Top 25 discussion

[Augie2020]

Does the NCAA committee use massey ?

[Augie2020]

Just like playing Pratt,Meredith,Goucher,Mary Baldwin .Scranton is stuck with Goucher nothing we can do there.Next year it will probably be worse with Lycoming and Wilkes joining.Cant do nothing when it comes to league play

[Augie2020]

Just to add to those notes... Transylvania was #2 in the preseason poll, is unbeaten, and yet isn't one of the four who have been #1 this year.
Started at #2 behind Hope, then fell to #3 after NYU moved ahead in week 1, then back to #2 when Hope lost in week 4, only to immediately get passed by Trinity (TX) in week 5, NYU fell in week 7 to get the Pioneers once again back to #2, then week 10 Trinity loses but Chris Newport moves ahead and they stay at #2.

I was going to do some homer talk in support of Transylvania, but I think I ended up coming to the conclusion that if I were a voter my #1 would be neither Transy nor CNU but the third unbeaten of Scranton.

great pick

[Augie2020]

Roundball:Not sure how massey works but when you go deep into the schedule CNU played one team under 40 and that was Santa Cruz #35,Transylvania played no one under 40 and just one team in the top 100 in massey  Scranton beat #13 Wartburg by 22 neutral court,#31 Tufts by 9 neutral court #34 Ithaca by 20 at home and #38 etown at home by 20.So I will never understand massey and how they work I just don't get how Scranton played all those teams and SOS is still thst high?

I don't pretend to fully understand Massey, but do know it's just math.  No such thing as an "eye-test" for Massey.  It doesn't cherry pick certain games or opponents, it considers all games and strength of all opponents.  I don't think it pays much (any?) attention to winning margins since that is too subjective and dependent on the coach; we all know that some of the strongest teams could have much higher average winning margins if they played their best players more minutes.  It does seem with Massey that if a team doesn't have a relatively strong SoS compared to other top contenders, it's very hard for that team to climb to the top of the rankings.

More explanation of SoS from Massey here:  https://masseyratings.com/theory/sched.htm

Thank you

[CNU85]

Roundball:Not sure how massey works but when you go deep into the schedule CNU played one team under 40 and that was Santa Cruz #35,Transylvania played no one under 40 and just one team in the top 100 in massey  Scranton beat #13 Wartburg by 22 neutral court,#31 Tufts by 9 neutral court #34 Ithaca by 20 at home and #38 etown at home by 20.So I will never understand massey and how they work I just don't get how Scranton played all those teams and SOS is still thst high?

I don't pretend to fully understand Massey, but do know it's just math.  No such thing as an "eye-test" for Massey.  It doesn't cherry pick certain games or opponents, it considers all games and strength of all opponents.  I don't think it pays much (any?) attention to winning margins since that is too subjective and dependent on the coach; we all know that some of the strongest teams could have much higher average winning margins if they played their best players more minutes.  It does seem with Massey that if a team doesn't have a relatively strong SoS compared to other top contenders, it's very hard for that team to climb to the top of the rankings.

More explanation of SoS from Massey here:  https://masseyratings.com/theory/sched.htm

Good comments. The above made me ponder and think for a moment, which made me look at the stats for CNU. Only two players average more than 20 minutes per game and that's Sondra Fan with only 20.9 minutes and Anaya Simmons with 21.9

[CNU85]

Just like playing Pratt,Meredith,Goucher,Mary Baldwin .Scranton is stuck with Goucher nothing we can do there.Next year it will probably be worse with Lycoming and Wilkes joining.Cant do nothing when it comes to league play

That's better than essentially not having a conference and trying to schedule anybody that will play you during their conference schedule!

[gordonmann]

Cant do nothing when it comes to league play

I think DeSales is having this translated into Latin and added to their team logo.

Look at their conference and then look at the Bulldogs' SOS. It's 111 positions -- almost one quarter of the Division! -- lower than Transylvania.

[Augie2020]

Just like playing Pratt,Meredith,Goucher,Mary Baldwin .Scranton is stuck with Goucher nothing we can do there.Next year it will probably be worse with Lycoming and Wilkes joining.Cant do nothing when it comes to league play

That's better than essentially not having a conference and trying to schedule anybody that will play you during their conference schedule!

