Skip to content

Questions and opinions about Pick Pocket

Does anyone know the exact formula for Pick Pocket success? It's certainly not a straight percentile die roll. Even at very high PP levels, failure is quite common. It's not just because of critical failures either (is that even a thing in this game?), since even at a skill of 180, it's not uncommon to see a failure rate of 20% or so.

It [i]seems[/i] to be related to the victim's level, and maybe class. I've noticed that commoners have a failure rate that more or less matches a straight percentile roll, but things get weird with other adventurers, and merchants that have adventurer levels. Thieves seem to be much harder than wizards and fighters from the same area, but I haven't taken the extensive data that would be needed to confirm this.

So anyways, what's the formula? I can't find anyone talking about it, or even acknowledging that it's not a simple percentile check against your PP skill. Just to avoid misunderstanding, I am specifically [i]not[/i] asking for the minimum thresholds needed to steal different items from different slots. I know about those and that is not what I am asking about.


Related question: The consequences of failure are stupidly high, worse than getting your thief killed. A failure basically breaks every merchant and every quest in the whole map. It's completely disproportionate, and totally at odds with the way NPCs react when you get caught stealing out of a drawer or chest. Pick Pocket has zero place in a zero or minimal reload game, which is a shame. Does anyone have their own homebrew rules for reasonable consequences? I'm not asking for a mod, I'm asking for guidelines that could be enforced via quickload and EEKeeper.

Comments

  • DhariusDharius Member Posts: 654
    edited May 2016
    Can't speak about the "formula", but are you sure it makes everyone on the map hostile to you? I thought it only applied to people within line of sight, so it might not necessarily ruin the game...depends on who and where you try to steal? Part of being a decent thief is making a stealthy retreat if things go a bit wrong, after all.

    It might actually be a "good" thing if you like incidents that reduce reputation, for some reason...
  • Clumsy_DwarfClumsy_Dwarf Member Posts: 112
    It is a percentage chance. The better loot you can get, the harder it is to succeed. Some targets, like Ribald, you need that 180-200% to have a reasonable chance but a commoner in the slums, not so much. Just remember, there is always a 1% chance of failure.

    So not quote me, but I think there is a bit of a bonus when you do it from stealth.
  • FringewareFringeware Member Posts: 29
    I appreciate the responses, but...

    I carefully laid out that I know some targets are harder than others. I am asking for the formula, not the qualitative effects, which I already know and already stated that I know. What affects the odds? The level of the victim? The thief's level? Does class matter? Is it the pickpockets score of the target? I'm not at all convinced that weight has any effect (it clearly has an effect on shoplifting, but after stealing a heavy bastard sword 30 times in a row from guard with a PP of only 95, I 'm convinced that weight doesn't matter when actually picking pockets). But if weight does matter, by how much? Surely someone knows the formula.
  • semiticgoddesssemiticgoddess Member Posts: 14,903
    I summon @Blackraven, who knows a lot about thieves.

    To answer @Dharius' question, turning neutral people hostile initially only affects people nearby, but the red circles can spread all the way across the map later on. If you turn one commoner hostile and no one else is around, no one else will turn hostile. But if that commoner starts running away, which he almost assuredly will, then everyone in his path will also turn hostile.

    You can contain the damage by charming hostile innocents and quarantining them in remote corners of the map.
  • mf2112mf2112 Member, Moderator Posts: 1,919

    I appreciate the responses, but...

    I carefully laid out that I know some targets are harder than others. I am asking for the formula, not the qualitative effects, which I already know and already stated that I know. What affects the odds? The level of the victim? The thief's level? Does class matter? Is it the pickpockets score of the target? I'm not at all convinced that weight has any effect (it clearly has an effect on shoplifting, but after stealing a heavy bastard sword 30 times in a row from guard with a PP of only 95, I 'm convinced that weight doesn't matter when actually picking pockets). But if weight does matter, by how much? Surely someone knows the formula.

    Not sure if this is still accurate, looks right though.
    https://forums.beamdog.com/discussion/22203/pick-pocket-changes
  • FringewareFringeware Member Posts: 29
    Those are the threshold values I was talking about in the OP. Those don't have anything to do with the odds of success or failure, they just set the minimum PP score you need to be able to even try to get those items. The hardest item on that chart is the weapon slot (95 prereq), but I can steal a sword froma guard at near 100% success with a PP 100 thief. Meanwhile, the same thief fails at about a 85% rate when trying to take a simple potion from Silke, and the failure rate is over 50% when trying to get "easy" objects like gems from certain NPCs. I want to know what is causing that, get some numbers, and be able to make informed choices about how high you need to get your PP.
  • magisenseimagisensei Member Posts: 316
    Personal rules for a more difficult play:

    1) no potion stacking of the same potion (eg potion of master thievery) - you can use the different type of thieves potion but you can't use the same one - this really limits what you can steal/pick pocket.

