Monday, July 8, 2013

Two Guys Walk Into A Bar...

Actually, for the purposes of this example, they haven't walked in yet, but they can peek through the window unseen.

First Guy 

GM 1: The bar is completely empty except for the bartender who is behind the bar, looking at the door, holding a shotgun.

Player 1: Ok, I got Sneak at 25% so I'm going to Sneak in and (roll r...)

GM 1: Hold it hold it hold it. The room is empty, the bartender is looking straight at the door. There's no way you can just sneak in there normally.

Player1: Ok, I'm going to set a trashcan on fire out on the street, then when the bartender comes out to investigate, sneak in.

GM 1: Ok, the trash can is on fire. The bartender (roll roll) does indeed go out to investigate after about a minute...roll to sneak...

Player1: Did it!

GM 1: Alright, you're in. And outside there's a trashcan fire. It's (roll roll) actually kinda getting out of control.


Second Guy

GM 2: The bar is completely empty except for the bartender who is behind the bar, looking at the door, holding a shotgun.

Player 2: Ok, I got Sneak at 25% so I'm going to Sneak in and (roll roll). Did it!

GM 2: Alright, you're in. Now the room was empty and the bartender was looking straight at the door. How'd you do it?

Player2: Mmmm...I guess I set a trashcan on fire out on the street, then when the bartender comes out to investigate, I snuck in.

GM 2: Ok. So she's out there and outside there's a trashcan fire. It's (roll roll) actually kinda getting out of control.


Some games work one way, some work the other way, some work both.

In D&D, for instance, a to-hit roll can work either way. You can describe exactly what you hope to do in an ambush, set it up, get modifiers, then roll it and then, bing, it goes off or (typically) you roll to hit an unlikely target and it works anyway because you rolled high and the GM or player then explains how it happened.


The 2 games above (and/or people playing them) look at the Sneak number on their character sheet in different ways.

GM 1 sees the Sneak number as representing how often the PC can sneak under typical conditions in the fiction.

GM 2 sees the Sneak number as representing how often the player gets to say "My character sneaks" and have that happen in the fiction.

GM 1 sees the Sneak number as a measure of PC aptitude.

GM 2 sees the Sneak number as a measure of direct player control over the fiction.


Both fictional situations are equally realistic (the same thing happens in both fictions).

Both fictional situations required the same amount of creativity (the same things were invented).

Both groups are visualizing the situation in the same amount of detail. (This is an ability distinct from just creativity.)

There is a difference though:

Player 1 is playing a game where she stakes her PC's life (that is, her right to play the game with that PC) on that creativity and that ability to  visualize the situation in detail, plus an in-game decision (go in sneaking) plus a 25% die roll gamble.

Player 2 is playing a game where she stakes her PC's life solely an in-game decision (go in sneaking) plus a 25% die roll gamble. The creativity and ability to visualize are there, but they are not gambled on.
(Player 1 is, also, staking her PC's life on her ability to communicate effectively with the GM. I've heard of adults having serious problems with this in their games--I've never seen it happen in real life.)
A big question in game design is, largely, what, if anything, do your players want to be gambling their players lives on?


  1. This is very interesting to me. Virtually all of the games I've run or played in have operated mostly in the first mode you described. An occasional situation might be resolved the other way, but the first way has definitely been the norm.

    However, is strikes me that it might be better in many situations to do it the other way (just roll the dice and then come up with a justification for the result), because the people I play with most often do have communication problems. There's one guy in particular (who is often the GM) with whom this is frequently a problem, not because he's a dick, but because he just has fundamentally different expectations about reality than the rest of us, which leads to constant misunderstandings. (Why keep gaming with this guy? a) He's a good friend, and b) for various reasons we don't have that many people to game with regularly.)

    It does seem to me that going about things the second way would help avoid some of that unnecessary conflict. The alternatives are to either constantly gamble on accurate communication (not a safe bet in this case) or to constantly slow the game down by questioning every last detail (which is what usually winds up happening, much to the annoyance of all involved).