True

[scottiedawg]

I find it very helpful to be clear what data and reasoning I'm using to make a statement.

It's very fun to talk about the top 25 polls, in large part because voters aren't instructed what data to use and what data not to use. We (referring to not just voters, but all fans) all have different opinions and viewpoints, and discussing those is a lot of fun.

Massey kinda tries to do what the top 25 polls do, but from a computer standpoint (which is more objective, but less fun). It has made decisions on how to value teams (e.g. using margin of victory as one data point).  While probably useful, we can see why the NCAA doesn't want to incentivize running the score up.

I have a model that does kinda what Massey does, but I've added my own spin on "metrics I think are important."

The NCAA has made it's own decisions on how to "rank" teams. They have largely decided on Objective measures (e.g. no eye test, no consideration for injuries, mid year transfers, no consideration for "peaking at the end of the year"). While those are indeed useful pieces of data for determining the best current teams, they are probably infeasible for the NCAA to set out as part of the criteria for ranking 430 teams.

The NCAA SOS indeed has it's cons. As Augie has pointed out, you can't control the league you play in. Some leagues have a bottom half or third with terrible winning  percentages, that really hurt the top teams in the league. That comes with those games being easy to win. 

WIAC or UAA teams automatically get awesome SOS by virtue of existing in the WIAC or UAA, but also have a harder time constructing a strong Win% than teams from weak conferences.

The NCAA Primary Criteria are seeking to, as objectively and fairly as possible, rank teams coming from many different "situations" (e.g. weak or strong conference, geography, travel budget) that those teams cannot control.

At the end of the day it's very very difficult to fairly and consistently rank a 900 WP / 500 SOS vs a 700 WP / 600 SOS, regardless of the objective or subjective criteria you use.

(as an aside, the overall winning % of teams in the Landmark is the 3rd highest of any conference.  Obviously if you're the best team you're the biggest reason for that high cumulative win% and you can't face yourself, but I digress).

When it comes to Scranton and Christopher Newport, examining only objective data (let's just use the NCAA Primary Criteria), they are soooo close.  Both have an undefeated Winning%.  Scranton's SOS is 0.577. CNU's is 0.554.  CNU is only barely ahead for me (again using only the NCAA Primary Criteria) because they have likely 8-9 wins vRRO, while Scranton likely has 7.

Even there you see shades of gray, not all RROs are created equal. It's easier to be a ranked team in a weak region than in a strong region. (but I digress once again).

Massey SOS, NCAA SOS, and NCAA wins vRRO are all trying to measure "did you play and beat good teams or not"

CNU and Scranton are very close among all 3. Transylvania is clearly behind in all 3.

I think CNU and Scranton are the only two you can make a "computer" argument for #1.    I think a "human" argument can be made for a whole handful of teams to be the current #1. And I don't think "computer" is necessarily better than "human."

Keep discussing!

[Augie2020]

I find it very helpful to be clear what data and reasoning I'm using to make a statement.

It's very fun to talk about the top 25 polls, in large part because voters aren't instructed what data to use and what data not to use. We (referring to not just voters, but all fans) all have different opinions and viewpoints, and discussing those is a lot of fun.

Massey kinda tries to do what the top 25 polls do, but from a computer standpoint (which is more objective, but less fun). It has made decisions on how to value teams (e.g. using margin of victory as one data point).  While probably useful, we can see why the NCAA doesn't want to incentivize running the score up.

I have a model that does kinda what Massey does, but I've added my own spin on "metrics I think are important."

The NCAA has made it's own decisions on how to "rank" teams. They have largely decided on Objective measures (e.g. no eye test, no consideration for injuries, mid year transfers, no consideration for "peaking at the end of the year"). While those are indeed useful pieces of data for determining the best current teams, they are probably infeasible for the NCAA to set out as part of the criteria for ranking 430 teams.

The NCAA SOS indeed has it's cons. As Augie has pointed out, you can't control the league you play in. Some leagues have a bottom half or third with terrible winning  percentages, that really hurt the top teams in the league. That comes with those games being easy to win. 

