2022 D3 Men's Soccer National Perspective

[Flying Weasel]

Every year the link doesn't work until right before the show starts.  Frustrating, but that's how it is.

[Flying Weasel]

I know paclassics has mentioned a number of times his desire for the committee to define/divulge how they weight the different criteria. Which feels like a call to make it more rigid and formulaic, more objective, removing the human element.  I disagree with that.  I wish it was less structured and less data driven and more flexible and subjective.

I think paclassics wants it to be more formulaic so that the selections are more predictable and transparent.  From the NCAA's standpoint, I think making it more rigid and formulaic allows the selections to be much more defensible and much less questionable than would be the case if the committee was given more flexibility and the process allowed for more subjectivity.  I guess I'm more comfortable with the "grey" that subjectivity would bring rather than striving for "black and white" clarity where the reality is complex and nuanced and not always accurately captured mathematically.

There's so many observations, comments, and arguments that could be made to support either side of this issue, but I'll just leave a couple that have me firmly on the side of desiring more subjectively and flexibility.

As it currently stands, for the purposes of carrying out ranking/selection duties, a committee member has no reason to watch a real game and formulate an opinion about a team.  In fact, an opinion formed from actually watching a team play (including their own team) doesn't seem like it could officially be used in the ranking/selection process.  In other words, numbers/data not only trump first-hand observation, they are the only thing that can be considered.  I'm not a fan of that.

The Massey ratings are 100% data driven and thus purely objective. I personally do not trust them. Overall it's not bad, but for me there are still way too many cases of a team having the numbers but not pasing the eye/smell test. If we followed Massey, wouldn't almost the whole NESCAC get into the tournament each year? A .500 or sub-.500 Colby should not be in the NCAA Tournament, IMO, but very well could be if the criteria becomes purely mathematical.  So, no, thank you.

In general, I want to see the best teams get selected with the at-large berths. Without allowing for subjectivity, I think it's harder to achieve that objective. And I would be good with some flexibility to select an undefeated Western Conn. even if the numbers don't quite make the cut. As I could be OK with some leniency on the SOS-front for west coast teams. Rigid criteria doesn't allow for this, at least officially.

All that said, I do get that by allowing for a bigger dose of subjectivity, you'd become highly dependent on having committee members who can handle that responsibility and do a great job even though they have relatively little time to invest in it. And maybe that's another driver for a more data-driven process--it allows staff to crunch the numbers and requires less time on behalf of the actual decision makers.  So I'm probably living in a idealistic, dream world when I wish for more subjectivity--both on the side of the time the committee can spend on this and on the ability to get enough committee members who would do an outstanding job with the subjective decisions they'd be free to make.

[jknezek]

I just can't reward 7 wins in 18 games. Winning less than 50% of games played is my line in the sand. Same for 10-4-7, at least that's a bit more respectable.

Well 7-2-9 when breaking out the ties is 11-6-1 right? So if you look at it from that perspective it might shed more light. And the 7 wins is low but the 2 losses is impressive. The ties, these dreadful ties, will be the key factor of how they are viewed.

In terms of how it's counted as winning percentage... yes. In terms of what actually happened... no. They won 7 games, they lost 2. They didn't win 11 and lose 6. If we are accepting that it is better for the game for teams to have these records, instead of about half as many ties using the 2 OT system, then we can't just say they would have won 50% and lost 50%. That's simply not true. There needs to be a better way. Personally I don't think ties should count in winning percentage. Because it's called "winning" percentage. So if you are 7-2-9, your winning percentage is 7/18 or 38.88889%. We can add losing percentage also, and they are at 11.1111%. So if you want to look at those two numbers, I think that is actually better than what we do now. They are more likely to win then lose, but 50% of the time they can't do either.

I look at Williams, winning 6 games, and think that in order for them to win the National Title, they are going to have to almost double their total wins for the year, or go through a lot of PKs. And yet they are likely going to get a shot at it. I don't like it. If you can't flat out win more than 1/2 the times you step on the field, I'm just not real impressed with your resume.

I don't disagree with you. Just stating that a tie is, on paper, half a win and half a loss. Is this how the committee will view it? I don't know, but it's a possibility.

Only in this case. Usually a tie is 1/3 of a win. 3 pts for a win, 1 pt for a tie. I wouldn't mind seeing that implemented as a percentage of total points. That way Williams would be at 6-1-10 and would have 28pts out of a total 51 or 54.9%. Western Conn at 20-0-2 would be at 62/66 or 93.94% of available points won.

That seems much more accurate than William's faux "Winning Percentage" of 64.7% (11-6).

[BaboNation]

Just once I'd like to make it through an NCAA selection show without the video crashing. 

[Maine Soccer Fan]

Thanks.

And I'm guessing they didn't change their clocks yesterday? Oh, wait, the NCAA is in Indiana...

[Flying Weasel]

Pool C's
UW-Platteville
Pacific Lutheran
Gustavus Adolphus
Middlebury
Franklin and Marshall
Cortland State
Christopher Newport
Lynchburg
Catholic
Johns Hopkins
Williams
New York University
Ohio Northern
North Park
Case Western Reserve
Kenyon
Carnegie Mellon
Tufts
Bowdoin
Vassar

[paclassic89]

No Western Conn.  Kind of ridiculous in my opinion

[jknezek]

No Western Conn.  Kind of ridiculous in my opinion

Yep.

