[Malarkey73]

I've always thought of it in the way that if you went to an area like Yosemite or Font (bouldering) - then for every 100 people messing about, then 10 of them would be able to climb (5.12, Font 7a+) and 1 of them would be climbing (5.13, Font 8a).

Those difficulty conversion are possibly way way off and this areas may attract particularly good climbers ... but my point is the scale is logarithmic in the sense of how many people can achieve each tick level.

[justinator]

At least in bouldering, John Gill proposed the, "B" scale, which measured basically how many people were able to top the problem. So if you established a climb, it gets a grade of, "B1", until another person gets sends it "B2". This scale basically favors the hardest problems being the most important, and adjusts the grade of the climbs, as the "sport" progresses.

Gill was a Math professor, so this is somewhat interesting to the discussion. He famously thought of boulders as, "problems", of course.