WIAC or UAA teams automatically get awesome SOS by virtue of existing in the WIAC or UAA, but also have a harder time constructing a strong Win% than teams from weak conferences.

The NCAA Primary Criteria are seeking to, as objectively and fairly as possible, rank teams coming from many different "situations" (e.g. weak or strong conference, geography, travel budget) that those teams cannot control.

At the end of the day it's very very difficult to fairly and consistently rank a 900 WP / 500 SOS vs a 700 WP / 600 SOS, regardless of the objective or subjective criteria you use.

(as an aside, the overall winning % of teams in the Landmark is the 3rd highest of any conference.  Obviously if you're the best team you're the biggest reason for that high cumulative win% and you can't face yourself, but I digress).

When it comes to Scranton and Christopher Newport, examining only objective data (let's just use the NCAA Primary Criteria), they are soooo close.  Both have an undefeated Winning%.  Scranton's SOS is 0.577. CNU's is 0.554.  CNU is only barely ahead for me (again using only the NCAA Primary Criteria) because they have likely 8-9 wins vRRO, while Scranton likely has 7.

Even there you see shades of gray, not all RROs are created equal. It's easier to be a ranked team in a weak region than in a strong region. (but I digress once again).

Massey SOS, NCAA SOS, and NCAA wins vRRO are all trying to measure "did you play and beat good teams or not"

CNU and Scranton are very close among all 3. Transylvania is clearly behind in all 3.

I think CNU and Scranton are the only two you can make a "computer" argument for #1.    I think a "human" argument can be made for a whole handful of teams to be the current #1. And I don't think "computer" is necessarily better than "human."

Keep discussing!

Thanks

[VT-Alum-NOVA]

I hope committee puts Scranton & CNU on opposite sides of bracket this year.  Still remember CNU traveling up to Scranton for 2019 sweet 16 and losing up to them by missing 2 free throws as time expired.

[scottiedawg]

I could absolutely be wrong, but I expect to see most of the top overall seeds on the same side of the bracket, due to geography.

If you build a quadrant around Scranton, but put it on the same side of the bracket as "west" quadrant (e.g. WIAC, Chicago, Hope), you set yourself up for Sectional weekend flights.

[scottiedawg]

This obviously can and will change a lot (the order) but due to geography I expect multiple (or all) of these teams to be in the same quadrant (CNU, Scranton, NYU, Messiah, Elizabethtown)


   
       
       
       
       
       
       
       
       
       
       
       
       
       
       
       
       
       
       
       
       
       

Team	My Guess at Overall Seed	Distance from CNU
Smith	1	556
Christopher Newport	2	0
Scranton	3	405
Trinity (Conn)	4	516
Babson	5	603
Ithaca	6	517
Hope	7	840
Trinity (Texas)	8	1581
NYU	9	398
Hardin-Simmons	10	1526
Chicago	11	856
Ohio Northern	12	620
Messiah	13	274
Elizabethtown	14	291
Baldwin Wallace	15	532
Rhode Island College	16	581
SUNY New Paltz	17	475
UW-Whitewater	18	964
WashU	19	889
Tufts	20	619

If that does happen, at least it will likely be structured so a potential CNU-Scranton matchup won't occur until the Elite 8.

[Augie2020]

Usually they do a great job separating the teams from the same region!For example look at last year's set up

[scottiedawg]

Usually they do a great job separating the teams from the same region!For example look at last year's set up

Agreed. There's likely scenarios I'm not considering.

Thinking in broad strokes, I think you could create a quadrant with New York, PA, OH, KY teams.    And then a "west" quadrant with TX, MI, IL, WI, IA, IN.   The tough thing is I think there are more than 16 "west" teams, that can't quite get to OH or PA, so you can't get 1st 2 rounds to work geographically.   So could be that being able to split Scranton and CNU up (which makes sense overall seeding wise), hinges on the # of Pool C teams from: (MIAA, WIAC, MIAC, ARC).

If you need a 5th "west" opening round host, maybe you put WashU in that NY/PA/OH/KY quadrant. But that means Scranton would not be able to host if chalk gets through to the Sectionals, OH team would likely host.

If you're Scranton you probably would rather avoid CNU in the Elite 8, even if it means you can't host Sectionals?