Friday, April 15, 2011

Procedural Lockpicking

Arkhein asked how to make lockpicking interesting.  Here's my attempt:

Your thief knows four ways to manipulate locks:
  • Bump
  • Probe
  • Rake
  • Twist
And that's it.  She knows the wrong actions might Jam the lock permanently preventing it from being picked or opened with a key.  She also knows some locks have a Trap with bad things happening to those that set it off by failing to pick properly.

But maybe she also knows you never Bump a Dwarven lock, ever.  And you never start a brass lock with a Twist.  And maybe as she gains experience (levels) she can Bump and still recompose herself when she realizes that was the wrong thing to do.
In this picture, a correct choice will move along the grey track to the next choice.  A wrong choice will result in the lock feeling stiff (a warning) and the thief can either choose the correct action from there or face the consequences.

A DM might have a whole folder full of locks-- different difficulties, different makes. The players would not see these images, but hear a description-- "It is a heavy brass lock, with a D inscribed on it."

Update: I got all excited this morning because I woke up with what I thought was a solution to generating these easily and then I realized Zak had already proposed it in his comment. Anyway, yeah, assign a number to each of the lockpicking verbs, the "moves," roll 3d4 and the results tell you which the thief has to use.  Use dice in a color progression and you know the order the lock requires them in:
So, for that roll, starting from white we get Rake, Probe, Bump, Bump.  But that just gives us a very simple (and unforgiving) lock that assumes it jams if the thief uses the wrong verb.  To complicate this, it seems like it depends on how common you want traps to be.  On a wrong move by the player, we might then roll a 1d6 and a 5=jammed, a 6=trapped, and the other numbers indicate a move to avoid jamming: "Oh, she was supposed to Bump and didn't, now she can salvage this attempt with a Rake otherwise it Jams."

But, all this fiddling is not getting us much for the player above just rolling a 25 on d%, so it would be essential that you as DM:
  • Have locks of certain types, that players can gain expertise on.  This is contradictory to our method of generation, but if you say that all brass locks start with Probe, you could then just randomly determine the other tumblers for that lock.
  •  Give some kind of advantages with level, that would allow for some skill at the game.  Like "one free get out of a Jam," or "Reveal the move on one tumbler in any lock."
Zombiecowboy asked about Bump and Rake.  The verbs here are just abstractions, but abstractions that are conceivably related to lockpicking so you can hang your roleplaying on them (those two provisions bulletpointed above.  You could use any verbs that would help you do this in your world.  I actually needed a fourth verb and just looked at the lockpicking page on Wikipedia to find "rake." But Pry, Tap, and Spin could work just as well.


  1. Neat! Could also work for trap mechanisms. Visually, I'd have the line between say, Probe and Rake be straight or at least curved, with the "Stiff" result in a different shape. At first glance it's hard to see that "Stiff" is what you get when you follow your Probe with a Rake.

    I'm presuming anything not on the chart (say "Bump") is a no effect.

  2. Thanks Roger, My idea was that if you didn't pick Probe-Twist then you would end up Jammed. Whatever other moves you tried.

    So you probe and then get twist right and then . . . what? Anything but Rake will result in the lock getting stiff and you knowing you messed up. Then you can try again and have a (50%?) chance of getting it right. (Although I realize that last jam doesn't make any sense if there isn't at least one more choice to be made before the lock opens).

  3. i like it but i feel like there must be a way for a DM to autogenerate a lock in seconds rather than have to prepare a ton in advance.

    Like: DM rolls d4 3d4 times and assigns a proper "step" for each number. Or there should be a chart of locks...

    Ok, if you don't do it I will.

  4. Yeah, that's the flaw-- to have to prepare these. If you can make a chart or something please do!

  5. This is the coolest. I really like using this for lock picking. just had the thought that it could possibly work for combat as well...

  6. I like it, very creative. It's like a finite automata!

  7. I like this but could you elaborate on the terms bump and rake?

  8. This is really nice. I can't wait to see the tables!

  9. wow, this "branching path" method is excellent. Like Zak says, it needs a way to generate a lock in advance (xd4+1 rolls, where x is lock difficulty, and doubles indicate a branch, second die rerolled)

    I like the idea of keeping a portfolio of previous locks too, because then you could get the PLAYER as trained as the PC is supposed to be. Higher levels get a redo if they jam a lock, or maybe a free step advancement on the path.

