This designation makes sense when you look at the trend in XP requirements for leveling up. The vertical bar graphs below shows the XP required for each level in Moldvay D&D. Below that is a line graph showing the average for all classes of the XP required for each level advancement.
|Graph 1: Each bar represents the XP needed for each level (x-axis) up to level 8.|
|Graph 2: Each bar represents the XP needed for each level (x-axis) up to level 20.|
|Graph 3: The line represents the average for all classes of the XP needed for leveling up.|
It appears that all classes progress throught the first seven or eight levels in a relatively similar numerical pattern; it is close to doubling each time. It isn't until human characters progress beyond level 8 that the XP required increases at much greater variability. Once level 14 is reached, the increments remain consistent on average (see graph 4 below). This supports my ranges for the three categories above. As you can see from the graph below, once demi-humans start dropping out of the picture, there is a range (levels 8-14) of increased variability in XP needed for advancement.
|Graph 4: The line represents the change in average XP needed for leveling up.|
What does that mean? For one thing, either a DM boosts the available XP for mid to high-level parties, or level advancement slows down considerably past level 14. It also means that mid to high-level play for demi-humans is significantly different.
On a side note, I was also surprised to observe that it was clerics who progressed far faster than magic-users. I had always assumed (without really looking at the numbers) that it would be fighters (see graph 2). It makes me wonder if there is really a need for differences in level advancement at all.
|Graph 5: Comparing the XP needed for leveling up between Clerics and Magic-Users.|
An idea I've been toying with is to use cumulative monster XP and compare it to graph 3 (above) to determine adventure difficulty. Such a method creates a singular difficulty number instead of range. This method also accounds for number encountered and special monster characteristics, something my older method did not. A major drawback of this method is the somewhat cumbersome and time consuming calculations involved. If the monsters are arrayed in a spreadsheet, however, either method would be equally simple. Perhaps I should start including such tables in my future publications.