Musings on Choice

A couple of recent threads (on plot and plot) led me to muse a little about the role of choice in an RPG, in particular, for the players of an RPG. Although it was not the original intent, I think this post quickly starts sounding like a DM advice article, so feel free to criticize it from that angle if you want.

The Basic Unit of Player Interaction
I do not know who first articulated the idea, but I wholeheartedly subscribe to the philosophy that every time a character acts in a game, the character's player ought to be making a choice. In a way, choice is the basic unit of player interaction with the game, and is one of the essential features that distinguishes it from a novel. Dwarf or Elf? Fighter or Wizard? Toughness or Weapon Focus? Sleep or flaming sphere? Go right or left? Talk or fight? Slowly shift into position or risk an opportunity attack to flank now? Stop to rest or press on? Rescue the villagers or chase after the fleeing villian? Ultimately, everything that happens in an RPG ought to flow from the choices made by the players (I'll get to chance in a bit). Hence, one of the key roles of the DM is to ensure that the players get to make choices in his game.

Choices and Consequences
Of course, in order for the players' choices to be meaningful, different choices should also result in different outcomes. A scenario in which the same outcome takes place regardless of the choices made by the players is seldom well-received as it means that their choices were largely irrelevant. The consequences of certain choices (in particular, those made in combat or during the straight application of mechanical sub-systems) are enforced by the game rules. As for the rest, it is essentially up to the DM to ensure that the players' choices matter.

Ideally, the consequences should be a reasonable outcome of the choices (barring complications such as incomplete or incorrect information - see next section). Of course, the key problem is that different people can sometimes have very different ideas of what is reasonable. One person's clever solution that should succeed can be another person's abhorrent scheme that ought to fail. When this happens to people on different sides of the DM screen, the DM may find himself either wondering why the players don't take the obvious approach, or aghast that the players are prepared to do something that he never thought they would. The consequences of a choice can sometimes seem overly harsh to the players, and can be a source of player-DM conflict. Because of this ambiguity, a DM should be particularly careful when using game-ending consequences such as character death.

In fact, in more complex scenarios in which the players have to make a series of choices, it is probably a good practice for the DM to envisage more than the two standard potential outcomes of complete failure and complete success. This may consist of having a variety of possible points along the success-failure continuum, or having multiple independent goals so that the players can achieve all of them, some of them, or none of them.

Choice and Information
Player choice can sometime be hampered though a lack of information, either because the DM has inadvertently or deliberately left out important information, the latter because finding the relevant information is supposed to be part of the challenge. In situations where the players may make choices without knowing all the relevant information, a DM who doesn't want the game to end abruptly should avoid using game-ending consequences, or ensure that the players get sufficient feedback before the consequence happens. This is for pretty much the same reason that a game of Hangman doesn't end after just one letter is guessed incorrectly, and why the game of Twenty Questions is not called One Guess.

Choice and Chance
There is a strong element of randomness in many RPGs, and occasionally, this means that even if the players make all the best choices, all they gain is a good chance at success. A DM who wants to reduce or limit (without completely eliminating) the role of chance in determining whether the players succeed or fail may decide to have certain consequences follow automatically from the players' choices, without requiring any dice rolls or other elements of chance. This works well with the "multiple independent goals" model mentioned earlier - the players may be able to achieve some of the goals simply by making the right choices. Other goals require them to make good choices and have luck on their side (or at least, not against them). This way, even though good choices cannot guarantee a complete success, they can prevent a complete failure.

As a DM, how do you ensure that the players get to make meaningful choices in your game, and that the consequences for good or bad choices are reasonable, bearing in mind that good and bad may be subjective? How do you balance the need to give enough information to the players to make informed choices with the need to avoid giving the players so much information that the best choice becomes obvious? What is your ideal balance between choice and chance? Approximately how much of the players' successes would you allow to be entirely due to their choices, and how much do you feel should be left to chance?

