RPI Is Broken (but not how you think)
The NCAA is putting out weekly RPI rankings, but there is more going on there than you might have noticed.
Warning: This one gets pretty nerdy later on. Don’t say I didn’t warn you.
I have spilled more digital ink on the topic of RPI (Rating Percentage Index) than I would like to admit. And yet, here I am again, ready to talk about its flaws. Normally, my complaints revolve around the simplistic model they use to determine the strength of a team. Home court and margin of victory and among the huge factors that RPI leaves out. This is a problem, but it is not the problem that I want to talk about today.
How does RPI work again?
I will keep this brief because RPI is simple (too simple if you ask me.) There are three components to calculating RPI.
Winning percentage (WP)
The winning percentage of your opponents (OWP)
The winning percentage of your opponents’ opponents (OOWP)
The formula to calculate RPI doesn’t take a mathematician.
RPI = (WP * 25%) + (OWP * 50%) + (OOWP * 25%)
Notably, RPI does not count every match a team plays.1 For National Collegiate Men’s Volleyball, only matches against National Collegiate Men’s Volleyball teams count. This is the NCAA’s way of saying only matches against D-I and D-II teams count. Matches against D-III, NAIA, and other divisions are not part of RPI.
Only matches where both teams are in D-I/II are in scope. But there is a catch.
Wait, who is included?
Since I have my own personal grievances with the NCAA and RPI, I calculate RPI myself.2 To make sure my numbers were matching the NCAA’s, I compared them this week and I was surprised that there were larger differences than I expected. The differences were small, but it did lead to two teams swapping positions in the top 10. This seemed odd, so I started to dig deeper.
I went to the latest RPI Ranking3 from the NCAA and scrolled down. That is when I saw it.
67 teams. This was one more team than I expected to see in this list. Thankfully, I didn’t have to look far for the missing team: UC Merced.
My first reaction was “wait, what?” UC Merced is definitely not competing in a D-I/II conference and I have heard talk about them potentially applying for the MPSF when they leave the NAIA. But I thought they were still competing in the NAIA. Right? Am I taking crazy pills?
Further down the rabbit hole I went. As of March 17th, UC Merced was in 2nd place in the California Pacific Athletic Conference on the NAIA website. Their schedule is full of NAIA teams in addition to a couple of MPSF members (Vanguard, Jessup, Menlo, & Stanford). Their schedule even includes their conference tournament for the NAIA, which they are hosting!
When looking at the university’s press release, they mentioned that the Bobcats would be competing at the D-II level starting in the 2025-2026 school year, with their first first of postseason eligibility in 2026-2027. As you may have noticed, it is still the 2024-2025 academic year. According to the school, they are not yet competing at Division II.
Then I turned to my nemesis (too harsh?), the NCAA. In a press release from July 2024, they announced that several schools were accepted into the first year of provisional membership into Division II, including UC Merced. For Merced, they are in a two-year expedited provisional membership track. This means that according to the NCAA, UC Merced is a Division II program (regardless of where the school is actually competing during the transition).
At first, I hated this. I have cooled down a bit, but I still have some qualms. UC Merced is in no man’s land. They have no reason to compete at the D-II level, since they cannot make it to the postseason. But the NCAA feels that their record must count for calculating RPI, even though other teams competing at their level do not count.
Counting these matches feels like it goes against the concept of RPI.4 It also improperly attributes an RPI to UC Merced when no other NAIA team is treated the same way.
That all aside, the real question is: does it matter? Does including UC Merced into the RPI calculation for 2025 make a difference? So, I looked into it.
The UC Merced Effect
Let’s start by looking at the NCAA’s top 10 in RPI.
Specifically, I want to point out 5 and 6: UCLA and McKendree. Side note: McKendree at six in RPI is incredible and I love it. When you take UC Merced out of the equation, this is what you get for those two schools:
McKendree - 0.64771
UCLA - 0.64562
Since UC Merced has played only MPSF teams (and lost to them), UCLA benefits from including UC Merced in RPI. This is enough to move them up a whole place. If you are like me, I wanted to know who benefits the most (and least) from including UC Merced. So here you go.
One team benefits the most: Menlo. Their RPI goes up by 0.0259 points. This is huge in the RPI world. All of the other gains are fairly limited, but UCLA and UC Irvine are notably near the top. Down at the bottom, there are two teams that actually lose RPI: Stanford and Jessup. Most teams are not impacted much, but there is a non-zero change to so many teams.
So now what?
I gave in and added UC Merced to my RPI calculation, because I want to recreate what the NCAA does. For those wondering, I re-compared my numbers to the NCAA and the largest difference was 0.00001 which is definitely due to how the NCAA rounds. I won’t fight them on that, but I mention it so that you can trust the RPI numbers I put out later this season. I do this for you.
I’m super interested in what you all think though. Did you know about UC Merced? Do you think they should count in RPI? Does it even matter since KPI (probably) counts all matches regardless of division? Do you think the NCAA even cares about this? I’d love to hear what you think.
Rankings Update
VBelo Rankings
It’s been a minute since I have put out the VBelo rankings and you know what, I think they are looking pretty decent (if I do say so myself). Long Beach is still at top but the gap between the Beach and the Bruins is not as large as I would have thought.
Hawaii and UC Irvine are essentially tied with USC and Loyola right on their heels. Then you have Lincoln Memorial and McKendree rounding out the top eight. McKendree is getting the love from the coaches these days (see poll later on), but LMU is not getting quite as much.
One of the teams that is on the rise right now is Lindenwood, coming in at twenty. They have been making a splash in the MIVA, post Hawaii, and it is yet another amazing story happening this year. What is wild is that Lindenwood is not even Receiving Votes in the coaches poll right now. They have two big matches against Loyola this week and this could be an inflection point. If they split the series, I think they should at least be receiving votes.
Coaches’ Poll
It finally happened. Penn State is no longer ranked. It took a lot for it to happen, but in the end there just has to be 20 better teams than one that is 5-13 on the year. George Mason moving up is a great choice. There are a couple of teams that I could see in the hunt, but no complaints from me.
Loyola got rewarded A LOT for beating CSUN. A couple of things here. First, this was a home match for Loyola. Also, the match went to overtime in the fifth set. I’m not convinced that they should be rewarded this much for their win. To show the disconnect between the coaches and the VBelo Model, the Ramblers had a 72% chance of winning this one. So yes, it is a strong win against a ranked opponent. But also, maybe the coaches have overvalued this one win.
Media Poll
Loyola continued to see a big bounce from its win over CSUN.I love that McKendree is still in the top 10 conversation with the media. The MIVA has been making huge headlines this year. And to think this year was going to be all about the West and how strong the teams are. I love to see good volleyball and McKendree is getting the attention they deserve for some great volleyball.
In’s & Out’s
A section for all of the random things that don’t fit anywhere but belong somewhere.
The march of the Ace King. Moni Nikolov is still on track to break the single-season ace record.
AVCA Player of the Week: Kory Grant (Opposite, Mount Olive)
Triple-Double Watch 2025: Nothing
They also don’t count matches where the original team plays their opponents in OWP. This means that the opponents winning percentage is really their winning percentage, except when they all played the team you are calculating RPI for. Yes, this was the hardest part to code for me, especially when I got to OOWP.
Is the opposite of a passion project, a rage project?
My favorite Easter Egg in the NCAA Stats site is that they named the section for RPI rankings Nitty Gritties. https://stats.ncaa.org/selection_rankings/nitty_gritties/43472
Am I defending RPI? This feels weird.