*Reposted from my comment on Whitehall Paraindustries*
A post I read recently on causality got me thinking about the "direction of causality" issue in gaming. I think that a lot of this discussion and disagreement comes down to people misunderstanding what the purpose of the "rules" in games actually do.
When we really boil down what game rules are for, it's very simple:
Something might happen in the narrative and you determine how likely/unlikely that is to happen.
Factors that are commonly considered are the difficulty of the action and the skill of the people involved. Add randomization, and voila - you have determined an action.
Pretty much every game rule, in all systems addresses some part of this basic equation. Rules are just shortcuts to these basic questions, and different rules address or combine these questions in different ways.
As an example, lets look at AC (or any defense rating, really). It's a shortcut for "how difficult it is to attack something", and the shortcut includes factors like physical protection and agility and overall skill of the defender.
Attack modifiers are similar - they are a shortcut that say "this is how good I am at striking aggressively".
Different systems use different shortcuts and probability structures to organize this stuff, but the basics are always the same.
It seems to me that the whole concept of "direction of causality" is mistaken - the only causality that exists is the consensual one that the players agree to. Different styles and rulesets imply causality, but they cannot create it.
Maybe an example will help, if only for myself. Let's take one action and look at the different ways that it can be handled, using different shortcuts. In game, a player says "I try to knock the monster into the pit".
You need to determine how difficult this is going to be - factoring in how tough the monster is to knock around, how skilled the player is at knocking things around, adding some randomization (if you like) and then determining the actual in-game effects.
For OD&D, this process is going to be largely up to the GM, with input from the players, and will be primarily based on AC, to hit bonus and a generous helping of "common sense", which really means deciding what you think might be realistic and then arguing about it. This is because OD&D doesn't use a lot of shortcuts.
In 4e, there are more shortcuts built into the game system. Rule of 42 gives mechanical guidelines for determining how difficult things are generally, and the player may have a power like "Tide of Iron", which is just a shortcut for saying "this character is hella-good at smashing things around by running at them".
The difficulty here is that the shortcuts are implying possibility - things like the rogue power that hits as a close burst on multiple targets, and can be done with a crossbow. Now... I've used a crossbow, brother, and there ain't no bursts with em. The power is a shortcut for saying "this character is really good at shooting a bunch of people in the face with missile weapons", but it asks you to agree that the possibility is there in the first place - which is where people who like OD&D have issues with 4e, it uses shortcuts that imply possibilities that they would rather not have.
I like 4e because it's made the mechanics of determining lots of this stuff more transparent and easier to use, but you need to be willing to use the shortcuts they built as well, and those don't sit well with everyone.