It needs to be more of a sandbox than a railroad?

[Sadras]

Shifting goalposts implies someone is keeping score. I'm not keeping score, are you keeping score?

No score is being kept, I assure you, the phrase was more utilised in the manner that you have taken the discussion on a tangent given the discussion that was preceding as explained in my post.

The third option you overlooked is where the GM says "this is what I had planned tonight, so it's either this or we play Munchkin". Because, again, social contract.

I would classify this as railroading. In/out game, its still railroading.

 

[mcbobbo]

I would classify this as railroading. In/out game, its still railroading.

This definition of railroading is far too broad to have a meaningful discussion about RPGs. If you define out of game things as railroading then we are discussing different things entirely.

 

[Grogg of the North]

For a really good railroading example, go here:

http://www.shamusyoung.com/twentysidedtale/?p=615

As for sandboxes, there are other traps as well. I've been in a sandboxy campaign that went like this:

PCs: We go north to the forest to see if we can gather food to feed the townsfolk.
DM: Okay, you're ambushed by gnolls [five levels above you]. You get captured and sold as slaves.

It felt like we accidentally wandered out of the starting area only to realize, too late, that we needed to be killing 10 rats in the sewers.

 

[Quickleaf]

I certainly agree with you that they are generally mutually exclusive, in that an adventure designed to be a railroad probably isn't going to serve as a sandbox, and vice versa.

I'm only disagreeing on the (relatively narrow) spectrum issue.

Ah, fair enough. I usually assume a coordinate system of multiple spectrums (weak-strong framing, sandbox-railroad, dungeon crawl-event based, etc) when describing adventures.

As to how to solve the design issue, my own view - based on intuition more than robust investigation and evidence - is that it can't be totally solved. Certainly, when I look at some well-regarded adventures from the 2nd ed era (eg Dead Gods, various Ravenloft offerings, etc) I can't see anything but hopeless railroads that I couldn't even think of running (because most of their scenes have no real hook outside the context of the railroad, and so don't lend themselves to adaptation to my preferred style).

My current thinking, and how I've been experimenting with adventure design, is about total adventure design that takes the best features of sandbox and railroad, and adds a touch of indie inspired design.

A sandbox gives the players agency (meaningful choices and consequences), but rarely has the same emotional impact as a traditional story. With more passive players it can lead to moments without a strong sense of direction or motivation and a sort of "Uh, what do I do, DM?" attitude.

A railroad gives that emotional narrative and provides a strong direction, but sacrifices player agency. With more active or critical players it can lead to moments without a strong sense of motivation and a sort of "why do I care?" attitude.

My goal with total adventure design is to have an adventure that:

1. Ensures the players have agency throughout. (sandbox)
2. Provides an emotionally gripping experience and strong direction. (railroad)
3. Creates/encourages strong motivations, i.e. for the players to know what they're after and to have a reason to care about it. (indie)

Currently I'm approaching #1 by...

• Building open ended encounters that can be solved in multiple ways.
• Providing indexes to the various sites & scenarios encountered in the adventure, for a DM to use in a plug-and-play manner.
• Giving multiple hooks for each adventure.
• Tying encounters to each other, noting how the outcome of one can affect how others play out. 
• Writing multiple endgame states to each conflict, and how the players' decisions/actions might lead to each.
• Trusting that boundary conditions to the adventure don't diminish agency.

I'm approaching #2 by...

• Making the main story arcs rugged enough to adapt to player abuse.
• Presenting the adventure in discrete chunks that are thematically connected, then choosing a default presentation of those chunks according to recommended experience level.
• Framing opening scenes to provide strong direction.
• Spending more time developing compelling NPCs and how their plots would unfold without player intervention, as well as notes on how they adapt to changes.

I'm approaching #3 by...

• Creating a choice at character creation (and beyond) for how individual PCs connect to the various story arcs and themes. Likewise, if there are types of characters (not) suited to the adventure, listing those up front.
• Referencing those choices at different points in the adventure, particularly when framing the scene for new conflicts.
• Giving players opportunity to see firsthand the unique outcomes/consequences of their decisions and actions.

 

[Sadras]

My goal with total adventure design is to have an adventure that...

Nicely analysed.

 

[pemerton]

Keep on the Borderland is a sandbox so adventures can be written that are sandboxes.

Both those examples cannot be "true" sanboxes, though, because you can't (typically) say "giants are dumb, let's go somewhere else". Or "that temple has been there forever, we can come back later".

<snip>

Implicit assumption is GM provides something interactive for the players

The last time I GMed KotB it was more like what mcbobbo describes in the final quoted sentence than what Emerikol describes in the first quoted sentence.

The players had both built (multi-class) thief PCs. Of all the classic D&D classes, in my view the thief actually provides the clearest hooks for the GM: players who build thieves want to engage in skull-duggery.

So, I narrated them as being in the Keep, and then framed them into conflict with the evil priest (it was a long time ago, so I've forgotten the details). The campaign ended up being about the cult (which I ran as a necromantic death cult), and in due course the focus of adventure moved from the Keep to a nearby town which had also been infiltrated by the cult. The PCs never went to the caves (I can't remember if they ever explored the "wilderness").

