KenPom is for Gopher fans what Nate Silver was for democrats, hopefully

[From the Barn]

Let us all take a deep breath, glance at the match, and hope the projected Big Ten win curve plays out like this.



Math triumphs over psychology, I hope.

 

[Gopher Warrior]

There is some bad math in there, though. There are assumptions the model makes that are not in line with reality. If you adjust for just a couple of key facts, things get (somewhat) worse. Gophers will be fine, but they're a tad bit overrated by KenPom right now because of how the model works.

 

[JDGopher]

There is some bad math in there, though. There are assumptions the model makes that are not in line with reality. If you adjust for just a couple of key facts, things get (somewhat) worse. Gophers will be fine, but they're a tad bit overrated by KenPom right now because of how the model works.

Can you tell us about the math? I don't mean that to sound like I'm doubting you, but I do really like hearing about math in sports so I'd love to hear about your thoughts on KenPom stuff.

 

[howeda7]

There is some bad math in there, though. There are assumptions the model makes that are not in line with reality. If you adjust for just a couple of key facts, things get (somewhat) worse. Gophers will be fine, but they're a tad bit overrated by KenPom right now because of how the model works.

So you're going to unskew it for us?

 

[Gopher Warrior]

Can you tell us about the math? I don't mean that to sound like I'm doubting you, but I do really like hearing about math in sports so I'd love to hear about your thoughts on KenPom stuff.

Sure... so, here's something from last week titled "Big Ten Kenpom Efficiencies: Raw vs. Adjusted". Near the bottom is a discussion of some adjustments that can be made when looking at Minnesota.

Kenpom is invaluable. It's just if you don't understand it very well, especially its limitations, you can miss things. (By the way, I did discuss my high level methodology for some adjustments I made for Minnesota with Ken.)

Any other questions on this, just let me know... I like talking numbers (except for polls & rankings - those are mostly worthless).