropriately punishes teams that had great regular seasons needs to consider about what this is. This is about the playoffs. It is about teams earning their way into the tournament on the merit of their postseason performances.
Heck, if you want to reward regular-season performances, why not create the Director's Trophy? Mark Emmert can present the trophy to the winner in a center-ice ceremony after the winner has completed the best regular season. It would become college hockey's equivalent of the Presidents' Trophy.
The hockey postseason is meant to be earned and contested on the ice. Worth is proven through competition. Rewarding third and fourth place teams with second life is antithetical to the values incumbent in the sport, no matter how great of a regular season teams may enjoy. The process should not overvalue what playoff successes a team should have had. It should reward the successes that teams did have.
So, yes, Leo and Josh are right. The "smoke-filled back room" did well. It would serve women's college hockey fans well if national-tournament contenders were selected in such a manner. Committees would make the conference postseasons increasingly meaningful and give fans the best tournament possible with the most deserving teams. So, yes, let's bring back the smoke-filled room. Wait, on second thought, smoking is unhealthy. Maybe put a fog machine in there.