Thanks.
I completely agree that there must be something that I/we are missing since it is necessary to go to a fourth degree polynomal to get the number right (+1/-1). Basically this tells me that the "(Level + 1)/2 x ..." is not the entire explanation which might be why it transtales into a crazy formula.
Problem is that I've tested the formula by eskuche and got HP 1.882 accumulated |ABS| error when calculating the theoretical versus actual HP values using the provided dataset from level 19, 20 and 21. However, although I calculated a 2 polynomal formula using trend analysis, the formula by eskuche is actually more accurate than mine, especially at 0 CON, so didn't post mine.
I've also tested the data using regression analysis, but this provided me with some wierd results that I didn't know how to interpret reasonably :-/
