The new forums will be named Coin Return (based on the most recent vote)! You can check on the status and timeline of the transition to the new forums here.
The Guiding Principles and New Rules document is now in effect.
Room 12: 2d6 Goblins, 23gp, and 2 Potions
Moridan
Registered User regular
Registered User regular
I want to talk about how role playing game in general are prepped for. Read any given D&D module, and you'll see the same basic format, a room number and what's in it. That's all well and good, but is there a better way?
I've been looking into GURPS, and taking the "thinking outside the box" train of thought, I got off on how RPGs adventures are typically put together, and then started thinking about how I wish there was a way to implement something like Oblivion's radiant AI into a pen-and-paper RPG.
Now obviously the advantage a computer has is that it can randomly generate thousands of possibilities in seconds, and you don't have that luxury at the gaming table. But what about a dumbed down version of it?
Here's what I came up with. It's essentially a random encounter table tied to each location in the "dungeon". For example:
This is tagged on a hallway. Simply put, the first time a character enters or looks into the hall, you roll 3d6 to determine if anyone of the Fort's inhabitants are in the hall, and who that person is. If it is an NPC they know to be located elsewhere, reroll or declare it empty. Same if that NPC is dead.
For a room:
This simulates NPCs who are active and don't just sit in their assigned rooms waiting to be killed. Now obviously, and good GM is going to have his NPCs react in a realistic fashion to the sounds of battle down the hall or in an adjacent room, but I think this adds a level of realism when the PCs are trying to sneak into a location, and it helps the DM fairly determine who is where.
Thoughts? Do you have any ideas that would improve upon this system?
I've been looking into GURPS, and taking the "thinking outside the box" train of thought, I got off on how RPGs adventures are typically put together, and then started thinking about how I wish there was a way to implement something like Oblivion's radiant AI into a pen-and-paper RPG.
Now obviously the advantage a computer has is that it can randomly generate thousands of possibilities in seconds, and you don't have that luxury at the gaming table. But what about a dumbed down version of it?
Here's what I came up with. It's essentially a random encounter table tied to each location in the "dungeon". For example:
Area A & B - Hallway Check for encounter each time the hall is entered. Check again if characters leave the hall for five minutes and return. If characters stay in hall, check every three minutes. At night, check after an hour, or every thirty minutes. Bandit Day / Night Tannis 3 -- Joryn 4 -- Kallib 5 3 Remus 6 4 None 7-11 5-14 Farber 12-13 15 Severn 14-15 16 Aeman 16 17 Branson 17 18 -Roll Twice 18 --
This is tagged on a hallway. Simply put, the first time a character enters or looks into the hall, you roll 3d6 to determine if anyone of the Fort's inhabitants are in the hall, and who that person is. If it is an NPC they know to be located elsewhere, reroll or declare it empty. Same if that NPC is dead.
For a room:
Area 9 - Training Two standing training dummies, a target board for arrows, a rack full of lower quality swords, spears, axes, and maces. Bandit Day / Night Barnell -- 3 Dorrien -- 4 Tannis 3 -- Joryn 4 -- Kallib 5 -- Remus 6 -- None 7-9 5-11 Farber 10-11 12-13 Severn 12-13 14-15 Aeman 14 16 Branson 15 17 -Roll 1d3 times 16-18 18
This simulates NPCs who are active and don't just sit in their assigned rooms waiting to be killed. Now obviously, and good GM is going to have his NPCs react in a realistic fashion to the sounds of battle down the hall or in an adjacent room, but I think this adds a level of realism when the PCs are trying to sneak into a location, and it helps the DM fairly determine who is where.
Thoughts? Do you have any ideas that would improve upon this system?
Violence, naked force, has settled more issues in history than has any other factor, and the contrary
opinion is wishful thinking at its worst.
- Robert A. Heinlein
opinion is wishful thinking at its worst.
- Robert A. Heinlein
Moridan on
0
Posts
As you said, size would certainly be a factor. The adventure I'm testing this with is an old fort full of about 9 bandits and a dog. A fortress full of fifty to sixty creatures would only work if you grouped NPCs into "units", which is still pretty realistic.
I didn't worry about line of sight and hallways relating to rooms. My approach with this adventure was figuring out what parts of the fort each bandit would typically have an interest in frequenting. For example, the leader wouldn't likely be found on the out on the lookout platforms, and the hunters would more than likely be out hunting, it'd be rare to find them in the living quarters during the day.
Of course, this requires a lot of prep work ahead of time.
opinion is wishful thinking at its worst.
- Robert A. Heinlein
"Ugh, Kallib isn't here. Let's reset the instance."