    1. That is one of the most articulate descriptions of why one might go with Way #2 that Ive ever heard.

    2. On further reflection, it also occurs to me that pretty much every time we've done things the second way, it's been treated as a mistake. As in, "Oh shit, we rolled the dice without roleplaying something first, better come up with a justification for what happened now!" (This is, of course, assuming that we *noticed* that we were just rolling the dice without roleplaying something in advance. God knows we do it all the time without noticing.)

      This underscores for me how much we assume (and by "we", I mean the people I game with most often) that the first way of doing things is the *right* way. The more I think about it, the less sure I am that's true. It clearly has some advantages, but so does the second way.

      I'm curious now how widespread that view is.

  2. I'm not really seeing the distinction (what is gambled on) here... possibly because you describe only the success and not the failure.

    It appears to me that in either case, there is a 25% gamble, and if it fails the character is noticed and killed. So either way life was gambled on the sneak roll.

    The only real difference I'm seeing is that in #1, there was a chance that the NPC didn't go look, and if he didn't then the sneak would not have been allowed; while in the second, the NPC action was assumed. The first case requires GM interpretation and permission, while the second case is entirely up to the player's creativity and is resolved with a single roll.

    How do you see this working in the case of failure? If the gamble is failed?

    1. In the first one, the game just doesn't move forward to the success until the player thinks something up. So that's a gamble.

      Plus (and this isn't covered in the post) the plan could be better (giving a bonus) or worse (giving a penalty).

    2. The gamble aspect of RPGs is probably one of their strongest allures for me, but the whole "the game just doesn't move forward" thing is a bit dangerous. I think #1 works best if attempts are a resource and something interesting happens if the game starts to drag.

      Wandering monsters checks are the classic example of this for me, where if the players start stalling about for too long debating how exactly to open a door the game throws some danger and excitement at them as a kick in the ass.

  3. Although I can easily see the benefits of Way #2, it basically just moves the problem - not solve it.

    There is still the desire to have a suspension of disbelieve, the GM (and in theory players) want to know *how* they got in, not just that they managed to do so. This is why the GM asks "Now the room was empty and the bartender was looking straight at the door. How'd you do it?" he's happy to have it happen, he just wants there to be an explanation for it.

    In the given example the player suggest starting a fire - but if the GM (and the bartender) knew the inn was warded against fire, and the bartender was smart and suspected sneaky thieves, he may well be even more weary and would certainly not go an investigate the fire! This would have been factored into the roll for Way #1, making the chance of investigation low to impossible.

    So the GM says "Well, the bartender is on his guard, he wouldn't check a fire started outside". It is important to note that the player has not have committed to this action (which is different from Way #1, who now has an on guard innkeeper and a fire to worry about) however they still need to think of a plausible reason to gain access.

    If they can't think of anything, then it's up to the GM to say how they do it. This can elicit a similar response from the player however, along the lines of "No, my elven thief would never crawl in through the *dirty* privy hole - they're far too sanitary for that". This can create an endless loop of suggestions and rebuttals for why stuff would never happen that way.

    This loop causes just as much aggravation as Way #1 does with forcing the player to 'gamble' on their creativity to progress - however it does so in a way that *does nothing to forward the story*. After a dozen rejected suggestions for how the feat was accomplished (despite knowing that it was accomplished) the Way #2 player and GM have done nothing in game.
    Whilst in the same situation in Way #1 of "the bartender is too wily for that ruse, and stays put with his shotgun trained at the door" a dozen times - half the street has burnt down, both bribed and regular law enforcement have gotten involved, and basically a huge pandemonium of *story* has occurred.

    This, in my opinion, is why Way #1 is superior - you have to *do* things. Even if those thing's don't work, stuff still happens and the GM can progress (even if it's only the progression of a train wreck).

