OT: Using a Points System to Seed the NCAA tournament
***Fixed Little Rock Duplicate**
This is my 1st Thread so be gentle.
I personally hate the ambiguity of having a panel debate who should get in and who shouldn’t using arbitrary arguments.For example, Michigan was 4-11 vs. Top 100 teams. This totally ignores that most of those loses were vs Top 25 teams! Should we really be punished for losing to a great team?
So I came up with a theory.
Why don’t we use a Point System (like the NHL) for wins and losses…how would that look?
Basics (Using RPI)
- Win vs. Top 25 = +6 points
- Loss vs. Top 25 = -1 points
- Win vs. 26-100 = +3 points
- Loss vs. 26-100 = -3 points
- Win vs 101+ = +1 point
- Loss vs. 101+ = - 6 points
**Reward Great Wins and Punish Horrible Losses
I also wanted to reward winning your tournament, so if you win your conference tournament, you are bumped up 1 seed.
So how does this look? (I’m using images in hopes that my formatting isn’t horrible)
- Green = Conference Tournament Champ & 1 Seed Bump
- Red = Downward Seed Bump (due to Green Bump)
I personally love this, but I’m someone that likes math to solve problems vs. gut feelings.
Also, since I’m God in this scenario, I’ll add a bonus rule:
Each team selects what region they are in from 16th seed going up (with the highest Seed Score going first).
This would mean that Kansas, for instance, would get to select which #1 seed region they would like, followed by Oregon, then UNC, then MSU after they see what each bracket looks like. This would let them choose their path to the final four.
I'd also be happy to post my excel file if anyone would like to dig into the math.
Cheers!
Source (http://espn.go.com/mens-college-basketball/rpi/_/sort/RPI)
2.) I think you need to have more differentiation towards the bottom. Teams in the 100s are different than teams in the 300s
3.) this probably hurts the mid to low majors in that it's hard for them to get bigger wins, so unless you centralize scheduling (which teams will never give up control over), it makes it tough to adopted by them
March 15th, 2016 at 10:02 PM ^
Ummm... the NHL rewards +1 for a loss.
Sent from MGoBlog HD for iPhone & iPad
Hell, where are my beloved Temple Owls...winners of the AAC regular season title (14-4 in conference) and owners of two wins over a UConn team that handled Michigan. Not a fan of a system that seems to only reward scheduling and penalizes the mid-majors. Sorry OP.
Ditto...
For example, Monmouth was 17-6 away from home, 2-1 against RPI top 50, won their regular season (but not conference tourney)... and didn't get an at large.
That's basically the NCAA saying, "hey, I want you little guys to go out, play the big dogs, and you'll get rewarded."
Monmouth says, "Ok!"
The NCAA replies, "psyche!"
It's certainly an interesting concept.
Are the points based on the RPI at the time the game is played or calculated at the end of the season?
Also, agree with some of the points made by UMGoRoss above.
Speaking of points....I wonder if the point counter will get restored inside of two years. That's how long it was gone last time.
Sigh.
Sent from MGoBlog HD for iPhone & iPad
March 15th, 2016 at 10:05 PM ^
was a joke, right?
There are several problems that you have to address. You are giving a rating system a tremendous amount of weight. If it's that reliable, why not simply use it to do the seedings and stop there?
I disagree with the premise of lumping in a range of results and treating them the same. A road win vs #1 should be worth a lot more than a home win vs #25. A road win vs #26 should be worth more too, indtead it's the same as beating #100. As we've seen with Michigan, some teams will have opponents concentrated in certain areas. Things do not just balance out. You are also giving equal weight to games played at the beginning of the year and under completely different circumstances to those near tournament time. There is one smaller seeding issue, and that's in trying to avoid matching conference opponents.
Trying to seed a tournament is complicated and any formula is going to have issues. One need only look to hockey, college hockey, that is. And hockey is easier with far fewer teams and conferences. Or we could look at some of the strange computer rankings in football. I think the committee is alwats going to come under criticism. That doesn't mean that taking them out of the process would be better.
March 15th, 2016 at 11:07 PM ^
Sent from MGoBlog HD for iPhone & iPad
March 16th, 2016 at 12:13 AM ^
March 16th, 2016 at 11:48 AM ^
you're a man after my own geek heart. I was actually just thinking about a system like this after the selection show on Sunday. I cannot agree more with this approach, why make the bubble teams (and just seeding in general) such a subjective decision?!?!
March 16th, 2016 at 12:18 PM ^
Do you take into consideration the range of defeat? A drubbing of a top 25 by a 26-100 has no impact overall. Obviously stated already is the difference in a home vs away win. I would like to see some kind of impact added for that. Also I'm assuming the rating of where a team is is based on final standing, not indicative of mid-season play? For instance, if Michigan played a top 25 team and won, but that team went on a slide and finished the season ranked 75. Does that count as a top 25 win or a 75th rank win?
I justed used the final ranking
Here are some more examples to answer the comments above:
IF there was a 1st four in this system:
Some interesting teams that would miss the field:
Indiana vs. Iowa
There are absolutely flaws in this system (home vs away, margin of victory, teams ranked 150+, the arbitrary nature of a 24th ranked team vs 26th, etc)
When I looked at teams on the bubble I looked at how many top teams did you beat (Top 25), how you did vs. your peers (25-100), and if you lost to any bad teams (100+). I don't think losing to a great team, or beating a horrible team should mean much. I then put a ranking system together that generally matched that.
I'm a bit of a nerd and liked the output of the numbers...hope some of you found it interesting too.
Cheers!