You happen to be privy to some ghost stories told by locals at the tavern. "I swear I saw this beautiful girl standing in the middle of the road. When I approached, she just disappeared, like she was never there at all! They say that her ghost appears once a year on the anniversary of her death."

If you happen to figure out the right day and approach, you'll find her showing you something, in Stir of Echoes style. Maybe it involves a little bit of murder mystery or dungeoneering.

Sure, this is a story plot that sounds like it could be the beginning to some off the shelf megadungeon, but it doesn't have to be. It could be more like a broom closet by comparison, and completely ignored by players without affecting anything. I grant you that it would be something preplanned to throw into a more open-ended world, but if you have a plethora of these adventure seeds (whereby you can add many of the monster and treasure details on the fly), then they might add to rather than detract from the game. 

I've been collecting these on my netbook for a while now. Check out this site: http://www.lythia.com/2008/05/short-adventures/
http://www.lythia.com/2008/01/review-pottage/
Or this list of one page dungeons:
http://critical-hits.com/2009/05/25/the-compiled-list-of-one-page-dungeon-entries/

I've been compiling some of these into rumors, jobs, and random encounters. My goal is to be able to have an NPC who presents the information be rolled up like a random monster. The information may then be less of a baited hook, and more of an unfinished puzzle.

The downside to all of this is how much work it adds to the game planning.
                

                Infinite monkeys will eventually create Shakespeare, but do you want to watch it in real-time?

Playing the Sims is fun, but it is a lot better because you fast forward through the boring parts (like sleeping).

As the DM, you have the power over time and probability.  As long as the dungeon passes the "sniff test" of being possible within the simulation, who are the characters to doubt that they have fallen on Hamlet by chimps?

You could create a thousand random encounter tables that carefully balance the likelihood of what will happen to the characters everyday.  But the odds of a really exciting adventure start to look like ape Shakespeare again.

I guess my assumption is that a true simulation of any real time period would be very boring for adventurers or lead to a quick death.  Don't you think 99% of level 0 NPCs would run away and hide from what must look like a gang of Hells Angels (the PCs)?  And 99% of the police will be shooting first (as will the leveled bad guys).

The game is a fantasy.  The DM is allowed to manipulate the perceived reality.  I think it feels much more true when those manipulations are done at a macro level rather than a micro one.  (It is much more immersive to have a storyline that allows the players to meet Caesar - a very low probability event - than it is to fudge on die rolls in a battle.)

BOTTOM LINE:  Your Roman sim is deeply immersive.  It is working well.  It can support the "dungeons" needed to make playing and running the game satisfying.


                

                I agree with where Jim is going.  You have a great advantage here with the basis that most dungeons that take place in buildings are already pretty much laid out.  A villa or insula is not going to vary much in layout from place to place.  So you don't need to worry about creating a random house from scratch.

I would say having three or four different layouts for each floor based on typical period floor designs would give enough variation for mix and matching to create the layout you need.

Placement of monsters might be a bit more difficult.  But I am sure a general idea of how you would make fortification for hideouts would be good with a few variations in traps, large low lvl numbers, or fewer higher levels as the initial barrier would work well also.

I think you will need to look at how instances are being done for MMORPG and what their good and bad points are to help you with this issue.
                

                You're the DM.  Make it so.  Who are we to say, "This wouldn't happen this way."  Just put it there and make it up later as we discover the clues.  I don't think we'd care too much even if the explaination was a cliche'.  Aren't the 'Red Tunics' just that - a cliche'?

Besides, just trying for a simulation game would exclude all the: races, magic, monsters, etc.  Not fun, or AD&D.  You have to have some things prepared for us to find, shouldn't you?

                

                I've been mulling this over this morning.  It seems to me that if you have a rumors table, as Nick suggests, each of those rumors (or some significant %) ought to have a mini (or maxi) adventure associated with it; Pre-generated or conceptual that will be filled in later.  Each of the rumors basically becomes a set of "Railroad" tracks we can choose to hop on or ignore.  This is not making it into a story driven game since it is up to the players to decide if they want to take the set of "tracks" or not.  But the DM needs to be ready and have a variety of them for us to choose from.  

