The NCAA Selection Committee does not target "best teams" for obvious reasons. They (a) want to get more of their own kind in (i.e. P5 -- it might be the most overt and obvious form of collusion in the United States), and (b) because there are so many undeserving teams that get in the tournament due to automatic bids coming from conferences that wouldn't have a chance in hell of getting in otherwise even if the tournament expanded to 128 they cringe at having to add anymore.

Here is a solution that rewards leagues for not being terrible as a whole, reduces the number bid stealers, reduces the number of unworthy conference champs and lessens the committees concerns with so many mid majors getting in the tournament (thereby - hopefully increasing the rationale to truly include the best teams regardless of what conference they come from).

Step 1. Set a minimum RPI number for what I call an automatic at-large. In other words, if you meet this RPI you are automatically in if you don't win a conference tournament. Set that at 30 or maybe 40...or whatever the right number is. In my mind, if you've done enough to get that high in the rankings there shouldn't even be a debate about it. The committee is allowed to use way too much subjective bull****. This reduces the bull****. Step 2 demonstrates how you are able to do this without it negatively impacting the selection process...

Step 2. The First Four is terrible in implementation but not necessarily concept. I would propose the following. Create a first eight that would match the 16 conference champions from the worst 16 conferences defined by conference RPI rankings. Instead of 16 auto bids going to conferences ranked 17 through 32 shrink that down to eight as they play each other for the right to move on to the field of 64. There is absolutely no damn reason that the SWAC or Southland, for example, should get auto bids. The Southland was 0-26 against Quadrant 1. The SWAC only won eight non-conference games all year and was 0-31 against Quadrant 1. The MEAC was 0-32. The bottom six conferences were 0-146 against Quadrant 1 teams. The NCAA Tournament has never been about putting the best teams in the dance.

This is the best way to get the best teams while also protecting conference champs.

Allow me to further illustrate why this is a better approach. In total, the bottom 16 conferences had a total of 13 Quadrant 1 wins between them. The West Coast Conf and C-USA (the two conferences with the biggest snubs) had 16 Quadrant 1 wins just between the two of them. Yet, quality conferences outside the non-privileged leagues are boxed in with the other 16 bottom feeder leagues.

Not only does this solution protect the traditional Cinderella from weak conferences it gives them a chance to win more than one game. It also frees up eight more at-large bids. Most lower conferences finish their conference tournaments a week early. The first eight could be played tonight (on the Monday) and Tuesday. Four games Monday. Four games Tuesday. If you were to do this you could either seed the First 8 with 1 to 16 seeds (initially) or just randomly generate the matchups. Let Penn go play Radford in a first round game for the chance to go play Kansas. They've already started doing this with two first four games for 16 seeds. They should do this for all the teams in the bottom 16 leagues.

And yes I realize this would have put C-USA in this place a year ago. That's life. We sucked as a league last year.

This year we would not have been, because we were better as a whole.

Basing it on RPI is part of the problem as it is. Its such a flawed metric. It served its purpose 25 years ago when no one knew better.

There are many examples, but the Bonnies come to mind. I think they are a top 25 RPI team and nowhere close to that level in real life.

(03-12-2018 12:02 PM)Gilesfan Wrote:  Basing it on RPI is part of the problem as it is. Its such a flawed metric. It served its purpose 25 years ago when no one knew better.

There are many examples, but the Bonnies come to mind. I think they are a top 25 RPI team and nowhere close to that level in real life.

I understand the flaws with the RPI but it's the only one that gives every team equal footing from the time the season starts.

The problem I have with others is that they are aggressively geared toward weighting what league you play in regardless of outcome instead of winning. That's why you almost never see teams from the non-power in the top 25 of those rankings.

Which metric would you use then? Maybe use an average of all of them. The human element of this process is failing because of power and greed.

I think it should be a pretty simple process that doesn’t need a committee. Take an average of the advanced metrics (RPI, KenPom, BPI, etc.) that the Committee supposedly uses. Choose the 36 highest ranked teams that did not get an auto-bid. Bam, you at least have something measurable to show as to why a team is in or out and are closer to getting the 36 best at-large teams regardless of conference affiliation.

(03-12-2018 12:23 PM)Hilltopper413 Wrote:  I think it should be a pretty simple process that doesn’t need a committee. Take an average of the advanced metrics (RPI, KenPom, BPI, etc.) that the Committee supposedly uses. Choose the 36 highest ranked teams that did not get an auto-bid. Bam, you at least have something measurable to show as to why a team is in or out and are closer to getting the 36 best at-large teams regardless of conference affiliation.

