James Maliszewski has asked about advanced old school games. His basic contention is that old school does not equal rules light. While he is true in chronological terms, and even uses the obvious example of just about anything published by Fantasy Games Unlimited there is a problem with including these games in the modern old school cannon.
The modern old school definition seems to have settled out onto three basic pillars. They are do it yourself, rulings over rules, and player skill over character skill. You can arguably add a four pillar which is literary sources over gaming sources. Advanced games strike directly at two of these, rules over rulings and player skill over character skill. The first of these should be fairly obvious while the second may not be.
To briefly cover the case for complex rules emphasizing rules over rulings you have to ask why more complex rules exist. In general they developed for two reason, which are somewhat interlocking. The first was realism, more correctly called verisimilitude. This was a major concern across the wargaming community throughout the 1970s and when wargamers began creating roleplaying games the concern crossed. Among readily available games today that date to the period the classic example of this is Rolemaster, most specifically Arms Law. The second reason is to cover all possible cases. The full Rolemaster system is also an example of this. It has twenty armor types, tables of modifiers, and hundreds of skills to cover every possibility in action and characters.
This last point shows how this starts to attack the player skill over character skill paradyme. This is really an extension of the process that brought up the theif. Complex games, with their breakdown of character abilities into finer and finer points as well as creating more situational modifiers encourages players to optimize against the rules set as opposed to optimzing against the situation. When you have a simple game the player’s do not have a laundry list of possible advantages and have to interact with what the judge describes. When you do have that laundry list players are encouaged to listen to the description through a filter created by that list.
At this point I have to ask, why is D&D3 not old school? I don’t think it is, but the only existing pillar it seems to really violate that complex older games like Rolemaster do not is the literary influences over gaming influences. Yet, I think most people will agree third edition was not old school even if it was recognizible as part of the family.
I think the difference, the fifth pillar, is up front choices versus continuous choices. In old school games your upfront choices were much more limiting than in new school games. Rolemaster opens every skill to every character and there are no per level forced abilities. In fact, leveling in Rolemaster beats the complexity of any third edition I played (although I quit before the insane prestige class prep that seems to have characterized late play based on my reading). Yet, a Rolemaster character is much closer to a pre-third edition D&D character in terms how his class defines him. He won’t multiclass, has no prestige class to graduate to, and will pay a high penalty for going outside his core skillset in terms of overall abilities. While you might emphasize character skill more that player skill in more complex old school games that skill is still more narrowly defined. When you choose to play a fighter of some kind you are playing a fighter at level ten, not a fighter/sorcerer/cleric hybrid who is going to join the Order of Draconic Lightbringers for some weird capabilities. Sure, your fighter is a two handed weapon specialist who knows ten psychic healing spells while their fighter is a bowman who also mastered savate for when the drunken dwarf calls a bar fight, but fighting is still their defining characteristic.
What really interests me is how we can apply this insight to make D&D3 a much more old school game. Perhaps d20 is the jumping off point for the advanced old school James is looking for.