Page 2 of 31 FirstFirst 123412 ... LastLast
Results 16 to 30 of 460
  1. #16

    Default

    Quote Originally Posted by Unregistered User View Post
    Great. The top 10 teams will perform like top 10 teams. It may adversely affect teams closer to the bubble.

    Nate Silver had it right from the jump.
    Silver's analysis was silly because he was taking the NET as it appeared very early in the season, before the data started to connect. The RPI was also very famously noisy early in the season. Over time, as more games are played, the numbers start to settle down and the teams start to sort out as you'd expect.

    Note that the one part of his analysis that he got right was that the efficiency metrics should factor MORE heavily.

    NET is imperfect, but it's MUCH better than RPI. It's biggest flaw now is still including some metrics that are similar to RPI.


  2. #17

    Default

    Quote Originally Posted by ltf View Post
    If anyone wants to take the time, interesting article about NET at sports illustrated. Its rather long but the crux is that, while the NET formula hasnt been shared,
    NET seems to empasize
    - efficiency
    - final score margin
    - road wins
    NET seems to de-empasize
    - strength of schedule
    The article gives examples to support its conclusions and speculates that power 5 teams will start to schedule low major teams on the road as this gives the best potential for a high NET.
    This is not really true. The most heavily-weighted component is still game results based on opponent and location, which is heavily dependent on strength of schedule.

    The 2nd least heavily-weighted factor is adjusted winning percentage which takes into account road games. So if you want to game the system by focusing on the 2nd least-heavily weighted factor, you can start playing road games against bad teams, but you'd be focusing on something that contributes probably 10% of your NET score. That would be really dumb.


    Last edited by justthefacts; 01-24-2019 at 07:59 AM.

  3. #18

    Default

    Quote Originally Posted by justthefacts View Post
    This is not really true. The most heavily-weighted component is still game results based on opponent and location, which is heavily dependent on strength of schedule.

    The 2nd least heavily-weighted factor is adjusted winning percentage which takes into account road games. So if you want to game the system by focusing on the 2nd least-heavily weighted factor, you can start playing road games against bad teams, but you'd be focusing on something that contributes probably 10% of your NET score. That would be really dumb.


    Liberty has a kenpom sos of 337 and yet they are ranked ahead of us. There are other mid and lows with awful sos also ranked ahead of us. Doesnt seem like sos is valued much.

  4. #19
    Join Date
    Oct 2012
    Location
    Fargo
    Posts
    844

    Default

    I bet next year we won't see many quality non conference opponents outside of what is already on the schedule (@Utah, OK St in Tulsa, ACC/Big Ten opponent at home, Gavitt games?) We will be scheduling teams we can crush.

  5. #20

    Default

    Quote Originally Posted by tmvander View Post
    I bet next year we won't see many quality non conference opponents outside of what is already on the schedule (@Utah, OK St in Tulsa, ACC/Big Ten opponent at home, Gavitt games?) We will be scheduling teams we can crush.

    We as a P5 school may not want that as we want other P5 schools on our NC schedule but that is what the majority of schools have begged for. Their biggest complaint over the last two decades is that the 'big boys' wont play lower level teams on the road. Doesn't this new system encourage that?

    I really don't know much about the NET Rankings other than what I've read on this site but if what everyone hear is saying is correct it seems like this system would not reward the Duke's of the world that play neutral site or home NC games. Isn't that a good thing for the game?

  6. #21

    Default

    Best coaches and programs still seek out great games as they believe you get better by playing great teams. The research reveals that to become a top program you schedule hard, and that does not mean middle or bottom tier power 5 teams. The problem is the games are hard to get. Trying to game the schedule to make the tourney is cowardly. As a fan and ticket holder at several venues i see high end power games in the non con and would love to see ISU and Marquette in home and home here. I do think the Gophers tried to schedule very good teams and could not help it that several of those schools that are usually very good turned out to be very good. It also hurt to get clobbered in the ACC game at Boston College.

  7. #22

    Default

    Quote Originally Posted by Powder View Post
    We as a P5 school may not want that as we want other P5 schools on our NC schedule but that is what the majority of schools have begged for. Their biggest complaint over the last two decades is that the 'big boys' wont play lower level teams on the road. Doesn't this new system encourage that?

    I really don't know much about the NET Rankings other than what I've read on this site but if what everyone hear is saying is correct it seems like this system would not reward the Duke's of the world that play neutral site or home NC games. Isn't that a good thing for the game?
    Yes, the SI article agrees with you that an unintended effect might be that pwer 5 schiols will start playing some mids and lowers on the road, which might be kimd of a good thing for a change. I suppose a counterargument would be that schools' fans may not be too happy about losing home games. Also, how much money would be lost through sacrificing home games. Still seems like needs to be tweaked to value sos more.
    Last edited by ltf; 01-24-2019 at 09:46 AM.

  8. #23

    Default

    I'm still in wait-and-see mode on the NET rankings. The biggest factor we still don't know is how the committee will use them. Do they get more weight than the other rankings (KenPom/Sagarin/etc.) in the committee's eyes? Or are they seen as equal to other rankings by the committee? Also, what does NET look like with a full season of data? Still too many unknowns to for me form a strong opinion.

    Every system is going to have its outliers and flaws. NET is no different. But I know I like NET better than RPI.

