Top 25 discussion

[scottiedawg]

It seems inconsistent that, when I say " x team's resume using Massey stuff looks super good, why aren't they ranked higher in D3hoops.com?" that you say (1) Massey and NCAA are different (implying I should ignore/discount/other the Massey data), (2) the top 25 has no bearing on NCAA selections.

If (2) is true (which I absolutely believe you it is), then why would any data (Massey, NCAA, other) be non-useful for the purposes of voting on a top 25 poll?

That's why I was curious as to what the D3hoops.com poll is trying to do.
If it's to determine the teams most likely to be in the NCAA tourney (which you're saying it's not), then Massey data should be ignored.
If it's to determine who the "best" teams are, then all data should be on the table. Right?

I've always interpreted polls as "the voters believe team 1 would have a higher win probability versus all other teams on a neutral court", "2 over all but 1", "3 over all but 1 and 2" etc etc.

[Pat Coleman]

With 20-plus at-large bids, it is now fairly unlikely that a Top 25 team gets left out of the NCAA Tournament in Division III basketball, although it does still happen, usually with teams in the 20-25 range.

Football is not really worth mentioning in any comparison, with just five at-large bids.

[VT-Alum-NOVA]

I checked only back to 15-16 season for WBB.  I specifically checked if teams ranked 15 to 25 in FINAL top 25 polls had made it into tournament.  All those teams made it into 64.  So it would seem the coaches do take into account whether a team in top 25 made the tournament.  To me, this makes sense.  To make final top 25, your team should have been selected to 64.

I guess converse would hold as well.  Your team makes final 4, your team WILL show in top 10 of the FINAL top 25 poll.

[Pat Coleman]

The final 25 takes place after the tournament. If you want to get the results I am referencing, you'll want to check the final regular season poll instead.

[Dave 'd-mac' McHugh]

Remember, the D3hoops.com Top 25 is made up of coaches, media members, SIDs, and others who know Division III basketball and balanced as evenly as possible across the entire country. It isn't just coaches. On the men's side, there is no men's poll. On the women's, there is the WBCA poll (for now) made up of eight "voters" who represent each "region."

[Roundball999]

Remember, the D3hoops.com Top 25 is made up of coaches, media members, SIDs, and others who know Division III basketball and balanced as evenly as possible across the entire country. It isn't just coaches. On the men's side, there is no men's poll. On the women's, there is the WBCA poll (for now) made up of eight "voters" who represent each "region."

Dave, is the WBCA poll going away?

[gordonmann]

Last year Augsburg and East Texas Baptist missed the NCAA Tournament but were ranked in our last regular season poll. This ranking was released the same day as the NCAA bracket was announced.

https://www.d3hoops.com/top25/women/2018-19/week12

Ithaca, Christopher Newport and Wisconsin Lutheran did better than our voters expected in the tournament so Augsburg and ETBU both dropped out of the final poll taken after the tournament was over.

[Pat Coleman]

Remember, the D3hoops.com Top 25 is made up of coaches, media members, SIDs, and others who know Division III basketball and balanced as evenly as possible across the entire country. It isn't just coaches. On the men's side, there is no men's poll. On the women's, there is the WBCA poll (for now) made up of eight "voters" who represent each "region."

Dave, is the WBCA poll going away?

Don't know that there is anything to say about that, but it seems like we will be moving to 10 regions in Division III basketball.

[deiscanton]

Top 25 poll is out for Week 8:

http://d3hoops.com/top25/women/2019-20/week8

No conversation needed with the voters this week.  Bethel moved up to the #5 ranking with 458 points.  Amherst and Wartburg dropped to #8 and #10 respectively, while Transylvania moved up to the #7 ranking.  Whitman moved up to the #9 ranking.

[VT-Alum-NOVA]

When will NCAA provide 1st regional rankings?

[PauldingLightUP]

When will NCAA provide 1st regional rankings?

February 12th. It seems late, but it falls right in line with previous years schedules with selection Monday not until March 2nd.

[scottiedawg]

Biggest vote gainers
+80, Bethel, beat Carleton (.353 WP, 118th NCAA SOS), beat Hamline (.688 WP, 232nd NCAA SOS)
+79, Whitman, beat Whitworth (.545 WP, 250 NCAA SOS), beat Linfield (.500 WP, 103 NCAA SOS)
+67, Trinity TX, beat Texas Lutheran (.375 WP, 252 NCAA SOS), beat Southwestern (.250 WP, 152 NCAA SOS)
+45, Chicago, beat NYU (.813 WP, 15 NCAA SOS), beat Brandeis (.750 WP, 43 NCAA SOS) WOW, great week!
+42, Transylvania, beat Mt. St. Joseph (.294 WP, 348 NCAA SOS), beat Anderson (.267 WP, 256 NCAA SOS)  Not sure how you look at Chicago's week and Transylvania's week and think they deserve the same number more votes. (granted, Chicago's % increase in votes is much higher)
+34, Oglethorpe, beat Millsaps (.450 WP, 265 NCAA SOS), beat Birmingham Southern (.650 WP, 266 NCAA SOS)
+33, UW-LaCrosse, beat UW-Stout (.647 WP, 87 NCAA SOS), beat UW-Eau Claire (.588 WP, 20 NCAA SOS)
+28, Mary Hardin-Baylor
+27, St. Thomas
+19, John Carroll
+18, DePauw
+10, Simpson

