The reason that no one has posted the formula is that the formula is not easy. I have spent the morning doing statistical analysis of D's data and still do not have a clue. Granted, my statistical skills are a bit rusty and the statistical software I have is several years out of date, but it is not a function of the square, the square root, or the natural log. The residuals do not plot in any helpful way. In fact, a simple plot of XP vs level shows anomalies--i.e., points that do not fall in line with the others. (I'd post the graph but I don't know how to do that.) The XP for for levels 22 and 38 are particularly out of line with the others. Either those points were reported in error or there is a second explanatory variable at play here (possibly HP or Pers?). Or see below for what I think is the most likely explanation.
Another quirk of the data is that from level 30 to 39, with 2 exceptions, the data is totally linear (first difference = 650). Not only that, but the two points that are out of line have average out to 650. Before point 30 and after point 39, the data are mostly non linear, with steadily increasing first differences--although there are a couple of places where two consecutive first differences are the same.
Also of a suspiscious nature is the fact that all of the Xp's starting with level 26 end in 0 and those above 50 end in 00. This is highly unlikely to be the result of a simple formula. My best guess at this point is that there is no exact formula--that the game just uses a table that the developers mad up out of their own heads.
alpie (I can't help it--sometimes the statistician in me takes over)