  9. #24

    Default

    Quote Originally Posted by tmvander View Post
    I bet next year we won't see many quality non conference opponents outside of what is already on the schedule (@Utah, OK St in Tulsa, ACC/Big Ten opponent at home, Gavitt games?) We will be scheduling teams we can crush.
    Bear in mind that the NCAA will still use NET to analyze wins against quadrant. Being 10th in NET (for example) but having only 4 Q1&2 wins (for example) will not play well on Selection Sunday. This is also the reason teams won't suddenly start scheduling Campbell on the road.
    Last edited by justthefacts; 01-24-2019 at 10:53 AM.

  10. #25

    Default

    Quote Originally Posted by bizzle22 View Post
    I'm still in wait-and-see mode on the NET rankings. The biggest factor we still don't know is how the committee will use them. Do they get more weight than the other rankings (KenPom/Sagarin/etc.) in the committee's eyes? Or are they seen as equal to other rankings by the committee? Also, what does NET look like with a full season of data? Still too many unknowns to for me form a strong opinion.

    Every system is going to have its outliers and flaws. NET is no different. But I know I like NET better than RPI.
    People should take a wait-and-see approach because the data is still sorting itself out. Lots of systems incorporate last year's results along with incoming recruiting and it doesn't really leave the system until the last 10 games or so. NET (for obvious reasons) doesn't have that luxury and so it's very noisy to start and gets gradually less noisy over time.

    Here's a good explanation of how NET will be used. I think the implication is that Sagarin, KenPom, and BPI will still be used, but I'd imagine it will be used only as a corner-case tiebreaker.

    This marks the second consecutive year the committee has made a significant change. Before last season, a quadrant system was adopted to place greater emphasis on success in games played away from home on the team sheets, which offer a snapshot of each team’s schedule and results. The existing quadrant system still will be used on team sheets, with the NET replacing the Rating Percentage Index to sort games based on the opponent’s ranking:

    Quadrant 1: Home 1-30, Neutral 1-50, Away 1-75
    Quadrant 2: Home 31-75, Neutral 51-100, Away 76-135
    Quadrant 3: Home 76-160, Neutral 101-200, Away 135-240
    Quadrant 4: Home 161-353, Neutral 201-353, Away 241-353

    While the quadrant system was widely deemed an improvement to the selection process, the NET is another significant step in addressing the recommendations the NCAA received from the NABC’s ad hoc committee, whose purpose was to make recommendations regarding the selection, seeding and bracketing of teams.

    Another change made last year to the team sheets was the inclusion of other metrics. These include the Kevin Pauga Index and ESPN’s results-oriented metric, the Strength of Record. The team sheets also included three predictive metrics: those managed by renowned basketball analytics experts Ken Pomeroy and Jeff Sagarin, as well as ESPN’s Basketball Power Index.
    https://www.ncaa.com/news/basketball...etball-ranking

  11. #26
    Join Date
    Nov 2008
    Location
    Shoreview
    Posts
    14,566

    Default

    The way I see it, the NET ratings are a politically correct way to consider margin of victory without technically weighting margin of victory. The various advanced metrics are a reliable surrogate for a score-based rating system. That's why KenPom and the like have always correlated well with Sagarin and the like.

  12. #27

    Default

    Quote Originally Posted by Bad Gopher View Post
    The way I see it, the NET ratings are a politically correct way to consider margin of victory without technically weighting margin of victory. The various advanced metrics are a reliable surrogate for a score-based rating system. That's why KenPom and the like have always correlated well with Sagarin and the like.
    It doesn't show up in the embedded tweet, but if you click the tweet you'll see that scoring margin is explicitly a factor in the ranking. It's the least heavily weighted, but it's an explicit factor.

  13. #28

    Default

    I did a comparison of all 300+ teams and compared NCAA Net to Ken Pom and the correlation is positive 0.986.

    I'd embed the tweet with the graph but don't know how, so he's a link.


    https://twitter.com/FaceTheFactsMn/s...00954498281472
    "Do Not Be Afraid to Be A Legend"

  14. #29

    Default

    Teams in the top 75 who are benefiting from NET over their KenPom are:

    NET KP rank School
    7 21 Houston
    13 24 LSU
    19 33 Marquette
    31 42 Kansas St.
    35 45 UCF
    39 51 St. John's (NY)
    47 63 Hofstra
    53 93 UNC Greensboro
    54 75 Temple
    62 90 Furman
    67 88 Memphis
    68 80 Yale
    69 84 Davidson
    75 107 Saint Louis


    Avg Strength of schedule is 156.

    On the other hand, teams most penalized by NET from KenPom are:

    KP rank NET School Difference SOS
    15 27 Auburn -12 28
    26 37 Florida -11 13
    31 45 Ohio St. -14 26
    35 48 Saint Mary's (CA) -13 108
    48 58 Creighton -10 7
    54 71 Oregon -17 68
    60 70 Arkansas -10 44
    62 74 New Mexico St. -12 174


    Average SOS of 58.5
    That average goes down to 42 if you toss out the last entry (NMSU).

    254 of the 353 teams are within 15 spots of the two rankings.
    138 are within 5 spots either way.
    "Do Not Be Afraid to Be A Legend"

  15. #30
    Join Date
    Nov 2008
    Location
    Shoreview
    Posts
    14,566

    Default

    Quote Originally Posted by justthefacts View Post
    It doesn't show up in the embedded tweet, but if you click the tweet you'll see that scoring margin is explicitly a factor in the ranking. It's the least heavily weighted, but it's an explicit factor.
    Got it - good to know; thanks!

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •