Lotta basketball still to be played, but we are on the right track. Go 'Cats!
http://m.espn.com/ncb/rpi
http://m.espn.com/ncb/rpi
RPI #15
- Thread starter backdoorpass
- Start date
We're ahead of Kentucky, Indiana and Duke. Computers sure are interesting.
So it's really a rating of athletic directors and not teams. I actually think it is a combination of SOS and record
OK, I looked it up. Here is the definition:
"In its current formulation, the index comprises a team's winning percentage (25%), its opponents' winning percentage (50%), and the winning percentage of those opponents' opponents (25%). The opponents' winning percentage and the winning percentage of those opponents' opponents both comprise the strength of schedule (SOS). Thus, the SOS accounts for 75% of the RPI calculation and is 2/3 its opponents' winning percentage and 1/3 its opponents' opponents' winning percentage."
RPI has less to do with how good you are, but much more about how good the teams you have played are.
If all are injured guys get healthy I don't see any reason why we can't have an RPI of top 10 we're going to win their next five games
The problem is, the poor records of the teams (except for Dayton) we play in the next 5 will lower RPI, even with NU winning.
Sounds like we'd be better off losing to improve our opponents winning percentages
Our next 5 opponents (including Dayton) are a combined 17-18. New Orleans just won by 16 at Washington State (which is a bad team, but a nice win for NO that slightly helps our RPI assuming we take care of business).
But, KenPom of our 4 non-Dayton opponents are 155, 249, 298 and 331, which doesn't bode well for those teams being very good in their leagues (except maybe IUPUI).
RPI does not equal batting average! I get how you can both love OPS+, WAR and WHIP and still love batting average and ERA. We grew up with them. But, can we PLEASE assign RPI to the dustbin of history?? No one loves RPI. It just appeared 25 years ago or so. And, there are far better metrics out there. I wish the Basketball Committee would make that declaration. Sorry. I am off my soapbox now.
You are right. It amazes me that the NCAA Tournament Committee reportedly still relies on RPI for seeding and bubble bursting. But it is indeed still a thing.
I am a kenpom subscriber (literally). I am starting to hear more sports journalists refer to kenpom, so I think that it is catching on. And, for the record, we are currently at #43 per kenpom. This seems about right at this point.
Each member of the NCAA Tournament Committee can reach his/her own determination of the T-worthiness of any given team. They do not have to follow any particular human or computer ranking system. Of course a member MAY do so (if s/he wants to). But there is nothing in the rules that forces members to follow any particular ranking system.
Members do get an info package that includes several rankings, including RPI's, Sagarin's and KenPom's, among others. It's up to each member to choose how to use that info.
Not even that, since your opponents may be relatively weak teams with inflated w/l records (for example by being relatively strong in weak conferences), which then may inflate your RPI, without it meaning your opponents are really good.
The RPI is very nearly meaningless. Its only virtue may be its simplicity and its "transparency" (anyone can compute it to confirm the ranking of any team).
A few years ago (I haven't researched this recently), I read that the selection committee doesn't actually take the team's RPI into consideration, but their opponents' RPI. They use it to gauge quality wins, which they do give considerable weight. The strength of schedule (opponents' records and opponents' opponents' records) are a significant part (75%) of the RPI calculation and can inflate the RPI, but without enough quality wins against that schedule, the selection committee will frown on that team.
Preventing that is, in theory, the function of the opponents' opponents' winning percentage component - but it's a garbage metric either way.
Yeah, but still 50% of a team's RPI is determined directly by its opponent's w/l records, regardless of whom they play.
Anyhow, many of the opponents of a top team in a weak conference could have decent w/l records (by playing many even weaker opponents).
I guess the fundamental problem is that w/l records could be very misleading when there are 300+ teams distributed among dozens of conferences of extremely different degrees of competitiveness, and any one team plays only a tiny fraction of the remaining 300+ teams.
I heard something like that (say, a team's record against top-50 RPI, against top-100 RPI, etc). The RPI might be just good enough for that.
But that may be an informal arrangement among committee members in a given year, as opposed to something that is built into the rules.
As far as I know committee members are 100% sovereign in their decisions (but they probably do try to come up with semi-objective criteria to justify their decisions if they come under fire).
But the major conferences do tend to have the best RPIs from top to bottom.
Well in effect, a lot of the 50% opp WL portion cancels itself out across all teams because everyone plays in a conference where all the teams play each other (which nets to ~.500). The balance that makes a difference is, in no particular order:
1) your own W/L %
2) the W/L of your OOC opponents
3) the OOC W/L of your conference opponents
4) the SOS of your conference opponents (which gets back into the same feedback loop but you get the idea)
So, not to repeat Styre, but in theory all of things are reasonable to consider in terms of trying to differentiate teams... it's just a garbage metric because nowadays we have so many other better ways to do so that are both more precise and more comprehensive. KenPom being the most obvious, but there are plenty of others beyond that.
Incidentally, going a bit off topic but I've always thought BPI and FPI from ESPN aren't particularly good either. They pop out too many wacky results and don't seem to put enough emphasis on actually winning games. They've also started showing strength of record more, which I think is better, though less comprehensive than eg KenPom or similar.
Now all of a sudden, a win against us might be what propels a team into NCAA consideration
This question came up a few years ago when we were in the hunt for a little while, but does anyone think the Northwestern factor would play into the committee's decision? Meaning, if we were a bubble team and the committee sat there with Northwestern on their consideration list and knowing full-well the historical significance of putting NU in the tournament, does that sway them at all?
I doubt it. Too many deserving at-large teams on the bubble, if they made it obvious they were picking NU out of sympathy or for publicity they'd lose credibility. When NU gets in, it will be because they earned it.
Never fear. Collins is building a winner. Bright days ahead, Cats make the Tourney in 2018.
Is it really that big of a story outside of NU fans? I think it will be similar to the bowl win - a little more national press than what would normally occur, but not a story. Now, if they make a run similar to the 1995 football team you have a story.
Not so fast. Yes every team plays in a conference. But every team has a lot of OOC games. Depending on the size of the conference, the OOC portion of the schedule could be similar to the IC portion, or even greater (for example, last season Ch St played only 14 of 32 games in its conference, the WAC).
AFAIK, conference records play no roll in the RPI. Hence, it is not necessarily true that the w/l record of in-conference opponents average about .500. It depends on each opponent's OOC record. Conceivably, most teams in a conference could have over (or under) .500 records, depending on how each did in its own OOC schedule. For example, in a very balanced conference, the in-conference record of each team may be close to .500. But if each played a soft OOC schedule, all may end up with relatively high W/L record (with each winning most of its OOC games, and roughly half of its conference games).
Inasmuch as Jill Stein’s alma mater (Harvard) ranks below both NU and Yale, she is petitioning for an audit of the calculations.
Last edited:
Similar threads
