For the determination of nickel in waste water a calibration
line is made.

The analist makes soultions with concentrations of respectively
0, 0.1, 0.2, 0.3, 0.4, 0.5 and 0.6 mg/l nickel.

Then he measured the extinctions of the solutions. He measured:

concentration nickel  |  extinction
----------------------|--------------
         0            |     0.064
         0.1          |     0.15
         0.2          |     0.26
         0.3          |     0.32
         0.4          |     0.43
         0.5          |     0.48
         0.6          |     0.53

Now he will calculate the best funtion trough the points.
He used the first orer polynom and shows the residue plot.
In the residue plot is to see that the first two residues
are lying below 0. He decided to use the second order
polynoom the residue plot shows now a random plot.

If you use magic numbers for polynom regression look always
to the residue plot, it is important that the residues are random.

In the future Magic numbers make ik possible to use the Goodness of Fit
test. With this test the significancy of a model can be indicated.