A major decison that I need to make regarging game-play.

[Sandblaster]

I think it's a very interesting idea.  I was toying with something similar a while back, since I wanted to make a roguelike-ish game that focused on the secondary mechanics (permadeath, random content, character progression) without using the core mechanics of combat, random dungeon crawling, etc.

Like Darren said, the main problem is the level of content you'll need.  In order for the game to stay fresh, each text encounter needs to provide a unique challenge.  The brute-force way to accomplish this is to have hundreds of unique text passages and choice branches, so that the player never sees the same one twice - but that's probably impractical.

I wonder if there's a way to semi-randomly generate the text and choices, to keep them interesting?  You would need a way to prevent the player from simply memorizing key words and picking the "right" choice for each passage.  You would have to make sure the outcomes of your choice were different every time, but at the same time you don't want to randomize them the way FTL does with some of its choices, because then the player's decision is no longer meaningful.

[Trystan]

If you added a ton of items and options that allow you to do anything then you'd end up with something like this:

You see a wounded goblin with a short sword to your immediate right.

Press right a bunch of times to attack it with your +1 mace.
Press left a bunch of times to run away with your boots of fleeing.
Press down to run into a dungeon wall.
Press up twice then 'c' to hide in a nearby room and close the door.
Press 'w' then 'a' to switch to your giant axe.
Press 'q' then 'b' to quaff your blue potion.
Press 't' then 'b' to throw your blue potion.

But then you'd get tired of seeing the same options for each encounter (move, 'w', 'q', etc) so those could always available and not explicitly spelled out unless you ask for it (by pressing '?' or something).

Man, that would be a cool type of game.

[Endorya]

If you added a ton of items and options that allow you to do anything then you'd end up with something like this:

You see a wounded goblin with a short sword to your immediate right.

Press right a bunch of times to attack it with your +1 mace.
Press left a bunch of times to run away with your boots of fleeing.
Press down to run into a dungeon wall.
Press up twice then 'c' to hide in a nearby room and close the door.
Press 'w' then 'a' to switch to your giant axe.
Press 'q' then 'b' to quaff your blue potion.
Press 't' then 'b' to throw your blue potion.

But then you'd get tired of seeing the same options for each encounter (move, 'w', 'q', etc) so those could always available and not explicitly spelled out unless you ask for it (by pressing '?' or something).

Man, that would be a cool type of game.

The first thing to do would be ruling out key presing and adding a mouse. I think it would be a major improvement.

[Trystan]

The first thing to do would be ruling out key presing and adding a mouse. I think it would be a major improvement.

Using a mouse can be an improvement (see Brogue) but one of the things I like about roguelikes is that everything has different uses at different times. Trying to turn that into a Choose Your Own Adventure system seems like a rather difficult and shallow way to do something that's not as interesting in terms of gameplay and game mechanics. Of course it's possible since the old CYOA books are just a state machine like a computer. You could have much better narrative and a lot of cool descriptive text if you really go with it though.

I don't know; maybe it would be interesting to move around in the overworld map using the Chose Your Own Adventure method (or dialog trees or whatever) and play the individual encounters as a roguelike. Maybe I just feel like being a butt head today. Your other posts about your game ideas have all be interesting. I look forward to seeing what you come up with.

[Trystan]

UnAngband creates small text descriptions of rooms that show up when you first walk into them based on their layout and terrain tiles. Maybe something like that could work? Analyze the region and come up with a text description. You can then choose to explore it manually or automatically rest, search for loot, fight until leveled up, etc and have it auto calculate what the result probably be. Sort of like a simplified auto explore on steroids. I don't know if it would be fun, but it's another idea that seems to combine several that have been mentioned here.

[Endorya]

I wonder if there's a way to semi-randomly generate the text and choices, to keep them interesting?  You would need a way to prevent the player from simply memorizing key words and picking the "right" choice for each passage.  You would have to make sure the outcomes of your choice were different every time, but at the same time you don't want to randomize them the way FTL does with some of its choices, because then the player's decision is no longer meaningful.