So, in last Sunday's game, you provided me with three rumors that I overheard while eavesdropping from the boat, two of which seemed worthy to write down: 1) Tertus is unhappy with the Tribune, and 2) Ruga Magnificus is a snake oil salesman.  Do you have an Idea where these rumors will lead?  Do you have Ruga conceived as an NPC?  Why is Tertus unhappy with the Tribune?  I think these are things the DM should have in mind before the rumor is disseminated to the players.

I feel that the game would benefit from some minor Railroading when it comes to it.  Not that we have to continue to follow the tracks, but tracks should be built.


                

                
Thanks for the comments, everyone.  It's incredible that I got this many in such a short period, so allow me to recover from my shock.

There.  All better.

Nick, I see where you're going.  It's a good method for a very open sandbox game where the players are kind of thrashing around with no particular plan.  However, this smacks of something that bugs me about MMORPGs and that is the guy with the exclamation point floating over his head, "I start a quest!"  This is fine for a computer game, but I don't like it in the tabletop experience outside of a tournament/convention setting.  I feel that it limits the experience.  That's not to say that I don't do this to some degree in the game, but every rumor does not lead to an adventure.  Or maybe it does.  :-)

Jim, you bring up some excellent points and thank you for commenting.  I'll take your questions in order.
1. Monkeys are hilarious and disgusting!
2. Much of the game so far has been the organic development you're alluding to here.  I don't have a plot in mind.  I presented a setting and the players have their characters interact with it.  As the game has progressed, mini-plots have arisen as they must in order to give the people that populate the world motivations and personalities and depth -- to give them realism.  However, much of what I have put in place has been random, and I plan to continue that going forward.  I find that the surprise factor is a major joy button for the group and making a great deal of stuff random contributes to that.
3. Nearly all Zero-Level NPCs run like hell at the first sign of trouble.  However, there are criminals and heroes everywhere.  You can figure that about 3% (3 out of every 100 people) of the entire population is leveled.  Some may even have better stats than you.   That means that for every Collegium, there is probably one leveled person and that gal/guy is probably the leader.  I've done some calculations for average level by social class, but that work isn't quite finished yet.  Suffice to say that there are plenty of people who aren't afraid of you, although after your last two encounters, the word is probably going to be getting out on your merry band.

Jason, thanks for commenting!  I really want to get away from the A-B-C quest mentality of MMORPGs.  I realize that to a certain degree I can't because that's just how adventure games work, but I've been driving for simulation over game where I can.  I think it's working, but we're coming to a point where although the players will continue to be attacked by pirates and street gangs as circumstances dictate, in order for the nifty fantasy aspects of AD&D shine we're going to need to introduce something fantastic.  I was looking for a way to get to this without putting together an MMORPG-style dungeon (I start a quest!).

Thom, thanks for commenting.  You always give me a lot to think about.  First off, you guys say, "This wouldn't happen this way!" frequently.  You're also very fond of being rules lawyers when something isn't quite going your way, but are more than happy to let it slide when things are.  That's our group dynamic and I'm OK with it.  In all honesty, I don't have anything prepared, per se.  I have myths and legends and I know where those are.  I now have some prominent NPCs (as they related to the PCs) and there are some enemies emerging.  What I don't have is some huge plot to play out over the course of the campaign arc.  I know that there will be another war with Carthage in about 20 years.  I know that Sicily is a hotbed of unrest right now and that over the next 20 years, Italia will fall pretty much under Roman control.

Again, I'm going to hammer on the point that every rumor cannot and should not lead to a quest.  Some rumors are just rumors.  Sometimes, there are no rumors and yet adventure awaits.  When you find a rumor, it's something is supposed to spur your imagination.  Who is this Rufus guy?  Is he a dick?  Does he have magic items?  It's supposed to make you want to do something.  Or maybe you're not interested.  I cannot create an adventure for every rumor.  If I did, I would be spending hundreds of hours building up modules and NPCs that would never be seen by anyone.  This is the Micro approach that Jim alluded to.  I have been detailing things from the Macro level (cities, trade goods, gods, legends, et cetera)   and what I'm discovering is that if you build an interesting and immersive world, your players will find their own adventures within it.  That's what you all are doing and I love it!

