Being a member of the Selection Committee is sort of like being a ref -- you're everyone's whipping boy.
Everyone blames you for their lack of success. Everyone marshals a few facts to prove that you are not simply wrong, but obviously wrong. Everyone is sure that you are a biased and incompetent idiot.
And yet, if put in your place, none of them could do any better, and most would fail spectacularly.
A week ago, I tried to take a close look and rank the bubble teams. I kept thinking I had it almost right, but then I would look again and see something that didn't make any sense. Team A seemed better than Team B, which seemed better than Team C, but C seemed better than A. I could never settle on any choices. After several hours, I simply gave up.
The fundamental truth is this: ranking sports teams who play different schedules is a deeply indeterminate exercise. There is no good way to do it. You have to answer unanswerable questions, such as:
(1) How much do you adjust for radically different schedules? How do you compare teams like Gonzaga and Virginia Tech? The latter has 11 losses, but 8 of those were to top-25 teams, and it has also beaten 3 top-25 teams. The former has played only 3 games all year against top-50 teams (one win, two losses). One has a lot of wins against an easy schedule, the other has fewer wins against a tough schedule -- how can you even guess which team is better?
(2) How much do you value consistency? Louisville, for example, has several nice wins a great record inside the top-100, but also some bad losses outside the top-100. Xavier, by contrast, has fewer quality wins, but not a single loss to a sub-100 team. Which matters more -- getting good wins or avoiding bad losses?
(3) How much do you weigh various criteria? The "nitty-gritty" report has 14 different elements. Is each valued equally, or are some more important than others? Do the values for the elements change as you move down the S-curve?
(4) How much do you weigh your own observations? Suppose you saw Purdue play recently in the Big 10 tourney, and they looked very good, but you haven't seen BYU play at all this year. Is it fair to factor in your own impressions as an extra plus on Purdue's side?
There isn't even a good way to go about debating these questions, much less to reach determinate answers. We are left running on ice.
The indeterminacy isn't so deep that objective meaning vanishes altogether. Some consensus is still possible. Everyone agrees, for example, that Tennessee deserved a higher seed than Delaware.
But the indeterminacy is deep enough that for some questions, there is and can be no consensus answer (no matter how much you look at the data). Not everyone agrees, for example, that Richmond deserved to get in ahead of Delaware. There is no non-arbitrary way to settle the dispute.
The questions are hard. Reasonable people disagree. At some point, we have to stop debating and make a decision so we can play the damn tournament.
We have, quite properly, granted decisionmaking power to an independent Committee that represents a variety of interests. It follows some neutral rules. It spends a lot of time looking at a lot of data. It does its best. It makes a call. So we can play.
I don't understand the post-hoc whiners. I want to ask: Have you invented some new perfect system to rank teams? Have you discovered the behind-the-veil principles of basketball equity that will settle all disputes? Have you thought up some better system than the one we have?
I doubt it. And consequently, I quickly tire of the whining and complaining.
Maybe you got a seed that's a bit lower than you were hoping. But if you want to get to the Final Four or take a title, you're going to have to beat a lot of good teams no matter what your seed. As Jim Boeheim said, you can overcome a bad seed, but you can't overcome having a bad team.
The decisions at the bubble are more painful; if you end up on the short end of those choices, you don't even get a chance.
But I still don't have much sympathy. Every year, there are ten or so teams on the bubble who have a roughly equal claim to five or so spots. The Committee is left with the responsibility of making an arbitrary choice among those teams. The teams, however, bear responsibility for losing and/or for scheduling powderpuffs, which is what put them on the bubble in the first place.
To Gonzaga I say: if you want to get an at-large bid next year, play some tougher teams. It's not your fault that the WCC has a bunch of bad teams in it, but it is your fault that you scheduled so many nonconference games against bottom of the barrel squads like Utah Valley State, Providence, Eastern Washington, Nevada, Boise State, Portland State, Denver, and Sacramento State.
If the Zags would have gotten in ahead of Va Tech, I would have said to the Hokies: if you want to get an at-large bid next year, win some more games. It's not your fault that the ACC has a so many good teams that are tough to beat, but it is your fault that you lost to Miami and Wake. If you would have won even one of those games, you wouldn't be in this position.
To the Committee I say: you did a very good job, as usual.
Sure, I would have seeded a couple teams (including my Gophers) slightly differently, and I would have made a couple different choices at the bubble. But on the whole, the Committee got it right. And given the indeterminacy inherent in the exercise, its choices are as correct as mine and as correct as theoretically possible.