Biggest vote losers
-110, Amherst, beat Williams (.737 WP, 12 NCAA SOS), lost at Hamilton (.667 WP, 48 NCAA SOS)
-86, Baldwin Wallace, lost at John Carroll (.882 WP, 204 NCAA SOS), beat Marietta (.471 WP, 22 NCAA SOS)
-75, Loras, beat Coe (.588 WP, 86 NCAA SOS), lost to Luther (.563 WP, 9 NCAA SOS)
-66, UW-Whitewater
-66, Wartburg
-15, Misericordia
-14, Illinois Wesleyan
-11, Cortland
-11, Pacific
-10, Austin

Amherst lost 21% of their votes.
Bethel gained 21% of their votes.

I would definitely still argue that the optics of a loss drop teams a LOT in the poll. And that a large reason teams increase in votes is merely because others are dropping because of losses. Which absolutely means that winning twice against meh teams is treated better than winning once against a good team and losing once to a "good to meh" team on the road.

I feel like if Whitman and BW started the week with the same number of votes, and Whitman beat Linfield, and BW lost to John Carroll, that I wouldn't move them at all. Is that putting too much stock in "having merely played a good team?"

As a thought experiment, I bet most would say that losing to the #1 team is better than (i.e. you should gain votes in the polls relative to) beating the worst team.  At some point though, you'd switch. What's your line?  Off the top of my head losing to a top 100 team == beating a bottom 250 team. I bet everyone's would differ.

[RogK]

I'd say voters should have a feel for how much separation exists between teams in their own ranking. Any loss should reduce a voter's opinion of a team, although possibly not an immediate reduction in ranking (if their accomplishments this season are still regarded as superior to the teams below them). If another loss ensues, a reduction in ranking seems necessary.
Opportunity should exist for other teams to move up.
Absolutely zero credit should carry over from any previous season to a current season. That sort of "traditionally strong program" credit unfairly blocks other teams from moving up based on their current season accomplishments. Similarly, pre-season rankings are based on 0 games played, so they should not give one team a step up on any other team.
Here's a hypothetical that I wonder how voters would handle : fictional Canarsie College is 22-0 but their four best players just slipped on ice and each broke a wrist. If a voter votes based on how good teams are at the moment, Canarsie may be demoted (and they get thumped in their next three games). The NCAA would still rate them as any 22-3 team, ignoring their depleted ongoing talent, right?

[gordonmann]

True. They would still be 22-3 in the NCAA's eyes but less than that in mine when I vote. I've dropped teams down or off my ballot when they lose a really critical player for an extended period and then lose games.

You've touched on a couple variables that are important, but I don't know how (or if) they are handled by Massey:

1) Timing: Team A is 15-3 with three losses in the last couple weeks, all to teams in the Top 100. Team B is also 15-3 but it lost to three teams in the 100-200 range early in the season and is 15-0 since then. Which team should be ranked higher?  Team A because it has better losses? Team B because it has figured things out and is now playing better?  Now add in the fact that Team A's conference is better than Team B. Or take a player off Team A (injured but will be back next week) and add one to Team B (recovered but playing a gradually increasing number of minutes)

2) Location: I know the NCAA SOS weights away games differently than home games. But does it treat the away game that's played an hour away over holiday break differently than the away game that's played at the end of a full week of classes following a really long drive or flight?

To make these factors more tangible, last week I moved Chicago up my ballot from No. 18 to No. 13 despite the Maroons' loss to Case Western. Why would I move them up my ballot after losing to a team that's now .500?

Because...

- That's the second day of a Pittsburgh-Cleveland road swing and things happen on the back end of trips, even in the NBA.

- Chicago's overall resume, including the results played earlier in the year, looked better than the other teams around them who beat less impressive foes.

- CWRU had been playing much better as of late. After losing its first four games, the other two losses were to No. 3 Hope and ARV John Carroll.

One of the message board posters gently chided me for moving Chicago up and I reconsidered my position, especially after CWRU lost two more games this weekend, one of them by a non-competitive margin at Rochester.  I considered moving Chicago down but I still liked their overall resume better than other options. And they added two good wins against NYU and Brandies at home. And Chicago passes my "eye test" for whatever that's worth -- they have the physical attributes of teams that win in the NCAA Tournament. So I don't feel like their resume is inflated by playing in a weak conference or relying on hot shooting from small guards playing against inferior foes.

So I ended up leaving Chicago at No. 13 on my ballot and now Chicago is edging toward my ranking.

My take isn't necessarily right, but that's how I try to process the different variables.

[VT-Alum-NOVA]

any thoughts on d3hoops providing regional rankings as well before NCAA provides them?  maybe top 5 in each region?  would be a ton of work.