Michigan Basketball's RPI Problem
Following Michigan's two big wins the last 5 days, I thought I would take look at our RPI as it stands today and how it will play out for the rest of the season. Fair or not**, the RPI is a big factor in determining seed lines. Brian and many others have been critical of Beilein's inability (or unwillingness) to game the RPI by playing #125-175 teams as opposed to #250+ teams.
RPI as of today: #21. Terrific, right?
In the next three weeks, Michigan will play #260 Detroit, #327 Jacksonville, and #347 Alabama A&M. Live-rpi.com has a neat little tool that allows you to simulate results to give you an estimate of RPI in the future. Let's assume that Michigan gets the three wins as anticipated.
RPI as of January 1: #52. And this is after three wins.
This demonstrates how much of a negative impact poor scheduling can have on your RPI. Most would agree that non-conference play (LSU game aside) has been a relative success this season. Michigan will likely be sitting at 12-3 (1-1) with multiple quality wins but will have the RPI profile of a bubble team. When Selection Sunday rolls around and Michigan comes in 2 seed lines lower than expected, this is probably your reason why.
**The RPI is just an incredibly stupid tool to measure the quality of teams, but the system is the system.
December 15th, 2017 at 2:10 AM ^
I think we let recency bias from last season impact our feelings on this excessively.
Def not systematically being underseeded.
Kenpom also does a thing that shows how little your seed matters for your odds of getting to the final four or winning it all. It does matter a little bit for sweet sixteen, but if you're underseeded and upset a top 4 seed, you get their path.
December 14th, 2017 at 6:38 AM ^
December 14th, 2017 at 7:11 AM ^
December 14th, 2017 at 8:34 AM ^
That being said, he can find a middle ground to avoid RPI problems and still help his team.
December 14th, 2017 at 8:39 AM ^
December 14th, 2017 at 8:41 AM ^
December 14th, 2017 at 10:00 AM ^
Texas was a 6 seed two years ago and has a future possible NBA lottery pick at center. UCLA was a 3 seed last year. UNC is UNC. LSU and VCU (and Chaminade, for that matter) were part of the Maui tournament, so we had less control over it, and we were unfortunate (since we lost to LSU)to play three of the four weakest teams in that field (but not the weakest, which seems to be Cal. seriously, they lost to UC Riverside and Central Arkansas and got blown out by Chaminade. They are butt).
The biggest problem with our schedule is that we lost to LSU. Replace VCU with Notre Dame and Marquette and this schedule looks just fine. It's also not Beileins fault that the rest of the B1G is shitting its collective bed. Also keep in mind that the
You can't ask a B1G team to schedule 6 top 100 teams in the non-conference and 6 more from 100-200. First, you'd have to play more road games. Second, it makes it much harder to develop freshmen. Third, if you play 6 games that you have a 75-80% chance of winning, you're probably going to lose 1 or 2, and a loss to an RPI 100-200 team is the inverse of a win over a team in the 25-50 range. I think Beilein wants to minimize the chances of a loss like that happening, and its hard to blame him.
December 14th, 2017 at 1:12 PM ^
Also, we should never schedule a team that has even a possibility of slipping below 250, and that happens regularly.
December 14th, 2017 at 8:59 AM ^
1. What has been Michigan's RPI after two conference games in January the last few years? Won't their RPI rise with a decent (11-7) conference record?
2. Playing bad teams will allow Mo to heal up.
December 14th, 2017 at 9:17 AM ^
But does this calculate Wins by our opponents? Like LSU, N Carolina VCU, UCLA Texas etc?
December 14th, 2017 at 9:31 AM ^
25% is your winning percentage, 50% is your opponents' winning percentage, and the final 25% is your opponents' opponents winning percentage. So the record of your opponents and the records of their opponents makes up 75% of the formula.
December 14th, 2017 at 9:24 AM ^
December 14th, 2017 at 9:46 AM ^
I absolutely love Beilein and think the "fire Beilein!' crowd needs to be negbanged.
That being said, why does he continue to schedule such terrible RPI teams? It's a lose-lose scenario because it's of no help when we win and then there's situations like NJIT...
December 14th, 2017 at 10:43 AM ^
NJIT finished 2015 with an RPI of 134. Eastern Michigan finished 2015 with an RPI of 148. Those losses were devastating to that Michigan team, which very well could have at least made the NIT if you replace those losses with tomato can wins. Those are the things that happen when you play a bunch of teams at that level: sometimes they beat you, and one (or two) bad losses isn't worth the chance of winning 6 games against teams like that. Hell, the 2008-09 team needed a big comeback and overtime to beat Savannah State, and if they lose that game they don't make the tournament that year even with wins over UCLA and Duke on their resume.
Put it another way: when making a case for or against a bubble team on selection sunday, nobody touts wins against Eastern Michigan. Games against anybody with RPI 100+ are largely ignored. RPI and SOS are tools, but they pale in comparison to the importance of big wins and bad losses. Beilein does not want bad losses on his resume, and he wants to schedule opportunities for quality wins. If you want to argue that he should replace teams with RPI > 300 with teams in the 200-250 range, then I can't argue with that, but he wants games that he can't lose so that he can safely treat them as glorified scrimmages to build chemistry heading into the conference schedule. I think that if Delaney hadn't pulled the crap with the BTT and the Indiana and OSU games were after Jan. 1, Michigan beats OSU because they have
December 14th, 2017 at 10:57 AM ^
I think this is a good point. I understand why it's frustrating to many people, but I'm more upset at anyone on the committee that uses RPI than I am at Beilein. Yes we will win 9 out of 10 games against the RPI 150ish teams, but that leaves us with a loss in 1 out of 10 games, so about every 3 or 4 years we'd have a loss that holds our resume down.
December 14th, 2017 at 10:53 AM ^
I doubt theres a Power 5 team in the country that doesnt have 4 or 5 cream puffs on their schedule so its all relative
December 14th, 2017 at 11:24 AM ^
You provide a great example showing how the RPI violates a fundamental principle of rational ranking. Winning over a lowly ranked team provides no negative information about a team's quality. Other than decreasing the schedule of relevant games played by one, it is at worst irrelelevant.
But I thought that the RPI was being phased out this year by the committee. It's astonishingly stupid that they kept using it this long.
December 14th, 2017 at 12:43 PM ^
December 14th, 2017 at 11:44 PM ^
That was a disaster: NJIT snowballed into the EMU/Arizona/SMU debacles. Just looking at the mathematical effect of a single loss doesn’t take into account the real-world impact.
So, from Beilein's standpoint, there’s some logic to scheduling super-weak teams. If the objective is to get a tuneup game, you might as well make sure that’s what you’re getting.