Leading the group around by the nose results in frustration.  Frustration for me because no one ever follows my clues or finishes my adventures, and frustration for the players because they want to do what they want, not what the DM tells them.  Railroading also utterly destroys suspension of disbelief.  It's like watching a movie and figuring out the entire plot in the first 10 minutes.  You know what's going to happen but instead of just leaving, you have to sit there for another 90 minutes before it's over.  I never want a game to go like that.  I want it to be almost as much of a mystery for me, the Dungeon Master, as it is for the group of players.  In answer to your query about the rumors, no.  I have no idea who these people are or what their motivations are. However, I do have the rumor notated and should anyone try to figure out what exactly Rufus is selling, or why the Tribune is a dick then we'll both get to find out!
                

                Carl,

I was avoiding the exact a-b-c mentality that you speak of in the quest progression and focusing more on what the map structure and general encounters for said dungeon would be.  Having pre-generated maps that you can ad hoc together with a general idea of defense or offense for said dungeon I think is more of what you need.

You are right in that we are developing our own story line and plots as you create this world from the macro down.  I know Lybos has a short/long term enemy now.  It will be interesting to see how it plays out.  Mainly because I know we are under a somewhat time constraint to find out about the missing flotilla.  Plus, he is not sure his companions are going to be gungho to do it right now.  But that might be just because we are bloody up.
                

                "...a mystery for me, the Dungeon Master..."

This sounds way too loosey-goosey.  I don't buy your argument.

What's the point of a DM, then.  Why don't you just play and let's all decide on how the mystery progresses, together, as the group.  

A DM should have some rough IDEA about what his rumors may lead to, be it an outline of events that have occured or will occur.  Winging-it leads to inconsistancy and inconsistancy leads to an unfun game experience.

                

                
Jason,

I see what you're driving at.  Having a few generic settings could be good.  I was trying to take the approach that any given "special" zone would be unique.  I also wanted to present a game world that would change with time and player actions.  If you kill off 12 members of a street gang in Ardea, that's going to leave a power vacuum.  When (and if) you return there, you should see the results of your actions.  Having plug-and-play dungeons as you describe can limit that sort of thing.  Still I see the wisdom in this and I'll take it into consideration when I'm planning to stay a step or two ahead of the players.

Thom,

I see my role as a Dungeon Master a little differently than you.  Rather than a force of control that opposes the players, I see my role as that of the world itself.  The world doesn't oppose you.  It is the place where you live and strive to achieve your goals.  It doesn't have a plan for you.  Rather it is the place where you create and act upon your plans.

Presently, we are all deciding how the mystery progresses as a group.  My job as the Dungeon Master in this is to play the part of the world and react to the players.  That (ideally) leads you to more actions.  That's an open-world style of play.  Do I know what happened to the ships?  Sort of....:-)  I have an idea of what might have happened to them.  It's up to you and the rest of the players to discover what really happened.

I honestly have no idea where the rumors are going to lead at this point.  I have some hints, as you do, as to what they mean.  Rufus, for example, travels a lot.  He could be some kind of flim-flam man or he could be an actual cleric/alchemist who sold a potion to someone it just happened to not work on them.  Or he could be something very, very different.  The point is that because I don't know yet, you certainly don't know, and that is a mystery and promises some sort of adventure.  Should you choose to pursue Rufus (although you should probably ask yourself why you're doing it -- I doubt you give a shit whether some charlatan is selling snake oil to the plebs or not) he may get a great deal more interesting.  Or he may not.  The dice will decide!

