Gaussian Wins Help
Posted: Mon Aug 19, 2024 9:17 pm
Has anyone done anything with Correlated Gaussian wins?
Looking into Calculating Correlated Gaussian wins from team's gamely Ortg, Drtg in Excel based on what Dean Oliver did in Basketball on Paper (chapter 11). And failing.
I won't pretend to understand the math. So keep it as simple as possible if you can help.
Don't know if Excel has changed since 2004 or me not understanding means I'm wildly messing up.
fwiw ...
I've got '84 Hawks season overall rtgs from Reference as
Ortg: 105.5; Drtg: 106.9
I've got game level stuff
(gamely Ortg in I4:I85 -- Drtg in J4:J84)
Variance for Org (in I88) 110.7371267
Variance for Drg (in J88) 99.52884741
Covariance (in J89) 30.63526175
Then I can't get the NormDist calculation to work as presented in the book.
Any ideas or exemplification for how it should work? And/or where I'm going wrong.
Thanks in advance for any help.
Looking into Calculating Correlated Gaussian wins from team's gamely Ortg, Drtg in Excel based on what Dean Oliver did in Basketball on Paper (chapter 11). And failing.
I won't pretend to understand the math. So keep it as simple as possible if you can help.
Don't know if Excel has changed since 2004 or me not understanding means I'm wildly messing up.
fwiw ...
I've got '84 Hawks season overall rtgs from Reference as
Ortg: 105.5; Drtg: 106.9
I've got game level stuff
(gamely Ortg in I4:I85 -- Drtg in J4:J84)
Variance for Org (in I88) 110.7371267
Variance for Drg (in J88) 99.52884741
Covariance (in J89) 30.63526175
Then I can't get the NormDist calculation to work as presented in the book.
Any ideas or exemplification for how it should work? And/or where I'm going wrong.
Thanks in advance for any help.