That would be better than what they have now. My opinion is KenPom is the best measure (and probably closest to Vegas lines).

I don't think any way is perfect. But like you said, it would at least be able to be quantified.

But MUH eye test!!

36 is the perfect number to use as a base number. If every automatic bid team fell outside the top 36 of whatever blended metric is used, that Top 36 makes up the rest of the field.

And I would take it two steps further.
1) Count the automatic bid teams in the top 36, then bring in the next half of that number from the ratings list as qualifiers. Using 2018 KenPom, there were 8 auto-bids in the top 36. So that would expand the group to 40 from the ratings list, 24 other auto-bids, and 4 at-large teams left for the committee to decide on.
2) For the committee at-large teams, figure three times the number of teams needed. For 2018 that number would be 12. Go to the ratings list and get the next 12 non auto-bid teams and start the discussion. Since there were 4 more auto-bid teams in KenPom #41-56, the 12 teams would have come from that #41-56 group.

(03-12-2018 08:45 PM)FriscoDawg Wrote:  36 is the perfect number to use as a base number. If every automatic bid team fell outside the top 36 of whatever blended metric is used, that Top 36 makes up the rest of the field.

And I would take it two steps further.
1) Count the automatic bid teams in the top 36, then bring in the next half of that number from the ratings list as qualifiers. Using 2018 KenPom, there were 8 auto-bids in the top 36. So that would expand the group to 40 from the ratings list, 24 other auto-bids, and 4 at-large teams left for the committee to decide on.
2) For the committee at-large teams, figure three times the number of teams needed. For 2018 that number would be 12. Go to the ratings list and get the next 12 non auto-bid teams and start the discussion. Since there were 4 more auto-bid teams in KenPom #41-56, the 12 teams would have come from that #41-56 group.

I like your methodology. (They’d never go for it BTW, but it is solid & waaaaay better than the current shaft).

What teams that missed out (besides MTSU) would get autobids this year?

I said this back when I first started posting on this forum back in '12....

unless someone files an antitrust suit, everything else is bs.....

they can do whatever the fokkkkkk they want.....

you can cry, you can whine, you can win..........it simply doesn't matter.....

until that happens the rest is just more bbs bs.....

disclaimer: that will never happen.....and why??? b/c you can't prove it in a court of law....

welcome to what it's like being a southern miss fan for 40+ yrs doing more with less and taking it up the arse many a time.....you'll get used to it....

posting drunk's posit in this one too.....

sooooooooo many threads covering the same topic.......how annoying.....

this is a dayum good idea where our conference currently sits if you want to avoid the 'bad loss' syndrome......and even then it may not matter.......however, it's better than any I've read to date....

Quote:ODUDrunkard13 Wrote:
Think it's time we change the conference tournament to a 4 team playoff, with the higher seeds hosting. Play on Wednesday and Saturday of championship week. We're just spread too wide to have a greatly attended conference tournament.

(03-12-2018 10:13 PM)owl at the moon Wrote:  

(03-12-2018 08:45 PM)FriscoDawg Wrote:  36 is the perfect number to use as a base number. If every automatic bid team fell outside the top 36 of whatever blended metric is used, that Top 36 makes up the rest of the field.

And I would take it two steps further.
1) Count the automatic bid teams in the top 36, then bring in the next half of that number from the ratings list as qualifiers. Using 2018 KenPom, there were 8 auto-bids in the top 36. So that would expand the group to 40 from the ratings list, 24 other auto-bids, and 4 at-large teams left for the committee to decide on.
2) For the committee at-large teams, figure three times the number of teams needed. For 2018 that number would be 12. Go to the ratings list and get the next 12 non auto-bid teams and start the discussion. Since there were 4 more auto-bid teams in KenPom #41-56, the 12 teams would have come from that #41-56 group.

I like your methodology. (They’d never go for it BTW, but it is solid & waaaaay better than the current shaft).

What teams that missed out (besides MTSU) would get autobids this year?

St. Mary's, Penn State, Notre Dame, Louisville, Baylor, and USC would have made the field by being in the KenPom top 40.

The group of 12 for discussion from which to pick the last 4 at-large bids would have been NC State, K-State, Arizona State, Maryland*, Oklahoma, UCLA, Rhode Island, Alabama, Middle Tennessee*, Marquette*, Syracuse, and Oklahoma State* (*teams not in actual field).

Teams that got an at-large but wouldn't have even been in the discussion under this system would have been Providence and St. Bonaventure.

Using a blended rating it would have been likely that Middle Tennessee would have been in, and Western Kentucky might have also been in the discussion.