One of the things about the Old School that’s not obvious to everyone is that what you don’t write down is as important as what you do. Sometimes I refer to this as the “metarules” of the game.
For example, the “1-n on 1d6, sometimes adjusted by ability bonus” mechanic, or some variation of it, is used in several places in the rules. But never do I explain it in any general fashion, and I’m not going to. Novice GMs will apply the rule only as written, but as they become more experienced, they find that the mechanic is more generally useful than it appears. If I chose to explain it, to make it an abstract mechanic, it would change its nature.
The “unified mechanic” of modern games is like that. It becomes a hammer… as in, when all you have is a hammer, everything looks like a nail. If I codified the d6 mechanic above, it would become a single hammer, rather than a set of similar hammers useful in different circumstances. Some GMs would be less willing to change the rule, while others would try to apply the rule in places where it was not really appropriate.
But instead, I provide similar, but not identical, rules in several places. As the GM gains experience (GMXP?), he or she notices the similarity, and the differences as well. When a player wants to try something for which there’s no rule written, the GM may remember the d6 rule, and see that some variation of it is appropriate to the situation.
I’ve learned from long experience that the things I figure out for myself I know far better than the things that are prescribed to me. For the purposes of the game, I choose to keep the rules simple and brief, but provide interesting, varied game mechanics to stimulate the imagination of the GM, rather than to try to figure out every possible thing that might happen and write a rule for it. Or, to write a “unified mechanic” that supposedly covers every situation, but isn’t necessarily appropriate to all of them.
Fans of the “unified mechanic” are howling as they read this. Bah, I say to you. Before Basic Fantasy RPG, I had a game called Project 74, and the earliest versions of that game were built around a unified mechanic. It sounded good on paper… ability rolls on 1d20, with a given target number, and the ability score bonus applied (along with other bonuses, like the Attack Bonus, called a Combat Rating in the earliest versions of Project 74). But in the first player character group we had a human, a gnome, an elf, and a half-ogre. The odds of the gnome forcing a door (he was a magic-user) were around 50% (11+ on 1d20, no STR bonus) while the half-ogre had a 60% chance (same target number, 16 STR). Say what? The number of times we saw the “pickle jar” conundrum, where the weakling succeeded at a strength task that the hulking brute failed at, were just silly. Oh, I could adjust the target number, but I couldn’t get a range of results I liked with a “standardized” target number rule. Raising the target to 16+, for instance, lowers the gnome to 25%, but the half-ogre is now down to 35%; the proportion is better that way, but it’s now too hard for the half-ogre to open the door.
On the 1d6 mechanic, the gnome needs a 1 (16.7%) while the half-ogre needs 1-3 (50%). MUCH more reasonable, in my opinion, and in the last versions of Project 74 I wrote it that way.
Sometimes, what is needed isn’t an overall rule. Sometimes it’s the LAST thing we need.