While Michigan’s non-con SOS certainly isn’t bad (I only start penalizing teams if they’re below 175 or so), it could be so much better.
In fairness, I imagine if UNC and Iowa State didn’t turn into dumpster fires, the NC-SOS would still be easily within the top 100 even with those low buy games.
My big issue with NC-SOS and SOS on the team sheets is that it uses the old RPI component and not something smarter. I see no reason why there can’t be a NET-based SOS.
I’ve been wondering about that. I saw some people espousing some team’s really hard schedule (don’t remember who at this point) then I looked at KenPom’s numbers and their schedule was nothing to write home about and I started to wonder where they get their SOS numbers. That’s really frustrating that they “got rid of the RPI” but still use it as their SOS indicator. Makes absolutely no sense.
I’m confused how Michigan can have a SoS of 52 given the number of Quad 1 games they’ve played and how good the Big Ten is this year. IIRC Kenpom has Michigan’s SoS in the top 10.
This is excellent, thank you. A 6 seed for Michigan would be great…I’ve always preferred the 6 seed to the 4 or 5 seeds, just to avoid playing 1 seeds for as long as possible. The gaps between teams get pretty small once you get past the 1 seeds.
Part of that too, I would argue, is avoiding the elite mid-major teams in favor of a middling bubble P5 11-seed as well. Plus, like you said, a three-seed is much better second round competition than a one-seed.
Ace Anbender at Mgoblog got Bart Torvik to look at Michigan’s splits with and without Livers, and we’re #11 in the country when Livers plays. Regardless of where we get seeded (presuming it’s in that 6-10 range as this article predicts), people are going to hate seeing us in their part of the bracket.
Yeah, I think that is really going to depend on the next month (hopefully) with a healthy Livers.
Going back and grabbing the Bahamas games will make any splits look good though 
Baseline for Michigan is to defend home court down the stretch. If it can win a couple road games then I think it is fair to start raising expectations.
Was thinking, the one upside to not being a protected seed this year: Pretty unlikely that we’ll have to watch them play Montana again.
I’ve seen some brackets where we are still in their Pod, which would be very funny. If we get a 7 seed and they win their conference tourney again there could be a theoretical 7-15 2nd round matchup
I’ll consent to this, so long as the 2 seed they beat is either Duke or Kentucky.
I’m all in for round 3 of Sayeed Pridgett and Timmy Falls
What’s the deal with the gap between the Kenpom rankings and the NET? And why is MSU so far ahead of Michigan in the NET? MSU is 15 and Michigan is 25 when IMO, the resume is better for Michigan.
NET resumes:
Michigan
Q1: 6-8
Q1A: 4-4 (wins over #3, 11, 15 and 33)
Q2: 4-1
Michigan State
Q1: 5-8
Q1A: 2-6 (wins over #16 and #34)
Q2: 5-1
So what gives? Why is there a 10 spot gap between them?
Here are the team sheets:
My guess is that the gap comes down to some of the efficiency margin stats at play. MSU is ranked 9th in Torvik (filtered by one day to avoid preseason priors) and Michigan is 14th. A few more blowout wins by MSU overall.
Where’s Purdue at in the NET? Is Saturday another opportunity for a Quad 1 win?
Yes, the upcoming games are on the bottom of the team sheet. Four of Michigan’s final five game are Q1 … although Wisconsin probably won’t stay in the top 30.
You can never really get a great answer to this because NET is a blackbox. As Dylan said, it’s probably just something to do with the idiosyncrasies of how NET handles effeciency (which is in some proportion to use MOV up to 10 points and also raw effeciency margin)
Hard to fathom that a home game against a Big Ten opponent is a Q4 affair but here we are with Nebraska languishing at 175 in the NET.
How is the SOS calculated on these? On the advanced metric sites, Michigan is 1st (Torvik) and 3rd (KenPom) in SOS. 57th here.
Same for MSU – much higher on the two sites above than the 46th listed (I get that the rankings for each set are different, which leads to small changes, but this gap seems really wide).
It is basically the RPI SOS component which is absolutely insane, but … oh well.