    2) No stealing and reselling to the same merchant

    3) No reselling the same item over and over again - you can only do it once the only exception are +1 magical weapons.

    4) no saves and reloading to try again and again

    5) if you fail you can reload but you cannot try again to steal again since you failed the first time - this allows you to play without the difficulty of hostile merchant (and everyone else)

    As poster said the consequences of failure is so incredibly harsh in BG2 that it is out of whack - how they implemented getting caught in BGEE was better - you can bribe your way out and lose 1 rep without being hunted down - although you do lose that merchant as well (no buying or selling to that merchant again) and screwing up your entire game.

    BG2EE should make stealing less harsh in terms of failure so it does not screw up your game - but make it such that it still has serious consequences -e.g. no buying from that merchant if not forever it must be for a good length of time and even then have the PC pay 100% or more for the privilege of buying again from that merchant - and of course a reputation loss (at least 1 or 2) - and maybe all merchants (regular not thief merchants) should charge more or something if this happens.

    Stealing from thief merchants should result in a fight or perhaps the option for a huge bribe - if it degenerates into a fight it should be with overwhelming odds against you and as such force you (the PC) to pay reparations for stealing - so those that will want to kill a lot of thieves will of course not be able to win such a battle and have to pay lets say 10,000+ gp to make amends - or maybe the thieves take someone hostage until you pay.

    In BG1 - when you caught by the Flaming Fist (during the last chapter) - they kill a member of your party just like that too show you that they can do it - so with thief merchants there should be something like that implemented for those that want to battle it out.

    Make the consequences of stealing and getting caught harsh but it should not be game breaking such that you need to reload for something that should not be game breaking.

    Stealing in BG2: One thing I noticed about stealing in the game is that the more expensive it is the harder it is to steal. This of course should NOT be a surprise just look at how regular stores protect their goods - expensive items are protected while cheaper items are less protected and inexpensive items almost have no protection. So it should not be a surprise that stealing an arrow is possible with a low PP but stealing arrows of dispelling is more difficult.
  • joluvjoluv Member Posts: 2,137
    mf2112 said:

    Not sure if this is still accurate, looks right though.
    https://forums.beamdog.com/discussion/22203/pick-pocket-changes

    Equipped weapons can't be pickpocketed anymore, so that's at least not totally accurate.
  • FringewareFringeware Member Posts: 29
    You can steal weapons from quickslots if they aren't equipped. I do it all the time. There are these guards in Nashkell that are shown carrying bows, but they also have swords. If you fight them, they can swap if they want to. They lose this ability to swap if you swipe their swords first, which makes sense. The threshold of 95 is also accurate. I've tried it at 94% and just got the generic "there is nothing you can steal at your skill level" message, but once you hit 95, BAM!, it's bastard sword city.
  • FringewareFringeware Member Posts: 29
    edited May 2016
    Okay, I did so quantitative testing, and PP success is affected by the target's own pick pocket skill. The victim's level has nothing to do with it, except in the case that it might give them more points to spend on pick pocket.

    After 600+ trials, I have enough data to tentatively conclude that the formula is simply the victim's PP subtracted from the thief's PP. The victim gets his entire PP skill, meaning that any racial or dex bonuses are counted, and these bonuses are applied even if the victim has no actual points spent in the skill. I can talk about the data collection process if someone really wants to hear about it.

    I can also confirm that the critical failure rate is 1%.

    Here are the racial mods, for reference:
    Humans/Dwarves/Gnomes: 15
    Halflings/Elves: 20
    Half Elves: 25

    Here are the dex modifiers up to Dex 20:

    16 or less: 0
    17: 5
    18: 10
    19: 15

    A Pickpocket skill of 140 should give you 99% success against any non-thief in the game.

    EDIT: Actually, it looks like only Bards and Thieves benefit from the race/dex mods, which makes sense. 99 points in PP is good for most of the game. I've looked through the cre files, and almost every thief and bard have no points allocated to anything. That's disappointing. Ravenscar is a 15 thief, but hasn't spent any of his points, and gets 99-percented by a PC with even modest skill (Ravenscar has a Dex of 3, which is bizarre, and his total PP skill, with race mods, is 0). This means that 140 is enough to pwn any thief in the game, unless there's an outlier that actuall yhas his/her points allocated.

