[mycotheologist]

I was assigned a computer based research project for college and the objective, as far as I know, is to find the optimum method and basis set for a particular class of octahedral chromium compound. He (the professor who assigned me the project) told me to start with Hartree Fok, then try the electron density functional methods in order to see if there are more effective methods/basis sets for doing calculations on this particular class of compound (an octahedral chromium complex). The results I've gotten so far are confusing me. I'm new to Gaussian so there are a few things confusing me. So from what the professor tells me, the list of density function methods in GaussView (B3LYP, B3PW91, PBEPBE, MPW1PW91, HCTH etc.) are ranked in order of least complex algorithm to most complex, meaning that as you go down the list, the methods require more computational power but yield more accurate results. He says the same thing goes for the basis sets (6-31G, 6-311G, LanL2DZ, SDD, DGTZVP etc.). So if thats the case, then the most accurate method and basis set combo should be HCTH/DGTZVP. I ran optimisations on the prototype compound and tested every method and numerous basis sets for each method and then compared the optimised bond lengths to the actual bond lengths which were determined experimentally and for some reason the most accurate method/basis set combo is B3LYP/6-311G. Am I doing something wrong? I can definitely conclude that density functional methods produce more accurate structure optimisations than hartree-fok but I don't get why B3LYP/6-311G seems to be yielding the most accurate results.