Honest 'ranking' question

Seth53

Seth53

Well-Known Member
http://espn.go.com/mens-college-basketball/rankings

Look at the RPI and BPI. Iowa is like #40 and #45, etc.

I happen to like those sorts of polls as they take into account some important factors.

Please explain to this somewhat ranking-novice, do they mean anything to the selection committee? or are they simply a unique academic exercise?

And I don't think ESPN likes us very much...not even a sniff in their latest power rankings.
 
Last edited:
Seth53 said:
http://espn.go.com/mens-college-basketball/rankings

Look at the RPI and BPI. Iowa is like #40 and #45, etc.

I happen to like those sorts of polls as they take into account some important factors.

Please explain to this somewhat ranking-novice, do they mean anything to the selection committee? or are they simply a unique academic exercise?

And I don't think ESPN likes us very much...not even a sniff in their latest power rankings.
Click to expand...


The RPI seems to mean A LOT to the committee. I don't really agree with that, as more advanced ranking systems like Kenpom aren't looked at or taken into as much consideration. I think the Kenpom and other more advanced ranking systems do a more complete job than RPI, they just haven't been around as long, and it seem the committee take the RPI as gospel.
 
I read an article a few years ago that the committee takes a lot of metrics into account when determining seeding and other things, and some committee members even have their own ranking systems. Didn’t save it but it was a pretty informative article that tried to outline their process.

I would say though that as far as who gets in, RPI has been historically the best indicator. If you have a RPI of ~40, you’re probably in the tourney. I would say as a high major since the expansion to 68, if you are in the 40s at all you’re probably likely to make it.

Seeding is where most of the debate happens and I think they do a pretty good job. The further along the year gets the closer the metrics get to one another typically. I’d really recommend looking at the Massey Ratings Index (http://www.masseyratings.com/cb/compare.htm) which compares many different ratings tools, it’s interesting to see the outliers for programs. For example, today Iowa is ranked 28th combining all of the different ranking systems, but it looks like they are rated as high as 46th and as low as 21st. most are in the mid-20s to low 30s though.
 
BryceC said:
I read an article a few years ago that the committee takes a lot of metrics into account when determining seeding and other things, and some committee members even have their own ranking systems. Didn’t save it but it was a pretty informative article that tried to outline their process.

I would say though that as far as who gets in, RPI has been historically the best indicator. If you have a RPI of ~40, you’re probably in the tourney. I would say as a high major since the expansion to 68, if you are in the 40s at all you’re probably likely to make it.

Seeding is where most of the debate happens and I think they do a pretty good job. The further along the year gets the closer the metrics get to one another typically. I’d really recommend looking at the Massey Ratings Index (http://www.masseyratings.com/cb/compare.htm) which compares many different ratings tools, it’s interesting to see the outliers for programs. For example, today Iowa is ranked 28th combining all of the different ranking systems, but it looks like they are rated as high as 46th and as low as 21st. most are in the mid-20s to low 30s though.
Click to expand...

Good info. Thanks!
 
The RPI matters but is only 1 piece of the puzzle. Whether the committee will admit it or not the last 10 games obviously matter. Iowa had to play in the "1st round" while Nebraska did not last year. Nebraska played well the last 10 games while Iowa played less than well. Neutral court wins are a plus and road wins are huge. Another thing Iowa has going for it is no bad losses. The 5 teams that have beaten us are all good. These are just some of the things I have noticed and are just my opinion.
 
BryceC said:
I would say though that as far as who gets in, RPI has been historically the best indicator. If you have a RPI of ~40, you’re probably in the tourney. I would say as a high major since the expansion to 68, if you are in the 40s at all you’re probably likely to make it.
Click to expand...

Not saying you're wrong, but last year, Tennessee and Iowa were 25 and 26 respectively and were among the last 4 in.
 
I think there is around 16 total auto bids, so rpi's of 40 through 50 are the bubble area unless you win your conference.
 
RPI matters and it's dumb that it does. I would think that group members representing institutions of higher education would understand why it's a flawed statistical measure.
 
ryno said:
exactly why RPI is so stupid.
Click to expand...

RPI doesn't take margin of victory into account, which means it is worthless. kenpom provides the most accurate ranking system, his numbers match Vegas 99% of the time, and honestly, he isn't too high on Iowa this year, but he's the best we've got. I would hope that NCAA committee members at least look at his rankings, but I wouldn't be surprised if they didn't.
 
It's not so much that RPI doesn't take margin of victory into account (you could still make a halfway decent model without it), it's that 75% of the model's weight does not consider a team's performance in any way, shape, or form. That's just ludicrous.

I mean SOS should be a factor, but a good model would not let it be 3 times more important than the actual results of the team.
 
You must log in or register to reply here.

Latest posts

Share this page

Top