PDA

View Full Version : 2019 NCAA Tournament Thread - Regionals are the best weekend of hockey all year



Pages : 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 [29] 30 31 32 33

Slap Shot
04-04-2019, 02:39 PM
I disagree. What's the point of having conferences, then? Here's an example:

I'll use the America East conference in basketball. Here are the top two teams:

<table>
<thead>
<tr>
<td>Team</td>
<td>Conf.</td>
<td>Overall</td>
</tr>
</thead>
<tbody>
<tr>
<td>Vermont</td>
<td>14-2</td>
<td>27-7</td>
</tr>
<tr>
<td>Stony Brook</td>
<td>12-4</td>
<td>24-9</td>
</tr>
</tbody>
</table>

Only ONE team from this conference goes to the NCAA tournament. Yet, 800 teams from the B1G go every year, even though some of these teams lose repeatedly to the top five teams in the conference. Why? Because the conference is considered "stronger." Great. But if you PLAY for one of these teams, you have no control over the schedule, or who you play. You're just trying to win games. Why should Stony Brook be penalized because they didn't play a "strong" schedule? The players did what they were asked to do...they won 24 out of 33 games. Compared to the competition THEY PLAYED AGAINST, they did well. That's all they can control. Why should a team that finished SEVENTH in their league get to go to the tournament over them? If they couldn't WIN against teams in their conference, why do they get picked over teams that won, in some cases, ten more games? I don't buy the argument about "strength of conference." Who cares? Just make Division 1 ONE conference then and pick the top 64 teams. The goal is not to get the "best" 64 teams...it's to reward those teams that excelled against THEIR competition. Otherwise, just relegate all those conferences to Division 2 because what's the point? If you're in a stronger conference, that means you recruited BETTER talent so you SHOULD be better. Which means you should be able to WIN against better competition. So by denying a team based on "strength of schedule," you have diminished WINNING. Why would I go to a school where I have to basically go 30-1 (if I don't win my conference tournament) to qualify?

This isn't an argument against AQs. This is an argument that a team should be judged against the competition it plays. If it is successful, it should be rewarded. Otherwise just have one giant conference and take the top x number of teams.

Stony Brook can control their NC schedule. And I'd rather the 64-tournament be as competitive as possibly after all the auto-bids are filled. That means sometimes a 2nd place conference finisher with a strong WIN% is left out because they didn't win enough games against strong NC competition.

chickod
04-04-2019, 02:52 PM
Stony Brook can control their NC schedule. And I'd rather the 64-tournament be as competitive as possibly after all the auto-bids are filled. That means sometimes a 2nd place conference finisher with a strong WIN% is left out because they didn't win enough games against strong NC competition.

Not my point. The B1G team that is 17-15 controls their OOC schedule too, but they still get it because of the league they play in.

ExileOnDaytonStreet
04-04-2019, 03:18 PM
Not my point. The B1G team that is 17-15 controls their OOC schedule too, but they still get it because of the league they play in.

I feel like this is a de facto argument against caring about strength of schedule, and Iím not on board with that straw man.

Because Iím not sure how you stop accounting for that in such a way that the 7th B1G team is denied an advantage over the 2nd AE team without having a law of unintended consequences where teams start only scheduling ****ty non-con games or adding crappy teams to their conference as cannon fodder.

MTUHuskies
04-04-2019, 03:22 PM
How the hell did I get subscribed to a **** bouncyball thread? Who cares how they pick their 64 +4 teams?

Sent from my SM-G960U using Tapatalk

ExileOnDaytonStreet
04-04-2019, 03:37 PM
Who cares how they pick their 64 +4 teams?Millions of people.

LikeÖ literally millions of people.

MTUHuskies
04-04-2019, 03:45 PM
Millions of people.

LikeÖ literally millions of people.But this is not the forum to discuss it.

Sent from my SM-G960U using Tapatalk

ticapnews
04-04-2019, 03:52 PM
Millions of people.

LikeÖ literally millions of people.

Not really. They care about filling out brackets and spending otherwise productive hours at work looking up what the **** a Chanticleer is or the 3-point % of the shooting guard at New Mexico State. They couldn't give two ****s about how those teams are picked.

Fishman'81
04-04-2019, 06:58 PM
This discussion seems to have veered-off the road a bit.

There is no way to prove a negative, even if a team goes 30-1 vs. demonstrably weak competition... Who's to say that that team wouldn't have gone 30-1 vs. better competition? There's no comparative data available there upon which to make that claim. So, all one has to go on is who some team played, and how it fared.

What we have right now is as good as it's ever been. I do like that. But the silly conference tournaments AQ's still exist, and they do not take into account the preponderance of data provided by a RS.

This "calculus" worked out fortuitously (for the most part) this year in both D-1 and D-3 -only a couple of worthy teams were jobbed by the system- but it is perennially a recipe for a hot mess, wherein a significant % of better teams lose berths to inferior teams.

dxmnkd316
04-04-2019, 07:23 PM
This discussion seems to have veered-off the road a bit.

There is no way to prove a negative, even if a team goes 30-1 vs. demonstrably weak competition... Who's to say that that team wouldn't have gone 30-1 vs. better competition? There's no comparative data available there upon which to make that claim. So, all one has to go on is who some team played, and how it fared.

What we have right now is as good as it's ever been. I do like that. But the silly conference tournaments AQ's still exist, and they do not take into account the preponderance of data provided by a RS.

