from #148 to #150 in the RPI. What does RPI stand for anyway? Ridiculously Poor Index?

The formula of those rating indices is so complex that games occurring three or four steps away can have an effect upon it. If teams that played teams that we played lose on a given night, it can affect us.

[quote='MICHAELSPAPPY' pid='3860823' dateline='1230831774']
The formula of those rating indices is so complex that games occurring three or four steps away can have an effect upon it. If teams that played teams that we played lose on a given night, it can affect us.

It just makes no sense to me that there is a rating system that rates you lower when you win a game. I'll never understand it, and why they would come up with a system so flawed really makes you wonder about the intelligence of the people running the basketball and football world.

(01-01-2009 04:23 PM)Dr. J. Wrote:  It just makes no sense to me that there is a rating system that rates you lower when you win a game.

1 part winning percentage
2 parts opponent's winning percentage
1 part opponent's opponent's winning percentage

The formula rewards teams that scheduled teams with above .500 percentages, that play other good teams. If you beat them, that's even better.

The formula does not reward teams that schedule weak teams, and then puts the exclamation point on it if those games are a loss.

If a conference as a whole wins OOC games, that's huge. If that conference beat teams that are good, then the conference games count a bunch. I believe that is why mid-major fans think the formula is unfair.

Beat good teams that play good competition and you've got a great RPI. It's just math.

From what I can tell it seems to be pretty accurate. When we have had very good teams it was reflected in the RPI. Wimp's NIT team was around 50 before we lost in the tournament. I think our senior laden team a few years ago was in the 60's. Good team=Good RPI

(01-01-2009 04:49 PM)Scotto Wrote:  

(01-01-2009 04:23 PM)Dr. J. Wrote:  It just makes no sense to me that there is a rating system that rates you lower when you win a game.

1 part winning percentage
2 parts opponent's winning percentage
1 part opponent's opponent's winning percentage

The formula rewards teams that scheduled teams with above .500 percentages, that play other good teams. If you beat them, that's even better.

The formula does not reward teams that schedule weak teams, and then puts the exclamation point on it if those games are a loss.

If a conference as a whole wins OOC games, that's huge.

Yep. And it doesn't really matter who those wins are against. If one of the non-conference opponents plays a weak schedule themselves that doesn't so much hurt you. Just say for example that Creighton plays a weak schedule, that matters a lot for them, but it doesn't matter very much at all for you. Because Creighton's opponents' winning percentage is only about 1/30th of 1/4th of your rpi. What matters more is Creighton's record which is about 1/30th of 1/2 of your rpi (counts twice as much).

Quote:If that conference beat teams that are good, then the conference games count a bunch. I believe that is why mid-major fans think the formula is unfair.

Yep. See the SEC. They buy games against nobodies. Then almost all of the teams rack up wins against the nobodies. Since you can't really control what conference you are in, and for us 18 regular season games are against the conference, we have very little control over our rpi. That is since about 2/3 of our schedule is against conference opponents and 1/2 of our rpi is based on our opponenets' records. So 2/3 of 1/2 of your rpi depends on the Sunbelt winning non-conference games. Not a happy proposition most years. I do have to say UALR does a very nice job of scheduling to win. You don't have to be 12-0. We just need the majority of the belt teams to find a way to win at least half of thier games and a few of the teams to win a few more than that.

Quote:Beat good teams that play good competition and you've got a great RPI. It's just math.

And in that sense, it is a fair rating. It is at least objective. You know the rules and you just have to work them.

I would like to see the various rating services put in a factor for close games. If you lose by one point in overtime on the road, that ought to rate more than losing by 40 at home.