So I did provide something interactive for the players, drawing on the material provided by the module. But, contrary to the paragraph from mcbobbo in the middle of the quotes, the players didn't have to take their PCs into the caves. All of us (players and GM) followed the adventure where it led.

Either your DM allows for you to leave the area and ignore Giants/Temple (playing a sandbox campaign) or forces the issue (railroads the campaign)

Similar to my discussion with [MENTION=20323]Quickleaf[/MENTION], I think it misdescribes the range of options to treat sandbox and railroad as two extremes on a spectrum. There are other approaches. For instance, if the GM describes the Keep being under attack by hobgoblins from the Caves, then that is "forcing the issue", but - provided the GM is actually framing the PCs into a situation of interest to the players - then they are not just going to have their PCs leave the area.

But this goes back to the issue of D&D adventure design, raised upthread by [MENTION=6670763]Yora[/MENTION] and [MENTION=23240]steenan[/MENTION]. D&D modules have a tendency to be very weak when it comes to the situation. So instead of suggestions for forcing the issue by dropping the players into the action ("You are in the Keep when hobgoblins assault it - how do you react?"), they tend to either set out a rather static situation (static, at least, as far as the PCs are concerned - eg KotB, GDQ, etc) or else set up a "hook" which the PCs have to follow if the adventure is to go anywhere at all (countless examples could be given, but Dead Gods and Expedition to the Demonweb Pits are two that come straight to mind).

That's one reason why I'm fairly choosy with the modules that I use.

A railroad gives that emotional narrative and provides a strong direction, but sacrifices player agency. With more active or critical players it can lead to moments without a strong sense of motivation and a sort of "why do I care?" attitude.

I think the tension in this paragraph brings out my own objections to railroading - they purport to give dramatic narrative but in fact frequently fail to do so, precisely because of that "why should I care?" problem. The player is, in effect, being prescribed a pre-given emotional response by the GM.

This can work in certain media - cinema is particularly good at it - but in my experience RPGing typically isn't such a medium. And the quality of the stories (plot, theme and especially actual, real-life pacing) is such that they don't work just as stories on their own terms. They work by way of buy-in. And the safest pathway to buy-in is co-creation.

Which rules out railroading.

1. Ensures the players have agency throughout. (sandbox)
2. Provides an emotionally gripping experience and strong direction. (railroad)
3. Creates/encourages strong motivations, i.e. for the players to know what they're after and to have a reason to care about it. (indie)

Currently I'm approaching #1 by...

• Building open ended encounters that can be solved in multiple ways.
• Providing indexes to the various sites & scenarios encountered in the adventure, for a DM to use in a plug-and-play manner.
• Giving multiple hooks for each adventure.
• Tying encounters to each other, noting how the outcome of one can affect how others play out.
• Writing multiple endgame states to each conflict, and how the players' decisions/actions might lead to each.
• Trusting that boundary conditions to the adventure don't diminish agency.

I'm approaching #2 by...

• Making the main story arcs rugged enough to adapt to player abuse.
• Presenting the adventure in discrete chunks that are thematically connected, then choosing a default presentation of those chunks according to recommended experience level.
• Framing opening scenes to provide strong direction.
• Spending more time developing compelling NPCs and how their plots would unfold without player intervention, as well as notes on how they adapt to changes.

I'm approaching #3 by...

• Creating a choice at character creation (and beyond) for how individual PCs connect to the various story arcs and themes. Likewise, if there are types of characters (not) suited to the adventure, listing those up front.
• Referencing those choices at different points in the adventure, particularly when framing the scene for new conflicts.
• Giving players opportunity to see firsthand the unique outcomes/consequences of their decisions and actions.

I'm not sure what audience you are writing adventures for. Are you talking about commercial publication? I think that certainly imposes some fairly stringent additional constraints on adventure design.

When I am preparing adventures for my own game, I focus on open-ended encounters, and strong framing based upon PC build and subsequent PC play. Well-conceived situations that hook onto the players' preferences (as expressed via build and play of their PCs) means that I don't really need a "main story arc". Good situations lead to engaged play, and with engaged play the story arc will take care of itself.

Here's a link to the Burning Wheel session I ran last weekend. To run that session I needed the players to build PCs with some clear motivations, plus my map of Greyhawk with its geography around the Wooly Bay. The rest "wrote" itself.

I'm not sure how I would present that as a module: maybe maps and some basic geography; NPCs with motivations and connections; some ideas for possible conflicts/developments; and advice on what sorts of PCs the players might build to hook into all this.

I think the 4e Neverwinter Campaign Guide is a possible model for this sort of thing. So is the Penumbra d20 module Three Days to Kill (though at a smaller scale).

 

[mcbobbo]

Is it not fair to say that since they didn't go to the caves, you brought the cult to them?

Unless maybe that nearby town was always there and not made up to facilitate the adventure?