Joe Mercenary may not be missed, but one of the main players may.
also, ineednosalt is right, this table would be more suited for an entire facility, while specific rooms, like their offices(in whatever form that may take), would have a greater chance of seeing them there. however "ugh, let's reset the instance" should be more like "let's wait for him here."
It's not much work at all after the first table. Just copy-n-paste and change the names and values as needed.
I never thought I'd say this, but I'm not looking for the result to "fit the story", I'm looking for realism. To better explain: For years I've been in the "fit the story" camp. However, I want to try something different this time. The "fit the story" play style has a major weakness, and that is predictability. More than likely, if I were designing this like I usually do, I'd make sure the leader of the bandits was the last person encountered, resulting in a "boss fight" scenario. Cool? Sure. Realistic? Maybe. Or maybe not.
The point is, the players would expect the above scenario. Or, if I switch it up on them, placing the "boss" closer to the front of the adventure, they are going to be looking for the twist. The point is, a traditional structured adventure is just that, structured, and the players are going to be second guessing the structure, and I'm going to be working to maintain it.
This way, there is no structure. This is random, and far more realistic. Will the end result be worth the effort? I don't know yet, but it should feel more real to the players, and it should allow their strategies to have a real impact that I don't have to shoe-horn into my adventure plot structure.
And I'm not playing D&D. It's GURPS.
opinion is wishful thinking at its worst.
- Robert A. Heinlein
If I wanted Radiant AI, I'd play oblivion.
That's my opinion, anyway. It seems like far too much work for something you can do yourself without dice. I get the impression the goal is to make a 'realistic' game easier, but you're putting more work in than you're getting out, and that makes it a redundant machine.
My plan for reactions is thus: Once an NPC is encountered, determine his reaction. For example, the look outs have bells they can ring. Others might just shout. Or maybe they killed the NPC with a quick assassination type attack. Regardless, once this is determined, I determine what areas might notice (adjacent rooms, hallways, etc.) then roll to determine who might be in them.
For example:
The PCs sneak around the side of the fort with an open window. Peeking in, one of the PCs notices Khallib standing by the fire. He lines up a headshot with a crossbow and drops him in one hit. Well, anyone outside the door in the hallway would hear his body hit the ground. I roll, and it pops up that Farber happens to be in the hallway. Did he hear it? Roll. Yep, he did. So then what does he do?
And I go from there.
opinion is wishful thinking at its worst.
- Robert A. Heinlein
But see, while it's more work up front (really no more than I normally do, but whatever. I tend to over prep.), it should be less work in game because I'm not struggling to maintain my adventure structure, I'm just reacting (and having the NPCs react realistically) to the dice rolls, which are bouncing off tables already set up to yield logical results.
It's interesting to me as a GM because I don't know for sure how it will play out, where as, in a structured adventure, I can tell pretty much exactly what will happen.
opinion is wishful thinking at its worst.
- Robert A. Heinlein
The "Room has X" layout allows a DM to look at their dungeon and then react accordingly to situations in the game.
The "Room might have x, y and z, roll 1d10" every hour means that the DM has to roll every time the players come to a room, intersection or hallway.
In many cases this can be a lot of empty halls and a lot of useless rolling. If the DM just knows what is in the area, they can react.
Here is the kicker...
Radiant AI is an attempt to get the computer game to be able to do the kind of thing that a DM does in a game. The Radiant AI places the enemies/NPC's at a start location, gives them an object and then says "go", it then makes choices for those things as they stimulate and are stimulated by the game enviornment. This makes the game a bit unpredictable, but only because the players/input are unpredictable.
You are trying to use a random mechanic to simulate an AI that was designed to simulate a thinking DM.
That can only end bad. You will do more work[people arent designed to figure randomness, we are designed to make decisions, which we do for others(I.E. NPCs/Enemies) very well], and the results will be less coherient and fun for your players.
You may or may not be correct. This will be an interesting test either way. Maybe I'm just looking for a way to change things up for myself, having been creating structured adventures for nearly two decades.
One advantage radiant AI has over a thinking DM however, is that the DM is usually not concerned with the creatures not currently being dealt with. They sit, lifeless, until the PCs encounter them. While radiant AI simulates a life unaffected by the actions of the PCs. Now, on one hand, how will the PCs know these dude have a "life" of their own? I'm hoping it will come through during moments of observation, when they observe but don't react.
But we'll see... maybe I just want a sort of semi-randomly generated adventure to mix things up a bit.
opinion is wishful thinking at its worst.
- Robert A. Heinlein
isnt that just like the old style random encounter tables?
