How the B/X thief gets a raw deal and what to do about it

 A while back I wrote about how the B/X fighter was underpowered relative to other classes, and offered some ideas to help bring it up to par. Perhaps even more underwhelming, though, is the thief class. 

The thief, as an archetype and concept, has a lot going for it. Tricksters, opportunists, sneaks, knaves, and rogues of all stripes fit neatly under the thief banner. There's a lot of potential for players who relish the idea of succeeding by wit, cunning, and deceit rather than magic or combat prowess. The time-honored rule of awarding experience points primarily for treasure recovered rather than monsters slain or any other metric seems practically tailor-made for thieves.

Unfortunately, the thief class as written in Classic D&D has some really glaring problems that make it a difficult and frustrating one to play.

Many have complained about the class's d4 Hit Dice, a gripe which wholly misses the mark in my opinion. Thieves are not meant to be durable stand-up fighters. When they must fight, they're best suited to the role of cheap-shot artist; going toe-to-toe with an opponent is the last thing a thief wants to do. The low HD, on par with the magic-user class and with ordinary humans, reflects this well. The thief's middling "to hit" progression also fits the role, being inferior to true fighters but good for opportunistic striking. 

The restriction to leather armor is entirely suitable for a stealth- and finesse-oriented class. B/X allows the thief access to all weapons, while BECMI sensibly restricts it to one-handed melee weapons and all missiles. Again, well in keeping with the archetype I'd like the class to model.

So much for stuff at which the thief is supposed to be mediocre. The real tragedy of the thief class is that the thief is even worse at the things supposedly within its niche. I don't think it's an exaggeration at all to say that the thief class is actually better at combat than it is at thievery, especially at low levels. Indeed, in my experience, low-level thieves contribute far more to the success of a party as missile combat support than as stealth, lock-picking, and trap-disarming specialists. A bow is easily a more essential part of the thief's kit than are his eponymous tools. 

Looking at the Thieves' Abilities table, we can see that most of the thief's skills have percentage chances of success between 10% and 20%. That's dismal, and exacerbated by the fact that in most situations, only a single attempt may be made to use a thief skill. Outside of a lucky roll or two, a thief's early career is most likely to consist of long, frustrating runs of failure. In fact, a beginning thief could easily go through an entire adventure -- or even a couple levels of experience! -- without ever succeeding at any of his trademark abilities. His chance of opening a lock is a paltry 15%; unless an adventure features a whole lot of locks, the thief is unlikely to log any success at lock-picking. At least in the case of locks, failure is immediately obvious, though. Who in their right mind is going to say with confidence, "I'll open that!" after the thief applies his whopping 10% chance to detect deadly traps to a roll made in secret by the DM? What thief in his right mind is going to trust to his 20% chance to move silently and 10% to hide in shadows to pull a daring heist against an ogre or dragon? No, it's painfully obvious that he'd be better off lending his short bow and Dex bonus to an all-out assault on the monster rather than try anything truly thiefish. The only abilities at which a low-level thief doesn't absolutely suck are Climb Walls starting at 87% and Hear Noise starting at 2-in-6 (33.3%). 

In spite of the fact that thieves are the only characters with any chance to do most of these things, they're so overwhelmingly bad at them that their value to an adventuring party is just north of nil. Not having a thief in the party can scarcely be considered a disadvantage relative to having one; in fact, more utility is gained by adding another character of any other class instead of a thief.

Another problem is that every thief, at least in the capacity of thief per se, is exactly the same at any given level of experience. Every 1st-level thief has a 15% chance to open locks. Every 6th level thief has a 45% chance to move silently. Every 4th level thief has a 25% chance to find and remove traps. The only ways thieves of the same level differ from one another is in their hit points and their ability score modifiers, none of which has any bearing at all on any thief ability. At least fighters vary a little bit in their fighting abilities based on their ability score modifiers.

The thief tables are also needlessly extensive, given that the difference between any two skills (excepting Climb Walls and Hear Noise) never exceeds 10%, and in fact the numbers tend to converge with increasing level. 10% may not be completely insignificant, but is it worth having a table with seven different columns for seven different skills when five of them vary by so little? 

Finally, there is the thiefly talent for striking unnoticed with a bonus to hit and double damage, which has a fair chance to one-shot an opponent of 1-3 Hit Dice but quickly falls off as the party starts dealing with more powerful monsters. To be fair, other classes (except perhaps the magic-user and elf) don't receive big damage bumps as they level up, but the thief's special attack has a much narrower niche than attacks in standard combat, requiring special opportunities to use, and each opportunity is generally one-shot only, succeed or fail. It isn't an unreasonable expectation that it should continue to be useful as the thief gains in experience levels.

So, how do we fix this quagmire? To begin, the percentages clearly need a boost, and this can be done without fear of making the thief too powerful, and while still allowing plenty of room for meaningful improvement as characters gain experience. Furthermore, we can easily condense all skills into a uniform scale of advancement. That is to say, there's no need to have different percentages for Open Locks, Move Silently, and Climb Walls; we can use the same scale for everything. (Climb Walls, in particular, needs a little tweaking to make it viable with lower chances of success, but I've taken a shot at that, too, in the link below.) You might object that this makes thieves even more uniform in their abilities, and so far you'd be correct, but hear me out...

Let's give individual thieves the opportunity to be especially good at one or two skills. We can do this by creating a table with two columns, one for Standard (or Average, or whatever) skill progression, and one for Specialist, with the Specialist column featuring a better starting percentage and faster progression. The easiest implementation would be simply to allow a thief character to choose one skill in which to specialize, with the remaining skills using the Standard progression column. If we feel so inclined, we could include a third column with progression somewhat inferior to Standard (labeled Basic or Below Average or whatever), and allow a thief to choose a second specialty in exchange for reducing another skill to the inferior progression. That gives quite a bit of range for thieves to be different from one another, but at the same time reducing the table to a mere two or three columns instead of seven. (Also happily avoided is the bookkeeping and number-crunching headache of the AD&D 2e point buy scheme.)

Finally, the special surprise or ambush attack can be easily scaled up by adding a die of damage at every fourth level, beginning with two dice at level 1, three dice at level 5, four dice at level 9, and so on. I personally like using d6 as the base for all such attacks, regardless of the weapon used, making daggers and clubs viable mechanically as well as aesthetically. 

If you're interested in seeing how all this might come together in practice, I've written up a complete revision of the thief class along these lines, including a full revamping of the thief skills themselves. Doubtless it could be further streamlined, and I may undertake that task at some point in the future, but for now it is what it is. Feel free to use or modify it however you see fit, even if only to crib a few ideas for your own house rules.