 

[Sadras]

Similar to my discussion with @Quickleaf, I think it misdescribes the range of options to treat sandbox and railroad as two extremes on a spectrum.

Either the door is open or the door is closed. You can certainly close and open the door as many times as you want, but the door will either be open or closed at a specific point in time.

There are other approaches.

I disagree, based on your example below. It is a railroad.

For instance, if the GM describes the Keep being under attack by hobgoblins from the Caves, then that is "forcing the issue", but - provided the GM is actually framing the PCs into a situation of interest to the players - then they are not just going to have their PCs leave the area.

I'm unsure the bolded text (emphasis mine) is even needed given that the DM is always going to frame things of interest to the players. If not for the players then for whom? Background? In some instances 'the interest' might be easily identifiable, other times it might be veiled.
In any event, just because it is in the interest of the PCs doesn't make it any more or less of a railroad. Your example implies that your perception of railroad means that the PCs interests are unaligned when railroading generally.

But this goes back to the issue of D&D adventure design, raised upthread by @Yora and @steenan. D&D modules have a tendency to be very weak when it comes to the situation. So instead of suggestions for forcing the issue by dropping the players into the action ("You are in the Keep when hobgoblins assault it - how do you react?"), they tend to either set out a rather static situation (static, at least, as far as the PCs are concerned - eg KotB, GDQ, etc) or else set up a "hook" which the PCs have to follow if the adventure is to go anywhere at all (countless examples could be given, but Dead Gods and Expedition to the Demonweb Pits are two that come straight to mind).

I'm of the similar opinion. I find one technique to encourage players to follow a course of action (progress the story), instead of forcing-the-issue in one or another way, is to "roleplay" more. Have the NPCs be more engaging with the PCs, and somehow implant an emotion/opinion that shatters the intial indifference of the players.

And when I mean "roleplay more" and "engaging" - I mean through the use of voices, gestures, being chatty, displaying NPC peculiarities, description...etc. Displaying an illusion of sandbox, but using suggestive techniques.

 

[Yora]

My current thinking, and how I've been experimenting with adventure design, is about total adventure design that takes the best features of sandbox and railroad, and adds a touch of indie inspired design.

A sandbox gives the players agency (meaningful choices and consequences), but rarely has the same emotional impact as a traditional story. With more passive players it can lead to moments without a strong sense of direction or motivation and a sort of "Uh, what do I do, DM?" attitude.

A railroad gives that emotional narrative and provides a strong direction, but sacrifices player agency. With more active or critical players it can lead to moments without a strong sense of motivation and a sort of "why do I care?" attitude.

The One Ring has a couple of pretty neat adventures that have a clear story, villain, and other NPCs, but leave very much room for the players to decide how the story will play out and end.

For example, one adventure has the players searching for a criminal who escaped on his way to trial when his two guards were killed in an orc ambush. He decided to leave his life behind and start somewhere else, but got captured by a large band of outlaws, whose leader wants him to become their friend and tell them about the local defenses. And he was never popular among his people and going to be exiled or hanged for his crime anyway.
It's up to the players to decide if they consider him a murderer who needs to be recaptured dead or alive, or if they want him to come back and confess the truth so he can plead for mercy. They may treat him as a captive of the outlaws, a traitor to his people, or a stupid kid who has no idea what kind of person his new friend is. They may want to kill him before he can tell the outlaws too much, drag him back to be interrogated and executed, or try to convince him to tell his people everything he knows about the outlaws or even lead them into an ambush to make up for his original crime.
There are so many possible ways things can go, which entirely depend on how the players interpret things and what they consider the right thing to do. The adventure shows no preference either way and is structured in a way that you still have pretty much the same encounters. What varies is how many opponents the players might face and who might be fighting on whose side. And the amount of opponents they are facing might in turn have quite some impact on the players descision to fight, flee, or negotiate. The adventure does not need any alternative branches to follow, it still uses all the material and visits all the locations. It's not that the writers had to write different paths and only one of them would be used.
That particular adventure is only 22 pages, but the same principle is used for the 140 page Darkening of Mirkwood campaign.

There's no reason to not make adventures like this for D&D, but WotC and Paizo want to do their linear dungeon crawls with balanced encounters. It's their choice, no necessity in any way.

 

[Umbran]

The last time I GMed KotB it was more like what mcbobbo describes in the final quoted sentence than what Emerikol describes in the first quoted sentence.

There is something that people forget. Sandboxes are *boxes*. Boxes have sides. There is (implicilty or explicitly) an agreed upon play area, and your'e expected to stay within it. If you step out and go over to the basketball court, well, that's your wish, but you are clearly outside the sandbox.

An adventure is a sandbox if, within the context of the adventure, the players are free to do what they wish.

A campaign is a sandbox if, within the bounds of the campaign, the players are free to do what they wish.

So, the adventure can be a sandbox. If, however, they players can't say, "giants are dumb, let's go do something else", then the campaign is not a pure sandbox, even thought the adventure still is one.