I hope the committee factors in how badly we beat some of them too. 20 plus trouncings of state, Indiana, Marquette, smu is impressive. I still can’t help be bitter on those close losses. We could be off the bubble and playing for a good seed had Walton woke up sooner/ we played d and rebounded with this intensity
1 Like
10 seed in Indy vs. board favorite Chris Mack
I’m not sure the board could survive such a matchup lol
1 Like
I’m going to echo Dylan’s tweet and say here that it’s crazy Furman is a top 100 RPI win (88).
I hadn’t even heard of that school until MSU played them in football.
Now we have to hope that Indiana doesn’t totally implode - they’re at 93 in RPI and it’d be pretty nice if they stay in the top 100. Nebraska at 86 and Marquette at 84 too.
Admittedly ignorant on the history and rationale here, but have done a bit more homework and RPI seems like a really awful metric to have a multi-billion dollar industry anchored to. I hope it’s priority is diminishing over time.
1 Like
Lunardi now has Michigan as an 11 seed and one of the last four byes.
Since there is literally a committee of smart (?), sports-related people assembled to make these decisions/rankings, you’d think using the ‘easy’ statistical models would quickly become an archaic method. I get that being on the selection committee isn’t their sole jobs, but as @JVS said, there is a lot of money–and peoples’ jobs, legacies, recruits, etc.–riding on the selection methodology; so much so, that I’d hope they do more than glance at stats, which is why it seems like paying attention to margin of victory, injuries, timing, etc. should be part of the informal equation
The thing about RPI is that even though it sucks, it’s probably a better stat for judging tournament seeding than something like KenPom. If you only because you have to reward people for the results of the games. I think seeding based on a predictive measure is bad because it diminishes the actual games being played. I’d be interested in some sort of RPI improvement with a similar concept, but I’d never want KenPom to be used to seed teams.
1 Like
I would LOVE to know how and why he has Wichita St as a 9 seed? Seriously, it’s ridiculous.
I understand not going fully off the traditional resume, but how are they better then a lot of those teams behind them?
I believe Kenpom likes them but that’s all they have for justification.
This is very true. I wonder if there’s a mathematical way to combine results metrics and predictive methods; like .5(RPI)+.5(Kenpom)=ranking
Is BPI not in some form a blend of Kenpom predictive elements, and quality/strength of actual performance? I know ESPN is pushing it b/c it is their measure, but curious if that solves some of the deficiencies of the extreme measurement methods (or if its too similar to one side or the other).
Keep in mind that KenPom is a “what have you done for me lately” metric. It strongly weights more recent results. It would be interesting how well his metrics predict tournament success.
Not as much as in the past though:
http://kenpom.com/blog/ratings-methodology-update/
There have been some other changes to the system as well. The weighting coefficients to handle recency and game importance have been changed. Essentially, recency is less important than it used to be and game importance is less sensitive to margin and opponent than it used to be.
Out of 87 brackets on bracket matrix 77 have them in with an average seed of 10.09, having them as a 9 seed isn’t that big of a leap. 59 brackets have Michigan in with an average seed of 10.58.
But why is my main gripe/question? What have they done to warrant that? (Not a dig at all on your data, I’m legit curious what in the world experts are seeing besides their Kenpom data)
I feel like they’re a perfect example of why do they get such a benefit of the doubt to go off of the eye test? I fully understand and love it’s not all strictly about the resume, but I strongly feel this year’s version of Wichita St doesn’t fit the eye test to warrant an at-large with that resume.
1 Like
Wichita State is 1-4 vs the RPI Top-100 that’s pretty crazy.
1 Like
Exactly! And that win was at home against Ill St without their best player. Their best non-conference win? @Oklahoma.
3 Likes
Clemson is another team I’m not sure about. They have a nice win @ South Carolina, and the RPI numbers aren’t bad…but they are 3-9 in the ACC. 3-9!!! Just hard to convince me they belong in the discussion right now.
2 Likes
Highest seed I’ve seen so far: Michigan an 8 seed according to Sporting News
1 Like