I wrote up a rumor table with no thought to it other than to make each rumor something that would cause discussion among the players, and that I be able to use it in pretty much any town in Italia.  Each rumor is a possible adventure seed.  It just depends on how the players want to act on them.  I add rumors to it from time to time to keep things fresh and replace the ones I've used.  Some of them will be red-herrings leading to chasing of wild geese (which may, in turn, lead to something cool).  Some of them will no doubt be recycled.  Complaints about rip-off wine bars and dickhead Tribunes will be more common than god/goddess sightings, but that's the way of the world, right?  Who doesn't have a story about a slimy politician?  Or some restaurant that you'll never go to again because the service sucks?

The inconsistency thing really bugs you, I know.  However, this is something that you're going to have to adjust to.  The game cannot be exactly the same each time it's played.  We may alter combat.  We may change how initiative and surprise are determined.  Spell effects will be altered as the game progresses.  Some spells may become unavailable.  Magic items won't operate consistently.  Weapon damage and reliability may change.  This is the nature of a shared fantasy game.  It is the advantage of playing with people versus playing on a computer.  Think about this: in this game, unlike a computer game, you can actually think your way out of dying (you made a deal with a goddess!)  I'm going to make rulings on things and they may vary from circumstance to circumstance.  That's my prerogative as Dungeon Master, and it's my responsibility as Dungeon Master.  The game isn't going to be fair and it's rules will not be consistent (they're more like guidelines anyway!), but I think you knew this.  The game will be fun, I promise.  Hell, I haven't been wrong so far.
                

                I understand what you are saying Carl.  I would imagine for long term things that are developing a special dungeon could be built.  But for that quickie this session we are raiding the thieves' guild of small ancient town, then having a quick dungeon map to pull out, rearrange, add some monsters, and use would be nice to have on the quick.
                

                I disagree that the characters won't follow your hooks.  We have a big hook right now - find out what happened to the Titanus and Hobgoblinus.

I think if we heard a rumor that there was a clue to their disappearance in a strange pyramid 10 miles inland, we would go look.  If you said the captain was rumored to be a zombie in a graveyard in Sardinia, we would try to go check it out.  I think we would follow any crazy lead to find out about the boats.

So, I'll argue that the "Find the Boats" storyline *IS* the dungeon.  You just need to decide what the steps are to get closer to the end of the story.

I know Raph Koster basically said that an MMO should just be a giant playground, and the players would make up their own stories within that playground.  I guess you could take that attitude in the current D&D game, but the "Find the Boats" hook has us players falling into the old habits of expecting a DM driven adventure with storyline.

Maybe we all need to get our expectations out on the table.  I guess I assumed I was in for a story-driven D&D campaign.  If the DM is instead planning to provide "SimRome", I can adjust to that.
                

                
Jim,  I had envisioned more SimRome than Final Fantasy.  The fact that a patron of yours lost some ships and now needs to find them is as much a part of the simulation as your character needing to feed himself.

I'm hoping for a playground.  So far it feels like one to me.
                

                Jim

Since this last Saturday was your first gaming session with us, I am afraid you have missed some of the development with the find the lost boats dungeon.  It was rumored that the flotilla was destroyed basically either of the two mythical creatures - I forget their names now.  But we as a group decided more on a more mundane reason - that it was one of the gangs vying for control of the docks in Rome.

I get the feeling based on how we the player characters decided to follow this thought pattern albeit after talking with the former Captain of the fleet rather than the mythical creatures route, the campaign for now has grown more to pirates and a nautical adventure (for now).

So I am not sure we would go investigate a pyramid 10 miles inland at this time if a rumor emerged for it.  What history and knowledge existing so far puts us looking into pirates and sunken ships near the straights.  I would say things are apt to change if after being unable to find the ships at said area and the last known part of call has the ships heading to Egypt instead.  Then that pyramid rumor might have truth to it. If said rumor was to emerge in game. ;)
                

                I feel the consistency of the campaign is found within the DM's nature. Whether the DM invents something while sitting at a table sketching out rooms, or invents something five minutes before it happens, the invention itself is born of the DM's mind. And because it comes from that single source, it is necessarily consistent.

