My Thief Fix

 In the words of Josh Josiah Joseph Josie “Worm Guy” McCroo, "Fixing the Thief" is like an initiate punching bowls of water in the Shaolin temple - an universally undertaken exercise that unlocks great kung fu later. 

This was said in response to my reading of Old School Essentials and realizing that the thief kinda sucked in a few different ways, depending on who you were talking to. The ways I found were twofold:

1. The thief, by dint of having “Skills” like climbing, picking locks, etc, locks off a lot of possibilities from the other classes. This is the complaint I'm mostly ignoring, because I grew up with Pathfinder 1E, a system rooted in that kind of exclusionary design. 
2. The skills suck. Simply by raw math, your average first level thief has functionally no chance to do things like detect traps. This is in part because OSR play is built around not using your character sheet and instead exploring manually through the GM and their fiat, but I say screw that. If you're giving me the tools and selling me the fantasy, then you'd better fulfill that fantasy.

Look at this. 15% to Open a Lock?

My solution for this also had to incorporate those charts, because they're in the book that I paid for damnit. This excludes ideas like the retooled thief by Whose Measure God Could Not Take or just flatly boosting the success chance.

However, I decided to draw from my beloved Blades in the Dark, and institute Tiers of Success!

• Firstly, if you describe your character doing the thing, you can often do it. If you describe yourself checking out the painting, you can find that there's a safe behind it.
• Second, if your class gives you a specific, relevant skill, you will always succeed on an attempt to do it, and you roll your chance to see if you critically succeed. This gets you some kind of bonus on top of disarming the trap or whatever, often avoiding the penalty associated with a Tier 3 check.
• If you don't describe yourself doing it, and you don't have a relevant skill, you just do a standard ability check. If you fail, you fail, but if you succeed, it comes at a cost: it takes time, a random encounter check is made, etc. 

Example time! Let's say you're trying to open a locked door. If Boris the Fighter tries to pick it, he'd make a Dexterity check, and doing so would use up a lockpick regardless. If Vadzlo the Thief picks it, he would automatically succeed, and then roll his lockpick chance to see if he gets to keep his lockpick.

This keeps the roll (because rolling dice is fun), while also making it less of a feelsbad when the thief inevitably fails his lockpick roll for the fifth time because he only has a 16% chance or whatever. 

This system does have some drawbacks. For one, Farmer Gadda pointed out that doing this would mean the thief was always the one to disarm traps and stuff. I'd personally argue that's part of the price of a system that differentiates between player characters: the Fighter is always gonna be the best at fighting, after all. For another, its a little bit of extra workload on the GM to adjudicate those kinds of mixed successes, but I personally feel like that kind of thing is fun and can effectively become its own procedure when done often enough.

Ultimately, I hope this kind of change helps players enjoy their characters somewhat more, and keep things flowing and frenetic in the dungeon with a built-in mechanism for generating complications.