This "calculus" worked out fortuitously (for the most part) this year in both D-1 and D-3 -only a couple of worthy teams were jobbed by the system- but it is perennially a recipe for a hot mess, wherein a significant % of better teams lose berths to inferior teams.

A 30-1 team will make the tournament every time. But nice straw man.

Fishman'81
04-04-2019, 07:26 PM
A 30-1 team will make the tournament every time. But nice straw man.

Well, yes. That's what I just said.

manurespreader
04-04-2019, 07:45 PM
I agree that strength of schedule should be considered, but in our sport I think it gets too much consideration.

Fishman'81
04-04-2019, 08:02 PM
I agree that strength of schedule should be considered, but in our sport I think it gets too much consideration.

Not picking a fight here, but how else can teams be rated comparatively, without some objective SOS metric involved?

Win % and SOS are really the only two salient stats involved, IMHO.

Slap Shot
04-05-2019, 08:58 AM
I agree that strength of schedule should be considered, but in our sport I think it gets too much consideration.

MSUM earned a #1 seed with the 40th rated SOS, UMASS the same with the 23rd. Cornell and Harvard earned at-large berths with the 19th and 22nd rated SOS's respectively, Providence the 26th and Bowling Green with the 46th.

Numbers
04-05-2019, 09:15 AM
I am copying this piece data here, as well as to a connected discussion on the Minnesota thread:

I'm not sure how to do the analysis of this data, but in regard to regionals, starting in 2006 and not including 2019,
#1s are 31-21 in round one
#2s are 27-25, with slightly better records in the last 7 years than before then.

This set of numbers seems to make sense in a sort of a way.....There should be little difference between 2s and 3s, given that it's often true in the final PWR that one game result different (which isn't much - just one fluky goal) can make the difference in seeding a 2 or a 3.
More parity, since 1s are not dominant, for sure. But the #1s still have a considerable edge.

So, in a vacuum, these seem like some what representative results.

However, what is interesting, but difficult to do scientifically, is something like this:
In games where one school had a significant travel advantage (mostly the 2 eastern regionals, for obvious reasons, but sometimes in the west as well)....and significant is difficult to define, the 'home teams', regardless of seeding, are running about 60% winners.

Take that for what you will in the idea of 'what's a good system'....

As for SOS....The best metric if all you want to count is game results, is KRACH adjusted for home ice.
The current PWR is close to that. Only replace ASU with PSU to get the usual top 15, and I am looking at post-regional results in some case, so that might affect ASU.

dxmnkd316
04-05-2019, 09:17 AM
Not picking a fight here, but how else can teams be rated comparatively, without some objective SOS metric involved?

Win % and SOS are really the only two salient stats involved, IMHO.

Margin of victory could be appropriate as well. Itís used in several advanced rating systems for football and basketball.

dxmnkd316
04-05-2019, 09:19 AM
Also, when using ratings systems, you need to understand if they are supposed to be predictive or not. Some systems are designed to pick winners of future hypothetical games. Other (most?) are designed to pick the teams that performed the best over the season.

Numbers
04-05-2019, 09:29 AM
Also, when using ratings systems, you need to understand if they are supposed to be predictive or not. Some systems are designed to pick winners of future hypothetical games. Other (most?) are designed to pick the teams that performed the best over the season.

Predictive systems, which are interesting to the user of course, seem inappropriate for an NCAA tournament. It seems much better to reward the teams for their performance already done. Of course, this is a point given to argument.

What is more interesting in this idea:
In the early days, the NCAA tournament was a sort of bonus. Most teams only played schedules of teams close to them. So, the tournament was a way to match regions of the country against each other as well.
It has changed, of course.

dxmnkd316
04-05-2019, 09:31 AM
Predictive systems, which are interesting to the user of course, seem inappropriate for an NCAA tournament. It seems much better to reward the teams for their performance already done. Of course, this is a point given to argument.

100% agree.

chickod
04-05-2019, 10:29 AM
Margin of victory could be appropriate as well. It’s used in several advanced rating systems for football and basketball.

Margin of victory is worthless in hockey. It may have some value in basketball, because of the larger sample size of points scored. I don't see how you can definitively determine anything by analyzing MOV in hockey. What? Does a 5-2 win "mean" more than a 6-1 win? Everyone knows that you can win 7-0 one night and lose the next. It's not a sport where enough "points" are scored to rely on a margin of victory stat.

And the only thing that irritates me more are those stupid "total goals" series that we used to have. The object of the game is to WIN. I don't care if it's 1-0 or 25-2. There is NO DIFFERENCE. And I don't even like total goals being used as a tiebreaker for some of these tournaments (like World Juniors, for example).

Kepler
04-05-2019, 10:40 AM
Margin of victory is worthless in hockey. It may have some value in basketball, because of the larger sample size of points scored. I don't see how you can definitively determine anything by analyzing MOV in hockey. What? Does a 5-2 win "mean" more than a 6-1 win? Everyone knows that you can win 7-0 one night and lose the next. It's not a sport where enough "points" are scored to rely on a margin of victory stat.

And the only thing that irritates me more are those stupid "total goals" series that we used to have. The object of the game is to WIN. I don't care if it's 1-0 or 25-2. There is NO DIFFERENCE. And I don't even like total goals being used as a tiebreaker for some of these tournaments (like World Juniors, for example).

The only reason they ever used total goals was to justify a two-game series and not be embarrassed by playing a mini-game after a team wins 8-0 and loses 3-4. The fact that both these rationale were stupid helps explain why that loathsome format is no longer used.