It could work, but I would be cumbersome. I acknowledge it being hard because I tried to simulate it and creating a function to interpret a cave object to recreate creating all text possibilities and actions for all chucks of the  cave exploration. It would be incredible cumbersome to do if you really wish to offer dynamism. I would not venture myself to do it alone.

[requerent]

I think you misunderstood me a little bit.

I do understand that exploring a site like a cave or a dungeon in third person is the way to go but the world map?

I'm not contending text-based adventuring nor am I suggesting that third-person is better, but that your methodology, from the mechanics you've described, is insufficient.

It feels like you're generating a simulated world. Simulation games can be interesting, but in this case there would be a LOT of mundane activities, so you're looking for ways to short-cut those so that boring portions of the game are skipped and interesting ones receive focus. Games tend to do this to develop a narrative.

So- what narrative are you looking to develop? What narrative does playing your game facilitate? It's not a free-roam exploration game, because you have special exploration mechanics that narrate the exploration in a particular way. So- are these narratives connected in some way? Or is it just a continuous stream of novelties?

Each one of these experiences may be independently interesting, but how do you plan on making them meaningful in a broader sense? How is the 'world map' a necessary part of building the narrative? Your game will offer a lot of breadth with 100k sites, but they may all feel meaningless without some depth connecting them.

Anyway, rolling the dice is what roguelikes are all about.

No. Probability is a 'feature' in a roguelike. What roguelikes are all about are making decisions, without clear certainty as to how they will effect us, in such a manner that increases our chances to survive. I recommend playing Brogue (https://sites.google.com/site/broguegame/) for some inspiration. The failure to survive rests entirely on the fault of the player, not the dice rolls. The dice rolling creates variation, not determination, in how events resolve. In many games, the dice-rolling is arbitrary- because if a player encounters a situation where dice-rolling is important, she's already failed. Permadeath requires that a player always caters to the worst possible outcome. This means that dice rolls represent an opportunity for a boon rather than a penalty.

Your mini-exploration game doesn't provide the player an opportunity to avoid dice-rolls. What you really have here is a system of Random Encounters. The player can't foresee what will happen when exploring before making the decision to explore, so she is typically obligated to always explore- otherwise she will be missing content and the reason to actually play the game. So- you have to play a mini-game to get to points where you can play the game... in effect, you actually have two different games connected by a database. This can be interesting, but you will likely get a better game if your development time is spent devoted to making one of those experiences superb.

In games like Final Fantasy, random encounters exist as a means to pace the narrative of the story, manage flow, and develop ambiance. You wouldn't play the game for the REs because they are irritating and boring as hell-- the only motivation to continue playing is the narrative and interesting boss fights (and power-gaming if you're into that). The issue with text-adventuring is that the game is telling me what my narrative is. If it doesn't do more than connect a stream of novelties, it isn't giving me a story. Instead of being able to min-max my survival chances directly, it's obfuscated by a wall of text and some dice-rolling. Since the narrative isn't hand-crafted, what are our motivations to continue playing?

Have you ever played Seventh Sense? Check it out. http://www.projectaon.org/staff/david/

he idea of the game is to explore a huge world with tons of sites scattered throwout it (100k places to explore). Some areas will be empty others not.

This is another point of concern. What role does the player's character play in this world? Is she an already established hero or are we going to get a bildungsroman? If this game is primarily about fun simulated mechanics, it seems odd to emulate the simulated mechanics.

Now imagine being halted while in third person by bandits and wild life, as you try to explore a portion of land, just to realize at the end that that area was empty.

Don't generate meaningless areas? It's meaningful if the experience augments the narrative. Just make every path tie into some overarching theme. Or find ways to incorporate these seemingly nonsensical elements into a theme. For example- if you slay some goblin nobody but he turns out to be the goblin tribes successor king, they may start hunting you down to the point that they ally with some opposing nation that seeks to conquer that lands that you're involved with and... yada yada yada. You can make it so that the scope of an adventure is predictable or influences future events in some way.

