Hack to translate 4 factors data into off/def efficiencies?

[Crow]

Anybody got a simple formula or excel macro to estimate team efficiencies on various selected four factors input values? Is this something Ed K., Dean O. or anyone else did in past? What are the change levels needed for each factor to raise / lower a team efficiency by 1 pt?

[xkonk]

I feel like Evan Z had regression weights on his blog, but I don't have a link handy.

[Crow]

Thanks, I've seen the regression weights before. This would be different, at least more direct to what I was looking for.