    Whilst Way #2 calls for either the same level of creativity, or for the *dismissal of logical outcomes*. If no acceptable solution can be suggested in Way #2, the group is forced to let the abstraction of success occur with no explanation of why or how it happened that way. As such both groups are no longer visualizing the situation in the same amount of detail, Way #2 has *less*.

    Of cause, in a story telling game where the GM and players both have the same command of the story - *and* the same in depth knowledge of all of the elements within the story (like knowledge of the fire-protected inn, the barkeepers secret identity and a full map of the layout of the inn, etc) then Way #2 works fine. However without this equilibrity of knowledge, Way #2 causes more problems than it solves, or at least produces less satisfactory results.

    (All in my opinion, of cause.)

  4. I don't care that much how much fiction is produced, so long as everybody at the table is having fun talking about or listening to whatever conversation's going on

  5. The second example is the way things work in "HeroQuest 2" (if I understood them correctly), a game by Robin Laws. The number in your character sheet represent the chances of winning a conflict. You roll against the opposition, that could be another PC, a NPC or simply "the world", represented by a given difficulty, and the result determines the level of success. Then, both GM and Player must decide what actually happened.

    Actually, "HeroQuest 2" is a fascinating RPG. I think the book is brilliant. Sadly, in play I found the experience too... abstract.

    1. Well, The truth is that when I play HeroQuest2 I use to play the first method. And I remember that before you roll, you have to say your tactic (Tactic is a step in the contest; naming the prize, tactics, opposition/resistence, roll...). I think the GM choose the method, like in others games.

      Burning Empires, Mouse Guard, in stealth contest is recommended to do the second method, I think in HeroQuest2 (exceuse my english) too. The authors think that if you do many dice rolls the failure is garanteed. And if you use the first method maybe you have to do many rolls.

      Well, in Heroquest2 is a little different. If you think stealth contest is very important for the story/adventure you have to do a extended contest (and there are two types, rising action and climatic scene, the last is blooded). And a extended contest is like D&D combat, you can do different tactics, you have rounds... But in HeroQuest 2, we have too the simple contest.

      The simple contest can look like second method, but like in extended contest, the GM rules and chooses. With a simple contest you can do a action or a "year" of actions, you can climb a tree or climb everest, but always in one roll. And you always have to say your tactics: page 21:

      "Tactics: Here the player(s) describe how to they're trying to get the goal. [..] if their suggestions seem unlikely based on the situation, the Narrator describes the circumstances more clearly or explains why the suggested course of action won't work in the game world. When a suitable tactic and governing ability is chosen, the [simple] contest is frame."

  6. The way ive always played is scenerio 1 with modifiers. You want to go in while the bartenders looking at the door, add 20% to your roll. If not, come up with something to improve the odds (trashcan fire, backdoor, wait for customers, etc)

  7. I'm agreeing with PseudoFenton in terms of which method I prefer, but like Zak, I don't care about the amount of fiction produced in the game. Instead, I prefer the order in which the fiction is produced.

    Method #1 requires commitment at the beginning of the turn. The player must declare an action, the DM must set the parameters for success, and then the dice are rolled BY THE DM. Then the DM & the player must deal with the result in a way that makes sense to the encounter and to the larger game world. The only failure in this example is that the DM rolls all the dice. It would be better for the player to roll the dice to see his chance of success carrying out his plan, giving the player more control in the outcome of events.

    Method #2 asks the player to roll for success or failure and then justify the roll, with the DM setting the parameters after the roll and figuring out the result of the bartender's actions. If he wanted, the DM could shoot down the player's justifications and then, as other commentators have pointed out, it could get into an endless loop of arguing.

    Method #2 appears to allow the player to be more participatory in the outcome of events, but it actually does not. The GM decides the result and the acceptability of the justification after the fact, waiting for the story instead of waiting for the roll result.

    Both of these examples move the story forward and include the player in constructing the story. Method #1 requires commitment to the story element the player suggests. Way #2 does not.

    I've played both ways, and I prefer #1, as it keeps the game moving forward.