    I remember people like Ribald being really hard, even with stupid high PP, but I can't find anything in his CRE file that would be causing this. I don't have an active BG2 game right now so I can't test it.
    Post edited by Fringeware on
  • Lord_TansheronLord_Tansheron Member Posts: 4,211

    I remember people like Ribald being really hard, even with stupid high PP, but I can't find anything in his CRE file that would be causing this.

    My Ribald at least does have 95 PP in his .cre data (as well as 19 DEX). I do run with mods, though I'm not sure they would affect that.
  • FringewareFringeware Member Posts: 29
    edited May 2016
    Huh, my BG2 install is completely unmodded. I haven't even played it yet, still too busy exploring BG1EE. I have no idea what could cause such a specific change. Now I'm wondering if the blank skills for the npc thieves in my bg1 install is some kind of bug.

    What class is your Ribald? Mine is a single class mage.

    EDIT: Poking through the CRE files, and man are things inconsistent. Most NPCs have zero PP, even if they logically should. Zhurlong has a PP of 60, which makes him one of the hardest marks in the game, but he has nothing that you can steal. Bentley Mirrorshades has the class [Innocent], but packs a 60 in PP. His wife, Gellana, is also [Innocent], but has no PP. The hardest mark I could find in BG1 is Naren Darkwalk, at a PP of 65. Ravenscar is super weird. He's the highest level thief in the game, but he has no skills and all of his stats are 3. What the hell?

    Anyways, there are maybe 4-5 NPCs with any kind of PP protection at all in BG1, which means that a skill of 100 should be fine. For the harder targets, it looks like 160-165 is needed for minimal risk. Potions of Master Thievery+ Perception will get you +60%.

    Not sure what's up with BG2. If Ribald really has Dex 19, PP 95, and is a half-elf, then his total defense is 130, which is consistent with what I remember from playing BG2 way back when. A quick check shows my Aran Linnvale with total PP of 130 as well, so he's not the only one, and my CRE files aren't completely borked.

    PP 130 means you need 230 to be 99% safe, which uh, yeah, not worth it unless you really, really like the Ring of Regeneration.

    I would love it if other people would contribute data and we could be confirmation/refutation. Incredibly, I think this is the first time in BG history that anyone has examined the mechanics of Pick Pockets, and I'm sure lots of other people would like tfis information as well.
    Post edited by Fringeware on
  • mf2112mf2112 Member, Moderator Posts: 1,919
    edited May 2016
    Last time I was there with yoshimo I had him drink three master thievery potions and I got the ring first try. I then had to reload a bit later but also got it on the first try again. I don't ever recall failing more than once after drinking three before at that point.
  • FringewareFringeware Member Posts: 29
    Oh yea, with regards to magisensei:

    I like your rules and they are similar to what I was considering. My current scheme is that if I fail a PP check, I get to reload, but I have to lose 1 Rep and pay 200 gold, just as if I was caught stealing from a container. I also can't pickpocket that character anymore.

    It helps a lot now that I know what the odds are. Pickpockets was a black box with wildly inconsistent risks and absolutely zero feedback on why some things were easy and some things were hard. I can now decide when to risk it, when not to, how many points I actually need, and when the right times are to pop those potions.
  • abacusabacus Member Posts: 1,307

    Oh yea, with regards to magisensei:

    I like your rules and they are similar to what I was considering. My current scheme is that if I fail a PP check, I get to reload, but I have to lose 1 Rep and pay 200 gold, just as if I was caught stealing from a container. I also can't pickpocket that character anymore.

    It helps a lot now that I know what the odds are. Pickpockets was a black box with wildly inconsistent risks and absolutely zero feedback on why some things were easy and some things were hard. I can now decide when to risk it, when not to, how many points I actually need, and when the right times are to pop those potions.

    If you ever get the mod bug (I know you said in the OP that you weren't there currently), Rogue Rebalancing does some nice things with PP... If you get caught, you essentially get a telling off from the "victim" and a chance to talk your way out of it subject to a stat check. If you get caught again in the same area then it's rumble time.
  • BongrizBongriz Member Posts: 6


    After 600+ trials, I have enough data to tentatively conclude that the formula is simply the victim's PP subtracted from the thief's PP. The victim gets his entire PP skill, meaning that any racial or dex bonuses are counted, and these bonuses are applied even if the victim has no actual points spent in the skill. I can talk about the data collection process if someone really wants to hear about it.

    I know it’s a while ago, but I for one would love to hear about the data collection!
Sign In or Register to comment.