I tend to think that meaningful, informed choices are crucial for games. At least the type of games I would enjoy (e.g. a game of chance is not interesting to me - unless I can "game" on the chance somehow, by determining odds, minimizing risk sand maximizing winnings.)

Some game mechanics can benefit from adding more "choice" or more meaning to a choice. For example skill challenges - if it only matters that I use the highest skill applicable, it seems pretty boring and easy. The choice isn't really all that hard or interesting. But if different skills lead to different outcomes (even if all of them are positive), it gets more interesting.

I wound up posting at The Mule Abides about choice as well, although I was responding to the threads about player advice rather than DM, and looked at it through a filter of experience with old vs. new versions of the game.

Using your framework, here are some of the conclusions I came to:

Basic unit of interaction: Well said. I subscribe to this idea too and I know the 4E design team does because reading about their goals pre-release was where I first heard the idea that choice is the essential currency of games articulated.

Choices and consequences: In that blog post, I talk about the granularity of choices. Large-grain choices have big effects on the story - do we go in this dungeon or that one? Fine-grain choices are more trivial - do we take the east corridor or the west one?

It seems to me that, by designing the rules to give players more choices, 4E causes them to be finer-grained and thus less consequential. For an old-school D&D fighter, basically the only choice you make each melee round is "Do I stick it out and keep fighting, or retreat?" That's a very consequential choice, because (especially at low levels) there's a real chance of dying each round you stand and fight. A new-school fighter makes many choices each combat round, but each of them is less consequential because the outcome is designed to hang not on individual actions (removing the "I win" button) but on the sum of all the choices of the whole party working as a team. Greg Costikyan has pointed out that the more random events there are, the more the outcome becomes predictable and the less impact each random event can have, and I think the same thing is true of choices.

Choice and Information: Good points from the DMing perspective - this is why it's good to enable players to scout things out, gather clues, etc. From a rules perspective, 4E gives players lots of information so that they can make well-considered choices. In practice that means removing unpredictability, so that you don't have magic items that you don't know what they do until you try it. I often see new-school players worry that, since OD&D doesn't have rules for so many things, they'll be robbed of the ability to make choices because they can't know how the DM will rule on the outcome. In practice I like to talk to the group and try to get a collective common-sense ruling, which has the advantage that you don't need to know a lot of rules to gauge consequences - you just have to visualize the scene & use your real-world experience of what's possible. (Obviously this works best for a gritty, "realistic" game where real-world judgments are more likely to apply).

Choice and Chance: A lot of the choices especially in old-school games revolves around how much risk you want to take. As above, having a sense of what the risks is important to making the decisions. If you know there's a 1 in 6 chance of a random encounter every turn, deciding whether to linger and search for secret doors becomes meaningful.

To answer some of your specific questions:

As a DM, I ensure that the players get to make meaningful choices by having that be the only way things happen. I run a sandbox game where the world is not moving according to any plan and it's up to the PCs to decide what to do (including which of the many story hooks I dangle before them they want to bite on).

The consequences for good or bad choices emerge from the rules and the situation. Finding treasure is good, being killed is bad. I don't do anything extra to punish bad choices; if I think the PCs deserve to be rewarded, I see which of my NPCs might share that opinion and how they might act on it.

The players have gotten good about doing threat assessments and calculating risk/reward. There are in-game resources, like a sphinx that answers questions and the ability to turn invisible and scout around, to help them gain information. They never have perfect information because information-gathering has a cost or risk (the sphinx charges a fee, even invisible scouts may have random encounters) and so there's always a chance of unforeseen consequences.

I like to have chance play a huge role in my game at a narrative level - I start with little knowledge of my sandbox, so random encounters shape what's there, and reaction rolls determine the NPC's attitudes so I don't even know ahead of time who will be their friend or enemy (unless it's obvious from the situation or the PCs' actions). At the same time, if the PCs come up with a great plan I'm happy to let it work without any randomness; not having a skill system in OD&D, I'm less tempted to say "Well, that sounds plausible but let's see if you get a good or bad roll for your skill check".