ESPN's Basketball Power Index ( BPI )

Re: Yes, ESPN does make up rankings...

ICHAWK said:
They eye test?

Lets say there are up to 78 teams who you look at to make the tournament. How are you going to watch all the games that those teams play to decide who looks best to you.
Click to expand...

Why not, RPI is based 75% on SOS which is arbitrarily estimated based on W-L, but not on how the games were played. There is a lot more that should be considered in an index of ranking. I just think RPI is so biased to what people actually think about teams that doing the eye test would be a very similar measure.
 
BrueCrew3 said:
I'm sorry .. but in my mind if you play one of the 'play-in' games, then you're a 16/17 seed. You're not a 13 or 12.
Click to expand...

I don't know why they don't do it this way. I thought it used to be the worst two teams in the tournament play the play-in game, winner faces the overall 1 seed in the first round. Thus it seems like the worst 8 should play in the first four and the winners play 1-seeds. That makes more sense than having VCU play in for the 11 seed a couple years ago.
 
volhawk50 said:
I don't know why they don't do it this way. I thought it used to be the worst two teams in the tournament play the play-in game, winner faces the overall 1 seed in the first round. Thus it seems like the worst 8 should play in the first four and the winners play 1-seeds. That makes more sense than having VCU play in for the 11 seed a couple years ago.
Click to expand...

Television is why, ask yourself which you would rather watch two #16 seeds going at it or two bubble teams playing?
 
Re: Yes, ESPN does make up rankings...

hawkfarmer said:
Why not, RPI is based 75% on SOS which is arbitrarily estimated based on W-L, but not on how the games were played. There is a lot more that should be considered in an index of ranking. I just think RPI is so biased to what people actually think about teams that doing the eye test would be a very similar measure.
Click to expand...

So if you lose a game by 6 pts instead of 12 pts that should count differently? I don't see anyone complaining about the B1G being rated as #1 in conference RPI...
 
Re: Yes, ESPN does make up rankings...

hawkfarmer said:
Why not, RPI is based 75% on SOS which is arbitrarily estimated based on W-L, but not on how the games were played. There is a lot more that should be considered in an index of ranking. I just think RPI is so biased to what people actually think about teams that doing the eye test would be a very similar measure.
Click to expand...

I didn't say RPI is a great measure. My point is that it is pretty hard for each member of the selection crew to watch every single game a teams plays and remember enough to have an objective measure. I guess you could say for a team like Iowa in a Power Conference you only need to watch conference games but that is 18 games plus one or more games in the Big 10 tournament. I guess you would probably want to watch a preseason game against ISU, WSU and VA Tech to see them against other power conference teams. Lets say you have to do this for 15 power conference teams who are your last group in/out. That would be 330 games you would have to watch. Now you have to move on to non power conference teams and figure out the best teams they played and watch.

The point is it's just too many games to watch. Iowa is a great example of this. They played terrible against Michigan so would you watch that game and say Iowa sucks. They played well against MSU, Indiana, at Minnsota so based on that you could say Iowa is good.

There is far too great of a sample size you would have to look at for everyteam if you want to watch everygame to get a true idea.

You have to have some sort of ranking system like RPI to at least let you narrow data down.
 
BrueCrew3 said:
I'm sorry .. but in my mind if you play one of the 'play-in' games, then you're a 16/17 seed. You're not a 13 or 12.

GO HAWKS!!!
Click to expand...

There is a reason for this when you think about it. The vast majority of teams that get seeds from 14-16 are automatic qualifiers from the smallest conferences(likely 1 bid leagues). These teams would likely "worse" teams than the at large teams.

That's the key here, you have to recognize the difference from those auto-bids and the rest of the at-large selections.

The setup they chose is four play in games, which are two separate sets. the two 16vs16 matchups are the 4 worst auto-bid teams, while the other two (12v12 or 13vs13 etc) are the "last four in" teams from the at-large pool.

It is actually an accurate way to put teams in their proper seeds based on the quality of the team.
 
ssckelley said:
Television is why, ask yourself which you would rather watch two #16 seeds going at it or two bubble teams playing?
Click to expand...

I would watch neither if Iowa wasn't playing. I didn't know the first four had huge TV demand. Seems like there's always something more interesting on.
 
Re: Yes, ESPN does make up rankings...

ICHAWK said:
I didn't say RPI is a great measure. My point is that it is pretty hard for each member of the selection crew to watch every single game a teams plays and remember enough to have an objective measure. I guess you could say for a team like Iowa in a Power Conference you only need to watch conference games but that is 18 games plus one or more games in the Big 10 tournament. I guess you would probably want to watch a preseason game against ISU, WSU and VA Tech to see them against other power conference teams. Lets say you have to do this for 15 power conference teams who are your last group in/out. That would be 330 games you would have to watch. Now you have to move on to non power conference teams and figure out the best teams they played and watch.

The point is it's just too many games to watch. Iowa is a great example of this. They played terrible against Michigan so would you watch that game and say Iowa sucks. They played well against MSU, Indiana, at Minnsota so based on that you could say Iowa is good.

There is far too great of a sample size you would have to look at for everyteam if you want to watch everygame to get a true idea.

You have to have some sort of ranking system like RPI to at least let you narrow data down.
Click to expand...


I didn't clarify, you don't need to watch the games but account for things like margin of victory etc. in an index. RPI is not very accurate, but everyone loves it. A better way to do it would be to take all teams that play each other and create a matrix which relates all common opponents and use that, plus a bunch of other variables to actually predict an accurate ranking. RPI is highly biased by SOS, which is biased by what people think is the best conference etc.

BPI is a better ranking system.
 
volhawk50 said:
I would watch neither if Iowa wasn't playing. I didn't know the first four had huge TV demand. Seems like there's always something more interesting on.
Click to expand...

You may feel this way but think overall which games are getting better ratings? The last 4 in play in games have been fun to watch and 2 years ago VCU made a nice run to the final 4 from that spot. The worst of the automatics playing each other are playing for nothing more than to get their brains beat in the next round.

ameshawkeye said:
So no response to the actual reasoning why the play in games aren't all 16/17 seeds?
Click to expand...

A couple of us have gave you reasons. If you do not like our response and want a better researched answer then give google a try.
 
You must log in or register to reply here.

Latest posts

Share this page

Top