. . . still interesting to me that at this late point in the season, the two primary sources for estimating RPI on a timely basis - Warren Nolan and Boyd - disagree on their top 3. Both have Florida #1, but Nolan has Arkansas #2, while Boyd has the Black Bears rated 2nd. They also have Stetson and Stanford reversed (at 4 and 6, respectively).
This is not so unusual the first 6 weeks or so of the season, but the second half of the year - particularly, by May - they almost always agree at the top. The RPI itself isn’t always the same, but the order of team is what I’m talking about.
You can’t check it against the real thing now. The NCAA stopped updating RPI once it had served its purpose, which is to select the tournament at-large teams last Monday. Also, OM may move past us once they finish off Tennessee Tech today.
The differences are sometimes so subtle that you can’t define them. What I mean is that it can be as tiny of a difference as how a game site was determined. Was it a neutral field or a home field? And, it might not be with a game involved with your team. It might be one that involves someone you scheduled and it determines a slight change in their RPI. Hence, it effects your RPI. For instance, is North Little Rock a neutral field for Arkansas? I’d say no. But it might be considered that. How about San Diego playing at San Diego State against Arkansas? What about San Diego playing San Diego State at San Diego State? What about San Diego State playing San Diego at the major league park? Some of those did not happen, but they could. And, if those teams played Arkansas, their RPI effects Arkansas. At least, I think they do?
You are right on target, Clay. For some of the reasons you mention, a team can actually win a game, but drop 3 or 4 slots in the RPI. Or, lose and move up. Such movement is based not so much on who they played in that game, or the result of that game, as it is what happened to some of the other teams they played earlier in the season - or, what happened to the teams those opponents had played. Remember - RPI is 25% the opponents of the teams you play - their opponents, not yours (well - in addition to your opponents). Only 25% of a team’s RPI comes from that team’s own W/L record. The remaining 50% does come from their direct opponents’ record. So, games that no one is paying attention to can have results that ripple through a conference and affect RPI.
Of course, there are little nuances to RPI above and beyond those 3 factors - things like record against different groups (top, middle, bottom tiers), road record (road wins count more; home losses count more), etc. - and those “other” factors are the “special sauce” that the NCAA does not divulge. Nolan and Boyd obviously have much of that info, but not all - or else their numbers would be identical, and they are not. And that’s what the OP of this thread was about; I’ve rarely seen them disagree this much, this late in the season.
And Jeff - I am aware that the NCAA shuts down their RPI machine once the field is set. Still, many of us are interested in the on-going pseudo-rankings as the tournament progresses. I’d think that you’d be one such person.
Not really. I quit checking RPI in hoops after Selection Sunday, ditto baseball after Memorial Day, because it’s being settled where it should be – between the lines. I don’t need to check RPI to know that Washington and Fullerton (and the Chickens) are surprise teams coming out of the regionals.