The RPI is dead, but will NET be an improvement?


#1

#2

#3

Not a fan of moving to a predictive model for this, I think I’ve said my piece on this subject before, lol.


#4

Not sure what’s going on exactly with NET yet as I haven’t had the chance to read any of the articles, so I might be talking about stuff that’s now irrelevant, but…

I’m not a fan of using efficiency rankings to actually decide who gets into the tourney. I think the best system to accomplish what the NCAA tournament sets out to do (Which I think most would argue isn’t necessarily to have the best team win) is merit based like RPI, one that is mainly going off of wins and losses. However, RPI was doomed when the NCAA tried to treat it as a ranking of the best teams AND a ranking of the most successful teams. This was shown through their use of opponents RPI and opponents’ opponents RPI as extremely high factors in a team’s ranking. 100% of a team’s strength of schedule was determined by a system that only determines success and not pure ability. If they used a similar system based on actual results and not theoretical predictive rankings, but used a composite of the efficiency ones like kenpom, sagarin, tarvik, etc. to decide how much to weigh each win and loss I’d be pretty happy.


#5

How about a Google black box?


#6

I’ll wait for someone to explain to me the difference between “efficiency” and “scoring margin”…


#7

Lmao, this is great. They just randomly threw pace-adjusted and non-pace adjusted versions of the same stat into the calculations?

That takes some impressive lack of knowledge.


#8

One is capped at 10 points


#9

Mushing all this predictive and results modeling into one number seems questionable, but doing it to come up with a proprietary, magic number, and then still having a committee to not follow that number, seems dumb – or like a money grab.


#10

And that makes sense how?


#11

“NET powered by Google Cloud”


#12

I have a hunch that these will be weighted heavier than RPI was in the selection process, which I do not think will be a positive step.


#13

If you are going to put this much time, money and press into a new metric… you are going to use it. Why not just eliminate the committee and generate the S-curve from NET?


#14

Once again, in their ultimate wisdom, the NCAA has secretly and behind closed doors figured out how to build a system without transparency. How the hell are coaches and fans able to build any trust and confidence in the committee.


#15

#16

Laughable.


#17

#18

#19

“What if we just grabbed every stat ever made and just threw it together into an amorphous blob?”


#20

Fun, cool, hip graphics! The NCAA is getting younger and streamlining its efficiency! We’re in the age of big data. This is a truly transformative era. Once we unpack this info we’ll be able to see big picture and move the needle on important issues!

I hope the /s isn’t needed.