First, you need to understand what type of land you're about to explore because it can be a dangerous place to go in at a specific time of the day, month, season or with certain weather conditions. You then define the pace at which you will explore it and define what you will be looking for. You will be able to actually define what each member will do during the exploration and if they should spread out to explore it way faster or remain together to increase their survival chances. Each choice shall bear an advantage and a disadvantage which only kin and careful players will sucessfully exploit.

I hope there are some options that will automatically do all of that for me. Will it tell me the danger level or do I need to discover and remember these things myself? It seems like another example of a poor mapping of feature to feedback. The player likely doesn't care about all of those things, just whether or not it's dangerous and how likely the area has something they are looking for.

Many events can be triggered during the exploration process like being ambushed or attacked by wild life, depending on the party skills that will tell who spots who first, giving the player the upper hand or a chance to evade an encounter.

The usage of skills, here, doesn't seem interesting in itself. The opportunity cost of skill development and the relative risk of choosing to go into an area could be interesting, but just having a skill floating there to modify random encounters can be disappointing. We want to get to POIs that are interesting- will these skills be useful in said POIs? It seems like we end up playing two different games. One is a dice game that determines what situations we end up in, and the other is a roguelike. They are connected only in that they share the same database. Sorry if I'm getting repetitive. Mini-games are fine, but, as this seems to be a game of exploration in many ways, your mini-game will be the main game and the roguelike will be the mini-game.

Other things will include unfortunate and fortunate consequences like falls and stepping on traps but also finding hidden treasure / items or special caravans carrying special cargo at a special price. Some situations will require player input to make important decisions like deciding if the exploration should be halted due to spotting a large enemy force or issuing shelter due to the weather changing, even though the area has been almost fully explored.

In a simulated world, locals become aware of POIs as a result of POIs likely being connected to civilization in some way (that is oftentimes why a POI is a POI). What are we exploring for in this game? The POIs could be unlocked via books or some other mechanisms that involve interacting with locals-- is there a 'need' for a world map? What incentive does the player have to explore apart from finding the fun of the game in POIs? If it's for special crafting opportunities (like in pokemon or something), why is the player motivated to do that? Is the player punished for not doing it?

I do think that the auto-explore feature does carry interesting decisions to make, in fact I believe its WELL above the options present in traditional roguelike exploration sessions. In ADOM the only options I had when exploring the world map were either [F]ight or [E]vade an encounter, that was it.

Okay- instead of paper-rock-scissors, it's now paper-rock-vulcan-lizard-scissors. It still isn't a mechanic that the player uses in a meaningful way. At the point in which it is, it's now the main game and the roguelike is the mini-game. Again- sorry for being repetitive.

Please understand, this post is all about world map exploration, and not about the best method on how to explore a dungeon or a cave, which is what this thread is all about.

I hope it's clearer now-- I'm not contending with your method of world map exploration, but rather the overall approach to developing the scope of your game.

I think that we need to know more about that to be able to rationalize whether your method of exploration is going to be interesting. The gameplay could be awesome, but I'm having issues understanding why that gameplay will be good. It seems like you're just trying to figure out a short-cut to manage a game that is too big-- this suggests, to me anyways, that the game is... well, too big.

[Vanguard]

"The cave exhibits tall, narrow and slightly leaned entrance with straight corridor leading to a dark and tenuous descent. A soft breeze can be felt coming out from it embraced in strange odor [check party total smelling skill] that you are [unable or able] to detect (or detecd as the odor of a specific mushroom)."

Options:
1 - Throw a rock into it and see what happens.
2 - Send one single member ahead to scout it.
3 - Enter the cave while leaving 1 or more members outside.
4 - Enter the cave with the whole party
5 - Tie each other with a rope before entering with the whole party

There's an old game called Darklands that does something like this.  It's a thoroughly excellent DOS RPG, and you get it from GOG.com.  If you want to take this approach, I urge you to give it a try.  Incidentally, it has by far the best medieval european setting of any video game ever made, so that's a plus as well.

As for the autoexplore thing, games are defined by their interactivity, and I think it would be a mistake to automate a potentially interesting mechanic.  If exploration is taking too long or is too uneventful, the obvious solution is to speed it up (eg shrink your map size or increase the area covered by one move) and more eventful (eg add interesting random encounters or increase the density of dungeons and such).

[joeclark77]

I actually think the OP's idea is very good, but I would just have it less verbose.  Instead of a text description, make it look like a data readout.  You could build a mini-map of the cavern as they click the buttons to "go left" or "go right" or "split up".  Use little green dots to indicate where party members are, and place numbers or status bars next to each green dot indicating how many torches that person has left, and how much free inventory for collecting mushrooms etc.  Red question marks could indicate where suspicious noises or smells seem to be coming from.  If you want, program it so that each person separated from the party leader operates on their own, by AI, so the party leader has to infer where they are (perhaps by making them trail a string) and try to bring the party back together once the goal is accomplished.  It makes it almost like a little puzzle game within your RL game.

[Endorya]

I'm not contending text-based adventuring nor am I suggesting that third-person is better, but that your methodology, from the mechanics you've described, is insufficient.

I'm not sure if you are referring  to the cave exploration example of the world map exploration. Either way, I've abandoned the text-mode cave exploration example.

So- what narrative are you looking to develop? What narrative does playing your game facilitate? It's not a free-roam exploration game, because you have special exploration mechanics that narrate the exploration in a particular way. So- are these narratives connected in some way? Or is it just a continuous stream of novelties?

It is a free-roam map exploration game without the third-person view of it, however it does go with the mundane third-person exploration model when exploring sites of interests, with the exception of settlements which will bring its own interface model that basically consist on listing all persons, buildings and services available on that settlement, so you don't loose time searching for people or specific buildings, mainly if we are talking about a huge city.

Your mini-exploration game doesn't provide the player an opportunity to avoid dice-rolls. What you really have here is a system of Random Encounters. The player can't foresee what will happen when exploring before making the decision to explore, so she is typically obligated to always explore- otherwise she will be missing content and the reason to actually play the game.

The type of land tells already what is expected to find if you read the right books (which will add this information on the tile general info). It's not "just" a system of random encounters, many things will be taken into account, as weather, season, time of day as well as specific creatures or resources might be only found in that particular piece of land. You will actually have WAY better view of what to expect to find than traditional roguelike games when exploring a dungeon, in which you have no clue whatsoever of what lingers about there, like in ADOM and CRAWL.

A type of cave, dungeon, or settlement will also provide important information on what is expect to find there. This way you build strategies and prepare your men accordingly before entering such places. This means it is not simply a matter of relying on luck, hopping that the Desert Dragon won't be found in the Small forest Cave

This is another point of concern. What role does the player's character play in this world? Is she an already established hero or are we going to get a bildungsroman? If this game is primarily about fun simulated mechanics, it seems odd to emulate the simulated mechanics.

That's something I really hate about roguelikes. I mean, having to choose a role at the beginning of the adventure having no clue whatsoever of the impact such choice will have during its journey (this is my personal opinion). In my game you don't get to choose roles before your mom enters in labor. You develop the skills want as you grow up, there is no skill restriction whatsoever, though you won't be able to excel in every category, obviously.

You will not born as hero. You will be a common being. Your decisions and actions will be the one thing to dictate how much loved or hated you are among different civilizations. You can die as a criminal or a common mortal or die as a noble hero with a proper ceremony, having your own statue at the hall of the famous.

Don't generate meaningless areas?

If I don't generate meaningless areas there is no point adventuring yourself into far territory because everything around you will be too rewarding already. Of course this is all dependent on what you are trying to achieve. If meaningless areas do not take too much of your RL time I don't think this will be a problem, at least it won't for me.

I will skip now most of your sayings because it would take a great deal of time to explain them all, word by word.

