It has been remarked that the mix of puzzle here lately is unusual. As far as experts go, we had 7 last month (June) and 6 in May. We have never had a month with 7 experts until now, and the only other month with 6 experts is way back in June 2007.
Having written a sudoku puzzle generator program, I have noticed that it rarely spits out experts. In fact I had to generate about 20,000 puzzles to get 365 Experts. The most likely result is an easy puzzle with mediums being less likely and hard puzzles even less likely. In fact, I have twiddled the algorithm a little and either the hards or the experts are least likely, depending on which algorithm I use.
Thanks for the explanation.
You don't have to use each puzzle, you could throw away a puzzle if it doesn't meet your requirements.
You could, for example, not allow 2 consecutive days to have the same difficulty level, or not to allow more than two of a given difficulty in 5 days.
or any rules you care to apply to ensure a more even distribution of difficulty levels.
Just to clarify, I have nothing to do with this web site except to solve the puzzles and participate in the chatterbox chatter. I am not sure what program DrGod has for generating puzzles, but I suspect it works something like the one I wrote. If DrGod just takes the puzzles as they are cranked out by his program then his program produces many more medium, hard and expert puzzles than mine does.
As far as the distribution of puzzle difficulties, I suspect that in the universe of Sudoku puzzles, there are far more easies than any other type. Since my generator program proceeds almost purely at random, I suspect that the output it generates can be viewed as a representative sample of the Sudoku universe. In other words, easies are very common, medium, hard and expert puzzles are rather rare.