So, it seems that the ‘ability check’ in the context of AD&D evolved semi-officially in the early 80s. Someone pointed out to me that the first time it was referenced in a NON-ad-hoc way was in the module I3: Pharaoh, published in 1983. So I did a little research, and these early references to the ability check involve the player rolling a d20 against one of his characters ability scores. For example, to determine whether or not a character slips on a stone surface covered with slime, he must roll his dexterity score or lower on a d20 to not slip. Thus, the higher the ability score, the better chance of succeeding (which is logical).
I’ve also been looking into other methods of making ability checks and found a few. Some of these may be from other games entirely, or were posted on various 1ed/2ed forums as house rules used by other DMs. The one that sticks out for me is the following:
Roll a two or more d6, the number rolled is determined by the difficulty of the task at hand:
Very Hard: 5d6
The player succeeds if he or she rolls equal to or less than the relevant ability score.
A perfect example happened with my gaming group during our last session. The party encountered a series of five swinging blades in a ten foot wide hallway. The blades swung all the way to the floor, and at different rates. Each blade was five feet away from the next one, so there was open space between them. We had to get by this obstacle to move forward.
The DM had each player roll two ability checks. The first was against intelligence, was considered ‘normal’ (3d6), and determined whether or not the PC could understand the pattern of the swinging blades in relation to one another, with the intent of passing through them quickly. If the first check was successful, we rolled another. This time, it was a ”hard’ (4d6) and was against our dexterity. This determined whether or not we could traverse the blades with our new-found understanding of how they worked. Failure meant that one of the blades hit the PC, causing damage. Success meant the PC was able to get through unscathed. All but one of us made it, and the one that didn’t took four points of damage… not much harm done.
I had suggested this method to the DM, and was pleased to see how well it worked and how the other players took to it. I like this method because the resulting roll will be somewhere along a bell curve, and not linear like the roll of just one die.
Here is a good site that examines the mechanics of dice rolling in RGP scenarios: