Can Of Worms
The Committee's Seeding Flip Flop: No Big Deal or Worth the Debate?
by Adam Wodon/Managing Editor
Why did the committee use non-sensical rationale for flipping Denver and Colorado College's seeds for this year's NCAA tournament?
I don't know.
But first let's answer a more fundamental question.
Why do we care?
There are some out there who ridicule those who love to tout "the numbers" and the over-reliance on them — the way we over-analyze the numbers; the way we drone on and on about the esoteric tweaks we want to see to the Pairwise; our whining over things like the "Teams Under Consideration Cliff"; the advocation of more "pure" systems like KRACH that only math Ph.Ds can fully understand (really, it's not that hard).
But using an objective-based system of some kind is generally regarded in hockey to be a good thing. And so we fight to educate, to make sure it's adhered to, and to improve it.
And when it's suddenly thrown out the window for no apparent reason, even if the effect is admittedly relatively miniscule, it's alarming.
First, let it be noted that I have regularly advocated for the committee to afford itself leeway for some common-sense manipulation during the seeding process, things that might avoid large injustices or create better matchups.
But I will contend that this simple little decision by the committee — to switch No. 2 CC and No. 3 Denver — has implications far more wide-ranging than would seem by the simplicity of the move.
We can quibble over the details of this objective-based system. We can argue that there should be more wiggle room. But that decision throws into question, for the first time, the entire concept of the objective-based system.
It was 1997. We were a few years into this so-called purely objective system that the NCAA men's ice hockey community decided to implement to pick the field for the NCAA tournament.
But, even among those who closely followed such things, the exact implementation was pretty vague. And without properly being able to explain it, and without a large enough vehicle to explain it, conspiracy theories were rampant. Even coaches had no idea, most of the time, how teams were picked or seeded. The smoke-filled room was over, but no one seemed to believe it.
That's when Joe Marsh shed some light on the issue. The St. Lawrence head coach was then the chair of the men's ice hockey committee. In an extensive interview, he explained how the field and seedings came about that year.
"It's all in the numbers," he said. It was simple. It was matter of fact. A step-by-step, building-block process, that even if it contained a couple small judgment calls here and there, could easily be explained and defended and re-created.
From there, evangelism was given its fuel. Those who appreciated the merits of such an objective-based system had a basis for explaining it to others, for singing its praises. And it had an outlet.
Over the years, the online media was so successful in educating the hockey community on what the Pairwise was — a summation of the committee's own process — that the hockey community itself just simply began to accept the PWR as is.
By 2003, the first year of the 16-team tournament, the process was complete. That season, the committee took the top 16 in the Pairwise (accounting for automatic bids, of course), and simply seeded the field as is (accounting only for host teams staying home, and avoiding intra-conference first-round matchups).
This was good and bad.
It was good because it indicated the committee's commitment to the objective system, and marked an acceptance, in a sense, of the educational process. It was hard to educate the public when even the committee itself sometimes poo-poohed what the media was showing. But now it was clear, to everyone, that the Pairwise was indeed just a summation of the committee's own process. We were all on the same page.
It was bad, too. By seeding the field so strictly, it tied the committee in to a system that wasn't perfect. Most of the hockey community embraced the PWR system as something with great merits — to avoid the smoke-filled back-room method of picking the field. But that doesn't mean PWR is flawless. Many of the criteria that make up the PWR have "issues," rehashed ad nauseum around here.
So the system isn't perfect, but it's at least an objective system. Widely considered to be a good thing.
Look Out Below
The committee has a specific set of criteria it uses to compare one team to another.
In this season's case, Colorado College defeated Denver on three of the criteria, and Denver, by virtue of its win in the WCHA tournament final, won the head-to-head matchup, 3-2. In the entire history of this system, this means that CC wins the comparison and should be seeded No. 2.
The committee decided to change this. It claimed that all the other criteria was a "dead heat" and that it came down to the head-to-head matchup.
As a result, Denver is set up to play the 15 seed while CC plays the 14 seed.
There was a time when the committee only used these individual comparisons to compare teams, one-by-one, one after another. Over time, the comparison "wins" became totaled, and the notion of individually comparing teams became less prevalent.
There have also been times that things were moved around in the seeding process for attendance reasons, or other similar factors, always with a good defense, even if arguable. And that's OK.
But this is the first time in 10 years the committee has ever specifically overridden its own criteria! Completely threw it out the window and said it was irrelevant.
It opens a completely unnecessary can of worms. Not because I want the committee to be so strict with its numbers. But precisely because this move was so innocuous that it opens a can of worms to "play around" with other seemingly innocuous moves, until those moves aren't so innocuous anymore.
What stops the committee from saying that Cornell deserves a No. 1 seed over Minnesota? Cornell went 17-0-1 down the stretch and won the ECAC tournament and regular season. There was a time that would guarantee a top seed. Minnesota went 10-10-1 down the stretch and lost two games at the WCHA Final Five.
What if Vermont defeated Colgate in the ECAC tournament consolation? In that case, Dartmouth would have qualified for the NCAA tournament based on the criteria. What would have stopped the committee from saying, "Wait, Vermont just defeated Dartmouth in the ECAC playoffs, let's put Vermont in."?
Some out there may be saying, "Yeah, why not? Those moves make sense to me."
And maybe you are right. Maybe those types of decisions should be made.
But the system is based on the idea that the entire season matters. And it was created to remove perception from the process. It doesn't mean the process is perfect, it's just more apparent, and less open to perceptions driving it.
The committee has now opened itself up to doing whatever it feels like, based on perception. This is the very thing the hockey community was trying to avoid. And I have heard almost zero outcry from coaches or other officials that they want to the objective-based system to go away.
So, Now What?
I have advocated over and over for better systems that are more pure and will give better results. There's many ways to accomplish this task, starting with KRACH.
But very, very few people want the objective-based system to disappear.
Sure, one solution is to scrap it all and allow subjectivity to pervade the process. You could argue that. I think it's a bad idea, and I think most people in hockey agree. It's simply not worth the controversy it creates. You think the current controversy is bad? Most of that is solely from the masses that remain ignorant of the process. Most other controversy — the type of stuff us writer types like to has over each spring — tend to be nitpicky, or esoteric, or frustrating little things as we advocate the desire to get a little bit more perfect.
So the better solution would be to start with a ranking system that's a whole lot better, and use that as your basis. And from there — because even the best system has a margin of error — allow the committee to legislate itself some wiggle room in seeding the tournament to avoid clear negatives and/or create clear positives.
But don't just scrap your system mid-stream on a whim, even — or especially — for something so seemingly small.
Think about this: To what ends did the committee make this decision? What benefit derived out of opening this can of worms?
As I've said before, if you have a compelling reason to make the change — specifically, avoiding an obvious injustice — please feel free to wiggle around all you want.
But why was this decision made? There is absolutely no compelling reason to have done it. Denver could just as easily been a No. 3 in Grand Rapids, and CC a No. 2 in Amherst, and no one would be the worse for wear. Or just call CC a No. 2 and Denver a No. 3 and then move them whereever you want.
That makes it somewhat irrelevant, on the one hand, but also scary, on the other.
This move was made for no reason — outside of the simple, yet scary, decision to completely throw the criteria system out the window.
The tournament will be great, as usual.
But that doesn't mean I can't be alarmed that warning signals have been sounded for next year.
If that makes me a geek, so be it.