'Inconsistency' in a campaign arises from within the player's perspective of the DM's inventive nature - and makes an easy accusation when it appears the DM's plan lacks cohesion.  It may be the player's resistance against the flow of the campaign, rather than the DM's apparent lack of a pattern.

But of course, the DM must TRY to be consistent - and that is the rub.  If the DM will insist on intentional inconsistency, or if the DM fails to be true to his or her own vision ... then certainly, Thom's point is well taken.  The campaign will become hopelessly loosey-goosey.

So the question to ask is not, IS Carl inconsistent, but, DOES Carl take steps not to be.
                

                
I think this is one of the most commented upon posts I've had in a long while.  Thanks to everyone for your participation.

Alexis, I maintain a campaign spreadsheet that tracks the macro and micro campaign.  I've been promising to send you a copy and so I shall.  It's a bit messy, but I'd like you to see what I've done with the material I've taken from you.  In that spreadsheet I keep campaign notes in an effort to remain consistent and not loosey-goosey.

Jason, your summary of events so far is great.  That is exactly what has happened.
                

                By making the dungeon a subterranian natural catacomb where anything and everything could exist, it eliminates any need for pre-planning. As such, it could twist into a volcano or lead out to the sea with entrances and exits being manmade(sewer entrances, root cellars etc.) or natural (caves, sink holes etc.).  Therefore, entrances/exits into/out of this "dungeon" could exist anywhere and at anytime with escaping critters emerging from its depths to be encountered above ground in any situation (town or country) as well as below. The natural catacombs can be endless lending itself to limitless eons of treasure hiders and seekers (with the creation of unlimited ghosts and gouls as well as fantastical creatures living in and amongst the fabled treasures, each with motivations of their own). A "dungeon" does not necessarily need to be a pre-defined enclosed area and the encounters could be entirely random. 
Any story overheard of treasure and creature could happen amongst any throng of humanity at any given time or randomly encountered and reiterated for glory and/or fear later. Therefore, a story/rumor does not necessarily need to precede enounter nor an encounter precede story/rumor. Randomness is the key. 
Once the group has decided to follow a particular course of action, then the particular confines of the encounter can be determined (eg. farmhouse basement). 
However, individual courses of action should not be dwelt upon in detail as it takes too long and isolates the rest of the players. Only group encounters should be detail actioned unless isolated individual action is pivotal to the whole. Most individual motivation and character building can be done in story format. Large actioned encounters should be group efforts.
For example: A character goes into the pub restroom and encounters an angry nymph - seems someone carved a toothpick from her sacred tree and she has traced it to this restroom- the person flees the bathroom to inform the group, they: 
1. Individual: decide to ignore it as drunken rambling and laugh it off - the nymph curses the character so that everytime he sits on the toilet he feels he has a toothpick poking him in the butt.
2. Large: Everyone decides to go check it out, the nymph is a daughter of Charybdis and she will try to drown the group in a whirlpool taking most of the town with her in her outrage until the sliver is returned and replaced in the sacred tree.
(options are: 1.the group can fight Charybdis (big battle) and try to escape, Or 2.they can fight Charybdis, try to escape, take up the quest = new storyline Or 3. the DM can state that another group has already started on the as-of-yet incomplete quest of returning the sliver and move toward another storyline, Or 4. the DM detemines the group has escaped the deluge to continue on their current storyline)
 
This saves the DM from needing to pre-arrange an encounter for every rumor/story/storyline and allows for large encounters whenever the DM has the time to flesh them out as well as allowing for large random encounters that do not even need to fit into the bigger storyline but adds adventure and excitement randomly to any game session. Ghosts and monsters do not attack when it is convenient and/or logical.  Hit the group randomly - it will be more fun that way. 
PS. You could find monsters from Roman mythology and assign them the stats from the D&D monster manual eg. Minotaur has the stats of the minotaur, Charybdis could have the stats of a giant water elemental, Notus could be a wind elemental(you decide its size) etc. That way you can throw Roman monsters at the group and have stats already made up for encounters. It would make random encounters quicker and easier.