DM's should be figuring what everyone in the dungeon is doing before they get there. Radiant AI is trying to similate a DM doing the job he ought to be doing, i dont know why you want an a mechanic to simulate a simulation of a job you can do easier and faster by just freaking doing it.
I dont have a name for the technique, but it goes something like this:
Determine the important parts of the plot (events/when they happen/whos involved/other key players or important figures), and keep track of that in your mind. A player take too long to try and save his kidnapped daughter? Whoop, she's dead.
However, when it comes to the people within a house/room/hallway/pub/sex shop, wing it. You're a creative person, you've come up with a plot, and plenty of NPCs, you can come up with a few average people and portray them interestingly enough, I'm sure.
Less rolling, less worry, less fuss, more game. This also avoids the "This room in the dungeon has 3d10 beholders with 1d12 gems and 2d4 pieces of art". I find that its pretty easy to get a knack for what your party can handle after a game or two, so you can wing monster encounters, too. Just have a list of things that would make sense, that the party can handle, and slip 'em in there.
Hell, even if the party is in over their heads, but it makes sense for a monster to BE there, and its not a complete aggro monster that attacks everything that comes within 100 feet of it, then throw it in there, too. See what happens (I had a low level party steal a purple wyrm egg due to some CRAZY good rolls on their part and crazy bad on mine).
Too much realism muxes up role playing sometimes. No matter what game you're running, theres something unrealistic. Fireballs for D&D, Zombies for All Flesh Must Be Eaten, Lovecraft Mythos for Call of Cthulu, Harrowed for Deadlands...even if youre running a real-world military game in which the players will be meeting perfectly human terrorists, the unrealism there is the possibility of someone who doesnt have a certain type of knowledge playing a character that DOES have that knowledge. The only way to make a truly realistic RPG is to have the players play themselves with NO fudging on stats or skills, and they encounter things like asshole teachers and boring classes, the worst of it all being something like shitty drivers on the freeway, the occasional serial killer, or scurvy.
So I come up with a variation of the traditional wondering monster tables, and suddenly I lack common sense? Ok.... o_O Thanks for the insightful comments I guess.
opinion is wishful thinking at its worst.
- Robert A. Heinlein
I never said you didn't have common sense, only that if the reader doesn't (lacking common sense being the only excusable reason for not including it in one's sessions), he or she really shouldn't be GMing.
"Wandering Monster Tables" are simply unnecessary, since they are neither as simple nor as effective as what even a mediocre GM should already have available. They're a holdover from earlier games which used randomization heavily to lighten the load on GMs and players, and are coming back in vogue as some games try to be more like their computerized counterparts. Compared to GM discretion, they're unwieldy, time consuming, and unlikely to produce satisfactory results on a regular basis.
...unless your measure of "satisfactory" is based upon how much the players are able to meta during the session by guessing how the DM has laid out the challenges.
I think it's a fair idea. However, I'm uncertain as to whether it's a good thing that players see you rolling and consulting a chart everytime they do X or Y. If the idea is for them to be caught offguard by nonconstructed nature of the adventure, it's bad because they'll catch on to the fact that much of the layout is being determined randomly. If the idea is to force them into playing from their characters' perspectives and prevent them from using meta-knowledge to make their decisions, then it's fine if they know you're rolling.
One thing that I'll bring up, though, is that you could save time in-game by making these charts and then determining the most likely routes they'll take and make those rolls ahead of time. This way, everything is still random because you're abiding by random rolls, but you're able to see what interactions are likely to occur, and then plan out some of the intricacies ahead of time.
Like, say there's a window that they can access. You roll and decide that when they approach that window, Farber is there. In case they leave to explore without disturbing him, you roll again. This time the result is nothing, so if they poke around and return to the window sometime later, the room will be empty. You think they might jump in the window to attack Farber right away, so you roll and see that both Dorien and Severn are in the adjacent hallway and will have a chance to hear that happen. What will happen in that case? You prepare their reaction ahead of time -- they might both join the fight, or one might run to sound an alarm while the other fights. If Dorien and Severn happen to be bitter foes, Dorien might take advantage of distraction to stab Severn in the back so that the party will be blamed for his death. Make their combat rolls ahead of time and see how that fight will play out if it happens.
This way, when the PCs jump in and assault Farber, they hear a gurgling cry from behind a door midcombat. After taking Farber down, they open the door and there's another one of the bandits just freshly killed. (Of course, if they don't go through that window initially, or if they do kill him but silently, none of that combat happens.) If you were using your system to roll these things as they happen, you might be hardpressed to decide on the spot whether or not Dorien carries out his dastardly plot. But this way, you have time to handcraft the initial encounters that are most likely to happen, and they're still as random as if you had rolled them during the game.
