Following on from my last post, here's the continuation of my explanations for my house rules:
Magic-users (and other arcane classes like Elves) prepare spells rather than memorising them, and don't need to have their spellbook to hand to prepare spells that they know. The spellbook is only required for learning a new spell (scribing it into the book is a necessary part of the learning process) and for any magical research; the spellbook is a workbook and notebook rather than a reference work. Scribing a spell from another magic-user's spellbook takes days at least (if not weeks) of uninterrupted work, making trading spells a risky measure of trust since spellbooks are very time-consuming to replace.
Magic-Users and Elves start knowing Read Magic, one randomly rolled first level spell, and one chosen spell at both first and second level. The second level spell can't be used until level 3 of course; it's advance study in anticipation of increased magical skills.
REASONS: This slight variant on spellbooks means that Magic-Users don't have to carry them around everywhere. I like this for several reasons. One is that since PCs don't automatically get an enemy MU's spellbook after killing and looting him, they don't acquire so many new spells so quickly and easily. Another is that it's not so hard to keep one-of-a-kind unique NPC spells out of player hands. Also, if a player MU loses his book, he's not so useless after a single day.
The rule for known spells for a starting character is similar to the standard rules, except that a PC gets a free randomly rolled level 1 spell before choosing a level 1 spell. This helps MUs be a little more versatile and encourages some of the less commonly used spells into play.
Before making a melee attack roll, a player can choose to make a power attack, taking a 4-point penalty on the attack roll to gain 1d4 to the damage if the attack hits. Stronger power attacks are possible: -6 to hit for +1d6 damage, -8 to hit for +1d8 damage, -10 to hit for +1d10 damage, -12 to hit for +1d12 damage.
REASON: I like the idea of the power attack system from 3e, but a 1-for-1 trade from to-hit to damage would be too potent in Basic (especially since the ability doesn't cost a feat). This system makes the exchange about 2-for-1 on average, but does allow for a power attack to hit extremely hard on a good enough roll, which helps give it the feel of a do-or-die gamble.
Players roll a "defence" against the attacks made against them instead of the DM rolling for the attack. The DM declares the number of attacks (splitting by type if applicable, e.g. claw/claw/bite) against a PC. The player rolls his defence against them: a d20 plus his AC, and declares the result. A lower result is better. If the result is equal to or higher than the attacker's THAC0 (including all modifiers into the THAC0 score) the attack hits, otherwise it misses. A natural 1 is an automatically successful defence, and a natural 20 is an automatic failure, suffering a critical hit (unless the attack is of a blunt type).
Note that this is a procedural change rather than a rule change: it's mathematically identical to the standard method of the DM rolling for the attacks.
REASON: As a player, I always used to prefer combat defence rules where I got to roll dodge/parry/whatever to ones with a D&D-style passive defence like AC. It took a while before I realised that if the chance of being hit is, say, 50/50 by a given enemy, it doesn't matter whether I'm rolling for defence or not – it's 50/50 – it just seems like I have more control if I roll the dice. So, this was the initial reason why the THAC0 Defence system appealed to me when I read it in the Dragon Magazine Archive (issue #177 page 24, for those who have it). The second reason was that one of the things I always found awkward about various versions of D&D was keeping track of PC ACs. Having to keep asking players to remind me of their AC, or realising after a while that I had one of their ACs recorded wrongly and I'd had them hit when the attack should've missed; it's a bane of mine across editions. With THAC0 Defence, the players are the only ones who need to know their own ACs, which helps them and me.
The Thief class gets 1d6 for Hit Dice.
Clarification on the scope of Thief skills: Starting percentages may be low, but they represent the chance of exceptional feats. "Move Silently" is the chance of complete silence, which isn't always needed - moving quietly is usually enough to give a chance of surprising an enemy, and that doesn't require a roll. "Hide in Shadows" is the chance of remaining unseen in direct line of sight with only shadow for concealment - having real concealment can improve the chance or give automatic success. Players may be able to find traps (and find ways around them) by experimentation and cleverness, without having to succeed at "Find and Remove Traps" rolls.
REASONS: The HD increase helps Thieves do damage under the “class-based weapon damage” rule, and I think it's good for it's own sake. Thieves have weak armour making them almost as vulnerable as MUs, but are more likely to get into combat (especially if they do any forward scouting or try Backstabbing), and in many groups are less likely to get protected like the MU would. I think Thieves are too weak per the standard rules, and increasing HP seems a good way to boost them (especially since it can help them survive any traps they set off accidentally, so it aids them in their class role).
The interpretation of Thief skills is another thing I've found on the blogosphere. It helps to keep the low skill percentages of low-level Thieves in context and suggest ways to get round their mechanical limitations (i.e. using descriptions to resolve thiefly activities so the player doesn't have to roll for those low percentages).
TWO WEAPON FIGHTING
A character can wield a one-handed weapon in each hand. Attack and damage modifiers are based on the characters' Strength or Dexterity modifiers - whichever is lower - and on a successful hit, the player gets to roll damage for both weapons and choose which one hit (this option does not add multiple attacks).
REASONS: Dual-wielding is a tricky beast to make both balanced and simple in D&D. My experiences in 2e and 3e are that adding more attacks for dual-wielding is either unbalanced (2e) or clunky (3e). Rolling damage twice and taking the better one is simple, and balanced because it's exactly the same as using a single two-handed weapon (in my version of “class-based weapon damage”). The limit of using the lower of STR or DEX means that classic brute-force fighters are going to be better off with the two-hander, which fits the flavour of dual-wielding and ensures that not everyone will dual-wield (2e failed badly on both counts); dual-wielding has its own advantages at high levels when magic weapons with special effects are common.