[stlawus]

Sorry but I find it ridiculous a team with 6 wins get a bid.  Negates the purpose of playing games. 

[PaulNewman]

First of all, you have got to be kidding me with that Kenyon draw.  And Stevens not much better.  More on that later.

By my count I got 20/20 prediction-wise but must have an error because I had CMU as a wild card and they're in....so 21? Or who did I have in that I missed? 

[Shooter McGavin]

Approaching this as if I am the committee and selecting the teams in order

Shooter's Pool C Picks
1. Hopkins
2. Kenyon
3. Gustavus
4. North Park
5. Newport
6. Midd
7. Montclair
8. Ohio Northern
9. Bowdoin
10. Cortland
11. Case Western
12. NYU
13. Williams
14. F&M
15. Lynchburg
16. UW-Platteville
17. Tufts
18. CMU
19. Vassar
20. Pacific Lutheran

First 4 Out
Catholic
Skidmore
Hamilton
Lycoming

19/20 isn't too bad 

[PaulNewman]

Oh wow.  Montclair?  Montclair didn't get in?  Wow, wow, wow.  That's my one miss but did have CMU as a wild card.

[Kuiper]

Oh wow.  Montclair?  Montclair didn't get in?  Wow, wow, wow.  That's my one miss but did have CMU as a wild card.

Didn't you have Montclair as a lock?  That seems like a big upset.

[PaulNewman]

Oh wow.  Montclair?  Montclair didn't get in?  Wow, wow, wow.  That's my one miss but did have CMU as a wild card.

Didn't you have Montclair as a lock?  That seems like a big upset.

Well, as a "lock to very likely."  There's no one on this board who wouldn't have picked Montclair. Not one.

[paclassic89]

I know paclassics has mentioned a number of times his desire for the committee to define/divulge how they weight the different criteria. Which feels like a call to make it more rigid and formulaic, more objective, removing the human element.  I disagree with that.  I wish it was less structured and less data driven and more flexible and subjective.

I think paclassics wants it to be more formulaic so that the selections are more predictable and transparent.  From the NCAA's standpoint, I think making it more rigid and formulaic allows the selections to be much more defensible and much less questionable than would be the case if the committee was given more flexibility and the process allowed for more subjectivity.  I guess I'm more comfortable with the "grey" that subjectivity would bring rather than striving for "black and white" clarity where the reality is complex and nuanced and not always accurately captured mathematically.

There's so many observations, comments, and arguments that could be made to support either side of this issue, but I'll just leave a couple that have me firmly on the side of desiring more subjectively and flexibility.

As it currently stands, for the purposes of carrying out ranking/selection duties, a committee member has no reason to watch a real game and formulate an opinion about a team.  In fact, an opinion formed from actually watching a team play (including their own team) doesn't seem like it could officially be used in the ranking/selection process.  In other words, numbers/data not only trump first-hand observation, they are the only thing that can be considered.  I'm not a fan of that.

The Massey ratings are 100% data driven and thus purely objective. I personally do not trust them. Overall it's not bad, but for me there are still way too many cases of a team having the numbers but not pasing the eye/smell test. If we followed Massey, wouldn't almost the whole NESCAC get into the tournament each year? A .500 or sub-.500 Colby should not be in the NCAA Tournament, IMO, but very well could be if the criteria becomes purely mathematical.  So, no, thank you.

In general, I want to see the best teams get selected with the at-large berths. Without allowing for subjectivity, I think it's harder to achieve that objective. And I would be good with some flexibility to select an undefeated Western Conn. even if the numbers don't quite make the cut. As I could be OK with some leniency on the SOS-front for west coast teams. Rigid criteria doesn't allow for this, at least officially.

All that said, I do get that by allowing for a bigger dose of subjectivity, you'd become highly dependent on having committee members who can handle that responsibility and do a great job even though they have relatively little time to invest in it. And maybe that's another driver for a more data-driven process--it allows staff to crunch the numbers and requires less time on behalf of the actual decision makers.  So I'm probably living in a idealistic, dream world when I wish for more subjectivity--both on the side of the time the committee can spend on this and on the ability to get enough committee members who would do an outstanding job with the subjective decisions they'd be free to make.

Very good post.  I should clarify that I don't think all subjectivity should be removed but I think it would beneficial to become more aligned with the D1 process.  Calculate an RPI throughout the year (it can be adjusted to weight SoS however the committee thinks it should be weighted) so there is an objective number to look at initially.  And then RvR, H2H, etc can be used as more subjective criteria for teams that are close in RPI.  Or use an adjusted RPI that bakes bonuses for ranked wins/losses into it.  I think the issue now is there doesn't seem to be consistency across the committees and there's a lack of transparency on how certain criteria are weighted.  Some regions seem to weigh SoS higher than others or RvR higher than others etc.   There is no such thing as a perfect ranking system but I do think this process can be improved.