how do I use a polynomial formula to extrapolate data?

[jmckeone]

I've been able to use linear trendline formulas to extrapolate future data without an issue. In a y=mx+b scenario where y is the usage level and x is the date. However when I try to do something similar with a polynomial formula such as y = 1E-254x^56.468 the math goes to pot. Can anyone help?

 

[hatman]

That's the nature of Polynomial trendlines: they tend to diverge quickly once you move outside the data set: good for interpolation, not so good for extrapolation. What often happens in these cases is that the best fit polynomial uses a much larger exponent than the ideal curve that should approximate the data. You've got to jump pretty deep into advanced number theory to get a better prediction.