Preliminary Release of NCAA Tournament Seeds

In a bizarre attempt to match the stupidity (or maybe futility) of releasing preliminary football playoff rankings, the NCAAT Selection Committee released a preliminary version of the top 16 teams (top 4 seeds) if the March Madness selection process were held today. The only way this could get any uglier, would be to release the entire field. With releasing the Top 4 seeds, people who care about the Big 10(or 14) will bitch about not having a team in the top four seeds. I expect this whining to blow over fairly quickly. But can you imagine the hand-wringing if a preliminary shot at the entire field were released?

In the past, I have made two separate attempts at playing around with seeding. The first attempt was semi-successful. I matched the NCAAT seeds with RPI Rankings and found a strong, linear correlation through the Top 6/7 seeds (ie top 28 teams). Sometime later I read in Jerry Palm’s old blog (collegerpi.com) that about 75% of the field is seeded within one seed of what you calculate based solely on the team’s RPI ranking. So best case, I showed that Palm’s observation was still valid.  Worst case, I wasted a bunch of time and accomplished nothing.   (At least reading Palm’s blog saved me the time it would have taken to update the correlation each year.)

The second time was in 2006 when State had an RPI ranking in the Top25 and was ranked in the AP Top 25. I was searching the internet to see where people where seeding State and eventually blundered across a website that collated various bracket predictions. I did weekly blog entries and we all discussed current seeds and how the team could improve its seeding. Then this happened:

Sad note…since 2006 when “NC State” and “NCAA Tournament” could be found in the same paragraph, I was focused only what it would take to clear the bubble and not seeding. (It’s not like I’m bitter or anything.)

Back to the present…

I thought that I would compile a table and look at the top 16 teams and the various stats that we look at on a weekly basis for the ACC teams:

I didn’t spend much time with the data, but I highlighted a few things that jumped out to me:

The Selection Committee picked two teams outside of the RPI Top 16 to fill their preliminary Top 4 seeds. Clearly, they didn’t use Pomeroy’s rankings.

We’ve seen poor OOC schedules send Bubble Teams to the NIT. It looks like the Selection Committee is not very consistent with their penalties for poor SOS rankings. The two teams that they jumped up into the Top16 have truly horrible OOC schedules (and their conf SOS ranking is lower than the worst team in the ACC).

I’ve heard some hype for UCLA, but haven’t seen them play. But looking at these numbers, I’m not impressed. I didn’t take the time to include conference standings, but UCLA is tied for third place in the 6th ranked conference. Once again, I’m not impressed.

I was surprised to see the L’ville has a losing record against the RPI Top 50. Obviously, that didn’t matter much to the Selection Committee

OK, this was just a quick and dirty look. What do you notice, hear, or read about this preliminary seeding?

About VaWolf82

Engineer living in Central Va. and senior curmudgeon amongst SFN authors One wife, two kids, one dog, four vehicles on insurance, and four phones on cell plan...looking forward to empty nest status. Graduated 1982

16-17 Basketball College Basketball

Home Forums Preliminary Release of NCAA Seeds

Viewing 10 posts - 1 through 10 (of 10 total)
  • Author
    Posts
  • #117227
    VaWolf82
    Keymaster

    In a bizarre attempt to match the stupidity (or maybe futility) of releasing preliminary football playoff rankings, the NCAAT Selection Committee released a preliminary version of the top 16 teams (top 4 seeds) if the March Madness selection process were held today.
    [See the full post at: Preliminary Release of NCAA Seeds]

    #117229
    McCallum
    Participant

    Did State make the Central Tar Heel 1A tourney?

    McCallum

    #117233
    Wulfpack
    Participant

    What stands out to me is Gonzaga, FSU and Butler.

    #117249
    VaWolf82
    Keymaster

    What stands out about those three?

    #117251
    BJD95
    Keymaster

    Artificially inflating the west coast teams. Again. The first half of UCLA/OR sure didn’t impress me at all.

    #117252
    VaWolf82
    Keymaster

    At one time, the Dance Card had a P12 factor to account for some strange decisions made by the selection committee. They discussed this briefly in the paper they published (linked in my Dance Card entry). Previously, they had concluded that the selection committee was biased against mid-majors.

    But they say in the current paper that they didn’t find statiscal support for conference bias since ~2012. But I’m pretty sure UCLA got in just a couple of years ago for no obvious reasons.

    #117254
    BJD95
    Keymaster

    ^ It just sounds like something Jim Nantz would say, doesn’t it – “The NCAAT is just better with UCLA in it.” BARF

    #117262
    Wulfpack
    Participant

    What stands out about those three?

    Just that they are so highly regarded.

    #117266
    VaWolf82
    Keymaster

    Of those 3, I have only seen FSU play. They’re seeded according to the RPI rank and they are 8-2 against the Top 50. Their struggles on the road against lesser teams make me doubt that they will go very far in a tournament format.

    Butler is placed two seeds below what their RPI rank would predict. So it doesn’t look like the committee is super-impressed either. They likely have little/no margin for error.

    Gonzaga gets the socialistic #1 seed to match their undefeated record. At least they backed up the committee’s confidence in them with a road-victory yesterday against St Mary’s..the toughest game left on their regular season schedule. So if they win out, they’ll have a #1 seed. The interesting question would be where do they drop if they stumble?

    #117267
    BJD95
    Keymaster

    Think they’d drop Gonzaga all the way down to a 4. 3 at bare minimum.

Viewing 10 posts - 1 through 10 (of 10 total)
  • You must be logged in to reply to this topic.