What you need is a simple solution for starting a particular part of a campaign(to get the players started), and then let them figure things out a little themselves. the railroading comes from a timeline of events that occur around the characters as they act/choose not to act.
This isnt to say "start the minute-glass, youre on the timer to figure out what youre going to do." however, there should be a balance between cinematic/major events and things that heavily involve the PC decision-making. it's great for the flow of the game.
The point is, I'm not doing this because I don't know how to do it the standard way, or because I doubt my skills in that area. I'm doing it to test out a new method. I'm not saying it's better than the traditional pre-structured adventure, I just want something different.
For me, setting up "scripted" events has become a little boring as DM, and I'm thinking that relying on a method of random generation (tweaked to give realistic results) would be a way to shake things up a bit, for my players, and perhaps more importantly, for myself.
What I'm asking for is advice on how to best set up this system. I don't need to be told that the traditional system is better. Better or not, I'm looking for a change.
opinion is wishful thinking at its worst.
- Robert A. Heinlein
Randomization isn't a very good way to discourage metagaming. Instead of trying to predict what you have planned, they'll try to predict what things you have planned as possibilities and will figure probabilities for what's going on. Being less predictable and spending less time in the metagame yourself are much more effective. Players tend to do what they're rewarded for, and if you reward for thinking in game, they'll tend to do that.
If you're just trying to spice things up, I'd keep the tables fairly large. Rolling dice is more fun when the results are less predictable, and they tend to be less predictable when there are more possible results. I'd re-roll bad results instead of declaring the room empty, since if you declare them empty, all the fun stuff will be stacked to the beginning. You may want to condense the tables by replacing the already-used results with either new ones or by expanding what's already there once you get to a certain point so you don't have to re-roll several times in a row. Definitely keep the "re-roll 1d3 times" result, since a small chance of having an extreme result keeps things from getting too monotonous. Randomization can be a fun distraction, especially if you have to make it make sense (why is he there? why are they together?) on the fly, but it can wear thin pretty quickly.
So you're getting bored with the old way? I'm going to say what I say in all the other threads like this:
RUN SOMETHING ELSE.
You're just making more work for yourself, and it looks like if you need Character X in Room B, that there is always a statistical probablility that they will never appear in that room.
In my almost 30 years of GM'ing, I've found that futzing with an existing system will generally just end up frustrating you and your players.
The thing with these sorts of games is that players will rarely, if ever, explore the same dungeon or lair twice, so they won't be dealing with the same encounters every time. This isn't true in a video game where a person can play through the story multiple times, thereby revealing how static the game world was designed to be.
If you want a bit of unpredictability, generate your encounters as you normally would, but set up variables as to when or where a certain encounter would take place.
For example:
Party A is exploring The Caverns of Monsterplace. They first enter Room B, which is guarded by a watch party made up of two Gnolls and an Orc. From there they enter a hallway which opens to Makeshift Kitchen C, Mess Hall D and Makeshift Officer's Quarters E.
Encounter table:
Room B: 2 Gnolls, 1 Orc
Room C: 1 Hobgoblin Chef, 3 goblin assistants
Room
Room E: 1 Gnoll officer, 2 Orc Watch Captains, 1 goblin scribe
Random/wandering encounters:
If the party encounters any of the creatures below, they will not appear again later. Roll 2d6 every time the party enters each room to determine if the following encounters are there. If the party does not encounter them in the room, cross that room off of the table and reroll those results on subsequent rooms.
Ogre Captain - the leader of the Monsterplace Clan.
Room B: 2, 12
Room C: 3, 11
Room
Room E: 5,6,7,8,9
Nashtoof Bonesaw - Orc butcher
Room B: 12
Room C: 5,6,7,8,9
Room
Room E: 2
This example shows a way that creates some unpredictability, but ensures that the party will encounter the special characters at some point, and puts them more likely to appear in rooms where they typically would be at (The leader in the Officer's Quarters, the Butcher in the Mess Hall or the Kitchen). You also have a general idea what to expect in each room as you have a general encounter set up, but you have added a touch of randomness to your dungeon. This makes it much easier for the GM to prepare each room (drawing up combat maps ahead of time, etc) and allows for plenty of story, but still encourages the party to stay on their toes - there is a slight chance that they will have to fight the Watch Guard party, the Ogre Captain and the Orc Butcher in the very first encounter of the scenario, which would certainly lead them to believe that the remaining battles will be quite difficult, as well.