The main problem here is that any option we are discussing can in fact work, it is not about having or not having meaningless areas, having or not having random encounters, having or not having third person exploration models, having or not having skills setting up what kinds of encounters the player is able to find (which actually won't work this way), having or not having to choose roles at the beginning of the adventure. Anything can function with the right parameters and balance. You are viewing these options based your own perspective while I'm viewing them with my own set of parameters in mind.

Then you have personal preferences. I prefer quick map exploration while others may not. I prefer not having cities being explored in third-person while others may. Some may prefer more text and less graphics and others may prefer more graphics to text. Some may prefer intense hack&slash while others might prefer exploration and survival. It is not about what is best but what do you prefer. I can't really state that a certain feature best fits the interest of all. It is not what I'm trying to do here. No matter how good features can be, some people will just hate it.

Some people love CRAWL while I personally can't stand it. It think of it as an INCREDIBLE repetitive and dull experience presenting no choices whatsoever besides either exploring the dungeon or leaving it; its combat is all about either you attack or you flee (mainly with mana-based roles). But you know why I mainly don't like it? Because I'm not into hack&slash anymore. So if I was to improve Crawl game, I would have a huge list of things to both remove and add. But I can't do this because the game has the purpose of being the hack&slash type.

What I'm really trying to do with my game is offering the player many choices and options and a game featuring a detailed, complex system, focused mainly in exploration and survival. You will be able to mine, skin, do pottery, build a music instrument, hunt and catch animals, learn how to build a house, build a ship, build your own tools, exploring the world, dig a tunnel, mine ore, mine gems, learn alchemy, do jewelry, create an armor from dozens of types of metals, create metal alloys, enchant your items, practice magic, explore caves, interact with NPCs, making new NPC friends, do quests, slay evil creatures, slay good creatures, put your items at sale, search for treasure, do bounty hunts, participate in a huge battle with thousands of troops, make love or murderer some one, visit a brothel, marry a NPC, purchase or create your own furniture, take a bath, recruit new members and gain their trust so you can become friends, butcher a corpse and have its internal organs used as trophies and ingredients, get a cold from cold weather, catch a disease, die from dehydration, perish from starvation, kill and eat your own party member, kidnap someone, The list goes on... and on... and then some...

And now you ask, why would I want to do pottery? Well, why not? If you have goods skills you can actually sell your pots for a good price. Why taking a bath? Well lets say your looks will have impact on NPCs. It really doesn't matter if I have a good explanation for all these options if you are just into hack&slash or if you prefer to play flight simulators. You may consider taking a bath something tedious and unimportant, which I fully understand but luckily, this option will be only checked during specific social events like going into a ball or into brothel, so it won't bother you during your regular expeditions and trading sessions. All comes down on how you do implement this features. Some features can actually be set to perform automatically so you don't loose focus or get consumed by tasks you may consider boring.

All I want is to provide the player with choices and new experience opportunities, something present in Skyrim (to some level) and something that most roguelikes lack (with the exception of DF and some other roguelikes). But implementing all this stuff in Skyrim would be very demanding as far resources are concerned, though "easy" (while being a slow task) to implement in a roguelike game because it does not focus on eye candy.

Only through a demo the project would be fully explained. And to make a demo showing what it can do, it means having most of its features implemented.

[EDIT]
I'm keeping a close eye in every single feature being implemented having the potential of becoming tedious or boring. I do hate tedious tasks! But I guess you know that by now

[Endorya]

I actually think the OP's idea is very good, but I would just have it less verbose.  Instead of a text description, make it look like a data readout.  You could build a mini-map of the cavern as they click the buttons to "go left" or "go right" or "split up".  Use little green dots to indicate where party members are, and place numbers or status bars next to each green dot indicating how many torches that person has left, and how much free inventory for collecting mushrooms etc.  Red question marks could indicate where suspicious noises or smells seem to be coming from.  If you want, program it so that each person separated from the party leader operates on their own, by AI, so the party leader has to infer where they are (perhaps by making them trail a string) and try to bring the party back together once the goal is accomplished.  It makes it almost like a little puzzle game within your RL game.

Thanks. But I've abandoned the text-feature. I would only venture doing it with the help of someone.