The monk violates one of the core precepts of class-based RPG design: it pulls a bait and switch. Most well designed classes stake out their core concepts and remain within them. In terms of game play, playing a fighter at 1st, 3rd, or 8th level is a relatively similar experience. You're good in melee, you have lots of hit points and a good AC, and you're probably at the front of any battle. Magic-users/wizards, clerics, and thieves/rogues all have similar, consistent identities.
Some games succeed in introducing some fundamental shifts in a class, but that's rare. AD&D and BD&D gave characters access to land holdings and groups of retainers at higher level, but most (all? I don't have my books with me...) of the classes underwent that change. It wasn't a shift in how the class played so much as a shift in how the *game* played at higher levels.
Rebuilding a class's identity over the course of gaining levels is bad for a number of reasons:
- A player might like the class at certain levels, but hate it at others. That makes for a lame experience. The player either grits his teeth at low levels, or loses interest at higher ones. Note that AD&D classes that had an "initiation" phase, like the wizard, don't fall prone to this trap. Sure, a 1st-level wizard is weak, but the key is that at high levels the wizard plays mostly the same. He has more spells and more powerful ones, but it's not like he transforms into a melee monster or a healer. The class naturally improves at its core abilities.
- Balancing power at high levels with weakness at low levels is a bogus design trick. The monk's weaknesses at low levels become strengths at high levels. Chewing through hordes of weak monks at low levels is cold comfort when a high level monk rips Orcus in two. The magic-user does gain in power at high levels, but its basic weaknesses remain the same. The class simply improves its core abilities. The monk's weaknesses go away. That's why the magic-user works, and the monk doesn't.
- The monk lacks a clear identity. Is the class a martial artist that excels in melee? The 10 AC and 2d4 hit points dispute that. Is it a replacement for the thief? Maybe, but at high levels the monk replaces both the thief and the fighter. When a class shifts so much as it advances, it either starts without a niche in the game or it expands to cover more than one. In either case, it meshes poorly with the rest of the game.
In comparison, 3e fell into this trap with its item creation rules combined with the rules for wands. A low-level 3e wizard, sorcerer, cleric, or druid plays much, much different compared to a high level caster with access to plenty of long duration buffs, wands, scrolls, and potions. Whenever you see a shift like that in a class-based design, you're probably looking at deep issues with the class or maybe even the system.
It's interesting to note that the other class dropped from 1e to 2e, the assassin, suffers many of the same problems. The assassin is a crappy thief with a cumbersome, difficult to integrate assassination mechanic grafted on to it. On the other side of the coin, the bard went from a strange, optional mishmash to a core class in 2e. The bard has a pretty clear identity in fantasy novels, one that was strong enough to make it a class that cleanly severed its druid/fighter/thief heritage.
Here's what I'd do:
1. Pick a a few core mechanics for the monk and stick to them. I think I'd focus on the monk's thief abilities, good AC without armor, and multiple attacks with unarmed strikes. I'd give the monk 2 unarmed attacks per round at 1st level, but never let him go beyond that number. Damage would start with 1d6 and improve to 1d8, 1d10, and 2d6 at its highest point. A monk's AC would start around 6 or 7, improving to AC 2 at its best, and allow for Dexterity to improve it.
2. For the assassin, I'd first make the thief more of a Gray Mouser than a... whatever character the thief is actually supposed to be. I'd institute a mechanic to give the thief an equal AC footing with the fighter despite light armor, maybe some sort of active defense. I'd drop backstab, because...
3. The assassin would drop the assassination tables, but then steal the thief's backstab ability. That'd be the class's core ability, with an improved bonus to the attack roll and an increasing damage multiplier with level.
4. I'd drop the thief's abilities entirely, instead embracing an OD&D approach of using traps to challenge players, rather than a character's stats. This is one area, though, where I'd mess around with the fundamentals of the game. As DM, there are plenty of times when deciding if an orc sees a character trying to sneak past him is entirely open to my fiat. There's little the player can do to affect the situation, making the dice an ideal arbiter. I'd give monsters a perception-based defense, similar to AC, and give each class a sneak rating that improves with increased levels. Like the attack matrices, you might have a sneak matrix broken down into categories, like this:
- Thief, monk, assassin, ranger, the best sneaky characters.
- Magic-user, illusionist, druid, the average sneaky characters.
- Fighter, paladin, cleric, the anti-sneaky characters.