The Mad Scientist Top 25 Ranking Debate Topic

zhawks
Posts: 881 (0)
Hall of Famer
208 66 38
Pro
Quote: Originally Posted By seble on 12/21/2009We're not trying to replicate the inputs that the selection committee uses, but to replicate the output of that committee. Based on our analysis, no matter what they claim to look at, those three things pretty closely determine seeding
I can understand that, while I agree with dalt in the fact that the RL committee doesn't use it I get that you are trying to mimic their final outcome, which might need a little different data input then they use IRL.

dalt, while I agree that the T25 has tried to be fixed (and failed) numerous times, seble has not tried to fix it yet so let's see what he has for us.
12/21/2009 2:22 PM
seble
Posts: 880 (0)
Hall of Famer
24 44 35 306 121 185
Prospect
The new rankings will be based on a combination of ratings and team performance, where ratings will be weighted heavily early in the season but gradually team performance will become the more important component.

Team performance will be a combo of Win% and SOS.
12/21/2009 2:23 PM
zhawks
Posts: 881 (0)
Hall of Famer
208 66 38
Pro
So it is not directly related to rpi as you mentioned above, it just uses many of the same factors?
12/21/2009 2:25 PM
colonels19
Posts: 4399 (0)
Hall of Famer
78 90 3 1 168 2 1 11
Prospect
I think WIS should use my BPI Sports CBB ranking format, but it probably makes too much sense and they don't want to pay me for its usage...too bad...
12/21/2009 2:34 PM
arssanguinus
Posts: 1120 (0)
Hall of Famer
282 1
Pro
What sort of things on team performance? Performance against high rated teams? Road performance? Margin of victory?

For example, under the current one, I am trying to figure out exactly HOW I got a #7 ranking which is rather higher than I would have put myself.
12/21/2009 2:39 PM
zhawks
Posts: 881 (0)
Hall of Famer
208 66 38
Pro
Quote: Originally Posted By arssanguinus on 12/21/2009What sort of things on team performance? Performance against high rated teams? Road performance? Margin of victory?

For example, under the current one, I am trying to figure out exactly HOW I got a #7 ranking which is rather higher than I would have put myself.


Quote: Originally Posted By seble on 12/21/2009
The new rankings will be based on a combination of ratings and team performance, where ratings will be weighted heavily early in the season but gradually team performance will become the more important component.

Team performance will be a combo of Win% and SOS.
12/21/2009 2:42 PM
arssanguinus
Posts: 1120 (0)
Hall of Famer
282 1
Pro
Oh. THanks. Just totally missed over that reading. (Facepalm)
12/21/2009 2:44 PM
zhawks
Posts: 881 (0)
Hall of Famer
208 66 38
Pro
No problem, just wanted to point it out incase you had missed it.
12/21/2009 2:58 PM
seble
Posts: 880 (0)
Hall of Famer
24 44 35 306 121 185
Prospect
We'll have to see how the rankings system works during testing. While it's tempting to want to factor in a million things, that can lead to a bunch of components counteracting each other or washing out. Basically a team is as good as it's record and the strength of it's opponents.
12/21/2009 3:01 PM
seble
Posts: 880 (0)
Hall of Famer
24 44 35 306 121 185
Prospect
By considering ratings, especially early in the season, it should temper some of the wild swings from lack of data.

My biggest concern is the problem of teams moving drastically early in the season. It's not necessarily a big deal in the long run as it settles down at the end of the season. But it can look funny. The flipside to that is that you don't want a team to hang around too long just based on ratings/talent if they're not performing.
12/21/2009 3:04 PM
vandydave
Posts: 2124 (0)
Hall of Famer
28 1 1 843 273
Prospect
so how will these changes impact wis tournament seedings the most as compared to the previous set-up?
12/21/2009 3:04 PM
zhawks
Posts: 881 (0)
Hall of Famer
208 66 38
Pro
Quote: Originally Posted By seble on 12/21/2009By considering ratings, especially early in the season, it should temper some of the wild swings from lack of data.

My biggest concern is the problem of teams moving drastically early in the season. It's not necessarily a big deal in the long run as it settles down at the end of the season. But it can look funny. The flipside to that is that you don't want a team to hang around too long just based on ratings/talent if they're not performing
But even IRL doesn't a team jump from NR to 15 after a good 2-0 start against good teams?
12/21/2009 3:09 PM
seble
Posts: 880 (0)
Hall of Famer
24 44 35 306 121 185
Prospect
That's not a big deal, but what if the #1 team loses the first game and drops from the rankings?
12/21/2009 3:13 PM
zhawks
Posts: 881 (0)
Hall of Famer
208 66 38
Pro
Quote: Originally Posted By seble on 12/21/2009That's not a big deal, but what if the #1 team loses the first game and drops from the rankings
Agreed, I didn't realize that is what you were going for. That shouldn't happen IMO. How close are the top 25 teams in terms of player ratings usually? Is it still using only the top 8 players overall ratings as before? Is it wise to use a players overall for this, or is it (should it?) be a formula to judge them?
12/21/2009 3:22 PM
dalter
Posts: 71 (0)
All Star
25
Prospect
Quote: Originally Posted By seble on 12/21/2009The new rankings will be based on a combination of ratings and team performance, where ratings will be weighted heavily early in the season but gradually team performance will become the more important component.

Team performance will be a combo of Win% and SOS
seble, by NT selection time, ratings should not factor into this whatsoever. At that point, it's the resume you've built over the course of the season, not how many 99 rebounders you have.

Do you agree with that premise?
12/21/2009 3:23 PM
◂ Prev 1...8|9|10|11|12...75 Next ▸
The Mad Scientist Top 25 Ranking Debate Topic
△Top△Bottom

Search Criteria

Terms of Use Customer Support Privacy Statement

© 1999-2026 WhatIfSports.com, Inc. All rights reserved. WhatIfSports is a trademark of WhatIfSports.com, Inc. SimLeague, SimMatchup and iSimNow are trademarks or registered trademarks of Electronic Arts, Inc. Used under license. The names of actual companies and products mentioned herein may be the trademarks of their respective owners.