View previous topic :: View next topic |
Author |
Message |
calbearman76
Joined: 02 Nov 2009 Posts: 5161 Location: Carson City
Back to top |
Posted: 08/22/18 5:51 pm ::: NCAA evaluation tool |
Reply |
|
The NCAA men's basketball selection committee has decided to replace the RPI with a new evaluation tool.
www.espn.com/mens-college-basketball/story/_/id/24445390/ncaa-announces-new-ranking-system-rpi
https://www.reuters.com/article/us-basketball-ncaa-net-rpi/ncaa-adopts-new-evaluation-tool-to-replace-rpi-idUSKCN1L71S8
https://www.ncaa.com/news/basketball-men/article/2018-08-22/college-basketball
Quote: |
NET will take into account game results, strength of schedule, game location, scoring margin, offensive and defensive efficiency and the quality of wins and losses.
|
Quote: |
Among the differences, late-season games won’t be given more weight in the rankings and margin of victory will be capped at 10 points to discourage teams from running up scores. |
Based on history I suspect the women will follow suit, possibly with a 1 or 2 year lag. As someone who has railed against the RPI for years I would like to think this will be an improvement but I am skeptical. In one sense anything is better than RPI, but this formula introduces several variables that may or may not be improvements. Capping a margin of victory at 10 points may be better than having no margin of victory included, but it may also wind up creating even more problems for mid-major teams. Not giving late-season games more weight seems like a move in the wrong direction.
The worst aspect of this announcement is that it was not accompanied by supporting data. What would the ratings have showed for the past 10 years. Who might have gotten in, or had a better seeding, if this system was in place? Hopefully this will be an opportunity to make real improvements in the selection process.
|
|
TheWildJacko
Joined: 06 Mar 2007 Posts: 301
Back to top |
Posted: 09/02/18 10:19 am ::: |
Reply |
|
10 points isn't a comfortable lead, so expecting teams to put on the brakes with a 10-point margin is unreasonable. In fact expecting teams to put in the brakes after a certain point as a blanket rule is unreasonable. If you have a large lead over the last team in the conference, you might put in your subs. But what if you're playing the game of your life and you have a large lead over UCONN? You have to keep the pressure on.
This rule seems designed to distort the actual game outcomes rather than reflect them. Either take complete margin of victory into account or ignore it completely.
|
|
ClayK
Joined: 11 Oct 2005 Posts: 11171
Back to top |
Posted: 09/02/18 11:36 am ::: |
Reply |
|
TheWildJacko wrote: |
10 points isn't a comfortable lead, so expecting teams to put on the brakes with a 10-point margin is unreasonable. In fact expecting teams to put in the brakes after a certain point as a blanket rule is unreasonable. If you have a large lead over the last team in the conference, you might put in your subs. But what if you're playing the game of your life and you have a large lead over UCONN? You have to keep the pressure on.
This rule seems designed to distort the actual game outcomes rather than reflect them. Either take complete margin of victory into account or ignore it completely. |
The other interesting thing from a statistical note is this: Margin of victory is the best way to find out which teams are better. Winning close games is luck; the good/great teams win games by large margins, and this arbitrary 10-point ceiling cuts off the evaluation tool from using a very valuable piece of information in its rating.
I would be OK with 25 points ... and let's face it, college coaches never take out their starters anyway, so what's the point of 10?
_________________ Oṃ Tāre Tuttāre Ture Svāhā
|
|
pilight
Joined: 23 Sep 2004 Posts: 67013 Location: Where the action is
Back to top |
Posted: 09/02/18 11:49 am ::: |
Reply |
|
A diminishing return on margin of victory seems reasonable. All the serious computer models (Massey, Sagarin, etc.) do that.
_________________ I'm a lonely frog
I ain't got a home
|
|
CBiebel
Joined: 23 Dec 2004 Posts: 1058 Location: PA
Back to top |
|
calbearman76
Joined: 02 Nov 2009 Posts: 5161 Location: Carson City
Back to top |
Posted: 09/02/18 7:04 pm ::: |
Reply |
|
I wouldn't really about this standard impacting coaching. Clearly a 10 point victory can happen as a result of a close game with fouling in the last minute or an 18 point lead whittled down in the final 2 minutes with reserves getting a chance to play. Making the cutoff higher could actually make coaches adjust to run up scores, which is not a good thing.
If you believe that winning matters (as I do) then the difference between a 1 point win and a 1 point loss is orders of magnitude greater than the difference between a 6 and an 8 point win. Even more, if you want to say that you are choosing the best teams based on what would essentially be solely a power rating you throw out the concept of competition.
I realize this isn't the NCAAs position, but I would like to have a rule that you have to have at least a .500 record in conference in order to make the tournament as an at-large team (and perhaps that you must finish in the top half of your conference.) I know this might eliminate from consideration some teams that may be marginally better, but I would rather see a team with 25 wins that hasn't had a chance to play top 50 teams than a team that has consistently shown they can't win.
|
|
|
|