  10. Thanks all, now go do this with traps and space battles and such.

    @Lasgunpacker, acch I was writing my update while you posted, great minds think alike. I guess once you think of the skill attempt as a simple game, a lot of the rest works itself out.

    But yeah, these, just like dungeons, could be crafted before hand, for example in the case of some elaborate Dwarven lock with various traps depending on which branch the thief stumbles into.

  11. I'd like to see how this goes in play. I'm uncertain that a whole several step mini-game is going to work for locks. It reminds too much of some iterations of cyberpunk/shadowrun when a session can get bogged down when the hacker/decker gets busy and the rest sit and watch.

    On the other hand, I've seen many OSR DMs rely on the percentage roll to pick locks or disarm traps and skip the heart of play: the interaction with the described/imagined environment. Your idea does get back to the type of play the OSR advocates.

    Myself, I've left pick locks to the die roll, but do use description and interaction for the trap on the door. So I've scoured the blogs and gaming pdfs for lists of types of doors, knobs and such. Different players respond differently. Many want to just roll, especially those who began playing with 3.x - so damn disappointing to me on both sides of the screen.

  12. Ok:

    How about this:

    Roll a d4

    An ordinary lock--most locks--only have ONE right move.

    The harder the lock, the fewer wrong guesses you can make before it jams: i.e. a 1-guess lock jams on the first wrong guess

    a 4-guess lock is easy


    Assume "bump" and "twist" are radical solutions--like dropping your ipod to fix it.

    Guessing wrong by 1 means you get to keep guessing (up until the jam limit for the lock). Guessing wrong by 2 means the lock jams.

    (Clever thieves will be able to realize what their options are after one pass according to what happens to the lock. i.e. after a "probe" that does not jam the lock, the thief will know the next thing to do is bump or rake.)

    For each level the thief has, s/he gets one "get out of a jam free" card.

  13. "get out of jam free" cards can be used once, ever, so even high-level thieves will only have a nhandful at most at any given time.

    Also, these "cards" take 20 minus (thief level) rounds to use.

    Having lock pick tools adds a number of cards equal to the gp cost of the tools divided by 25.

  14. This is great, and I'm going to use this somehow. Getting my DM feet wet by running a solo adventure with my ladyfriend using stock AD&D 2e rules. Lots of Thief-esque heists in mind and this will work to make picking a lock a lot moe interesting than rolling d%! Just need to figure out how I want this to work with 2e's thieving skill system...

  15. I just use a combination of up and down.


    If you miss one, it resets and you have to try again. The lockpicking roll is just to see if the pick broke if you get one wrong. Enterprising players will have written down the sequence as they find which ones work, to avoid excessive lockpick loss.

    This idea is very cool. I will be definitely searching for a way to implement it.

  16. Thank you all.

    So many good ideas I'm still soaking them in.

    One aspect here is the level of abstraction and another, but related aspect, is the time it takes to prepare and use this in play. A third might be feedback in the mini-game.

    Todd mentions dice, and if you got blank dice, you could write different words on it. and facilitate the speed of this.

    Zzarchov, in a comment at Zak's blog, mentioned the Mastermind boardgame which I think might work well as a prop in a single player game to help a player keep track of feedback from a lock/trap. (Maybe for you and your player Forest.)

    Feedback would be easier with the Up/Down combo you mention steelcaress. But with only 50% chance of missing a step you might want a resource built in to get spent, like the way picks break in Oblivion.

    @Zak: I'm really interested to hear how your system works with your players. I was thinking of your "types of players" post a while back. Presuming the player that can pick locks will also be one that really likes to sneak, I wonder if they'll care about messing with a mini-game.

    "get out of jam free" cards can be used once, ever"
    I like this quite a bit and would use it. Thieves' tools I would just require before they could even attempt.

    @Red: The odd fact is that I don't even have thieves in my campaign and one of the reasons is that they seem to be much more solo operators. So I'm with you that it could bog down if it took too long. I think as a DM I would try to use locks on more interesting doors/chests too.

    As an aside, I'm always hearing interesting things about Shadowrun. Which version would you recommend if I were to buy one to look at?

  17. I wrote a table that generates the name and material of a lock, as well as assigning BURP traits to six tumblers

  18. Cool, I'll take a look at it. Thanks.

  19. This is fun. As a 4e DM, I would probably use a thievery check at the beginning of the trap, the check would determine how many failures the player is allotted. Very cool, thanks.

  20. The aforementioned table is up. See here;