Originally Posted by Gorgonian
Putting the problem's rating as evidence that it is "supposed to be easier or harder" is disingenuous, too, since the ratings are a direct result of which players solved the problem and which players failed it. So if it has a lower rating than another problem, thus far it has indeed proven to be easier. That could change over time, though, as more data is collected.
It's only easier because more people randomly stumble into the right move and/or realize that the main alternative, though good, isn't good enough for +1.75. As for how hard it is to fully work out the solution - they don't even try to measure that.