Are puzzle ratings supposed to be like this?

@chuigda said in #4:
> When you're solving puzzle, you know there's an answer to that puzzle, and then you start carefully selecting candidate moves and then calculate them. However in real games sometimes you cannot easily figure out whether there's a tactic. You may play a move without considering alternatives, without calculating in depth, or blunder due to some tunnel-visioned calculation (for example you capture your opponent's queen but then got checkmated). Thus when you see a position in puzzle training you may be able to solve it, but in a real game you may not play the best move. Thus puzzle rating tends to be higher than game rating.

Problem is, my puzzle rating is less than game rating.

My take on chuigda's posting in #4 is that the arguments right up to the last sentence are spot on. But they don't explain why most puzzle ratings are higher - ratings can be calibrated however you want them to be, and they only show the difference in strength between players in that rating category, not the absolute value of the strength. Instead, the arguments demonstrate why it makes no sense to compare puzzle ratings with game ratings even if the two are calibrated to have the same median. Solving puzzles and playing games are two exercises which are very different in nature.