Everyone blames you for their lack of success. Everyone marshals a few facts to prove that you are not simply wrong, but obviously wrong. Everyone is sure that you are a biased and incompetent idiot.
And yet, if put in your place, none of them could do any better, and most would fail spectacularly.
A week ago, I tried to take a close look and rank the bubble teams. I kept thinking I had it almost right, but then I would look again and see something that didn't make any sense. Team A seemed better than Team B, which seemed better than Team C, but C seemed better than A. I could never settle on any choices. After several hours, I simply gave up.
The fundamental truth is this: ranking sports teams who play different schedules is a deeply indeterminate exercise. There is no good way to do it. You have to answer unanswerable questions, such as:
(1) How much do you adjust for radically different schedules? How do you compare teams like Gonzaga and Virginia Tech? The latter has 11 losses, but 8 of those were to top-25 teams, and it has also beaten 3 top-25 teams. The former has played only 3 games all year against top-50 teams (one win, two losses). One has a lot of wins against an easy schedule, the other has fewer wins against a tough schedule -- how can you even guess which team is better?
(2) How much do you value consistency? Louisville, for example, has several nice wins a great record inside the top-100, but also some bad losses outside the top-100. Xavier, by contrast, has fewer quality wins, but not a single loss to a sub-100 team. Which matters more -- getting good wins or avoiding bad losses?
(3) How much do you weigh various criteria? The "nitty-gritty" report has 14 different elements. Is each valued equally, or are some more important than others? Do the values for the elements change as you move down the S-curve?
(4) How much do you weigh your own observations? Suppose you saw Purdue play recently in the Big 10 tourney, and they looked very good, but you haven't seen BYU play at all this year. Is it fair to factor in your own impressions as an extra plus on Purdue's side?
There isn't even a good way to go about debating these questions, much less to reach determinate answers. We are left running on ice.
The indeterminacy isn't so deep that objective meaning vanishes altogether. Some consensus is still possible. Everyone agrees, for example, that Tennessee deserved a higher seed than Delaware.
But the indeterminacy is deep enough that for some questions, there is and can be no consensus answer (no matter how much you look at the data). Not everyone agrees, for example, that Richmond deserved to get in ahead of Delaware. There is no non-arbitrary way to settle the dispute.
The questions are hard. Reasonable people disagree. At some point, we have to stop debating and make a decision so we can play the damn tournament.
We have, quite properly, granted decisionmaking power to an independent Committee that represents a variety of interests. It follows some neutral rules. It spends a lot of time looking at a lot of data. It does its best. It makes a call. So we can play.
I don't understand the post-hoc whiners. I want to ask: Have you invented some new perfect system to rank teams? Have you discovered the behind-the-veil principles of basketball equity that will settle all disputes? Have you thought up some better system than the one we have?
I doubt it. And consequently, I quickly tire of the whining and complaining.
Maybe you got a seed that's a bit lower than you were hoping. But if you want to get to the Final Four or take a title, you're going to have to beat a lot of good teams no matter what your seed. As Jim Boeheim said, you can overcome a bad seed, but you can't overcome having a bad team.
The decisions at the bubble are more painful; if you end up on the short end of those choices, you don't even get a chance.
But I still don't have much sympathy. Every year, there are ten or so teams on the bubble who have a roughly equal claim to five or so spots. The Committee is left with the responsibility of making an arbitrary choice among those teams. The teams, however, bear responsibility for losing and/or for scheduling powderpuffs, which is what put them on the bubble in the first place.
To Gonzaga I say: if you want to get an at-large bid next year, play some tougher teams. It's not your fault that the WCC has a bunch of bad teams in it, but it is your fault that you scheduled so many nonconference games against bottom of the barrel squads like Utah Valley State, Providence, Eastern Washington, Nevada, Boise State, Portland State, Denver, and Sacramento State.
If the Zags would have gotten in ahead of Va Tech, I would have said to the Hokies: if you want to get an at-large bid next year, win some more games. It's not your fault that the ACC has a so many good teams that are tough to beat, but it is your fault that you lost to Miami and Wake. If you would have won even one of those games, you wouldn't be in this position.
To the Committee I say: you did a very good job, as usual.
Sure, I would have seeded a couple teams (including my Gophers) slightly differently, and I would have made a couple different choices at the bubble. But on the whole, the Committee got it right. And given the indeterminacy inherent in the exercise, its choices are as correct as mine and as correct as theoretically possible.
posted by Ted at 2:35 PM
