Is DM fiat okay?

[Crothian]

If only because people who have a definition agreed upon have nothing to disagree about?

That might be it, I don't always know the reasons why things work like they do. I just know how they work.

 

[IcyCool]

The GM is not required to agree to that intent. If he does agree to that intent, and the roll is successful, he can't then go back on his word and subvert that intent. In D&D, the only thing prevent the DM from doing so is his sense of fairness. If he wants your Move Silent roll to mean nothing, it will mean nothing.

So the GM isn't required to agree to the intent, good to know.

I was trying to point out that it's a difference, yes. The main difference is that the roll means something, and the GM can't change that.

Fair enough, but neither can the GM change it in D&D. "If he subverts that, he's not playing the game." It is equally true in D&D and in BW. It's just that in BW, it is more obvious if the GM subverts the action. That's just bad DMing.

 

[buzz]

Fair enough, but neither can the GM change it in D&D.

Sure he can. That's the point I've been trying to make with guards example. The DM can subvert your successful Hide/Move Silently roll by simply saying, "You manage to be whisper quiet and hide in the shadows... but a guard decides to check on the king's fish tank. He walks into the room, sees you, and starts shouting for backup."

In BW, there isn't the same specific task focus, and the achieved intent is sacrosanct. The GM can say that a guard walks in to feed the fish, but that can't change the outcome of the roll. If you were successful, the in-game narrative would be that you managed to duck under a desk as the guard walked in, and you watch him feed the fish from your hiding spot. He leaves, and you go back to picking the lock or whatever.

If the GM said that the fish-feeding guard negated your roll, he'd basically be saying, "You know when I said that a successful roll meant you stayed hidden? I lied." At that point, you're not playing BW (or similar games) anywhere close to the rules. The GM is simply not given the power to do stuff like that.

It's like if you were playing Monopoly, and one player insisted that since they were using the racecar, they didn't have to roll dice; they could just "drive" to any spot on the board, and even run over other player's pieces. They simply are not playing Monopoly anymore. Ergo, we're beyond talking about the pros and cons of Monopoly.

Agh. Sorry to keep going off on this tangent, Croth.

 

[D'karr]

Well, I'd agree that the above situation is not going to work in any system, as you're basically describing a GM who has created undetectable uber-guards and is staunchly refusing any intent that changes that fact. FWIW, at least I know he's doing this ahead of time in the conflict/intent method, and can just not roll and try and figure out another way to get what I want.

And that is the part I don't "like" about a system like that for D&D. How would a character, in D&D, know what his chances of accomplishing X are? He has to ask the DM. A reasonable DM will provide a reasonable answer. But an unreasonable answer does not immediately mean that the DM is being abusive.

Anything reasonable, not covered by the rules, should still have a reasonable chance of success. You can call it a DC a Percentage or anything else that suits. However, every action has a consequence, or should have one. So for example if I want to try to sommersault over my enemy I have a DC of 25. In other games I might have a percentage chance. In the "agreement" game the DM and I agree that if I make the roll I make the sommersault. Up to this point there is no difference between the three systems. BTW, in D&D this is already covered by the rules.

But what happens if I try something reasonable not necessarily covered by the rules. Let's say I want to jump off a cliff and land on the back of the flying Roc just below us. There are many ways a DM could handle this. He might chose to assign an ability check with X DC and a grapple check or a jump/tumble check and an attack roll or any myriad ways to resolve it. The fact is that at this point a reasonable DM is trying to come up with a way for the character to do X. In every game that does not have rules specifically for this he would have to do the same. If the "agreement" game has a mechanic for it then it can be used. If there are no rules for it the DM is using DM Fiat. He is in essence improvising. And improvisation is DMing, because no matter how hard they try, no game out there can ever hope to have rules for all the crazy things that players will attempt to do. And even if the game has rules that generally cover that, what about those times when the rules just get in the way?

In D&D, the DM doesn't have to tell me any of this up front; I could roll and he can ignore the result and my character is caught. You can hear the train coming down the tracks.

Yes, and in D&D he doesn't have to tell the player what the AC of creature X is either. But we still trust him to be reasonable about it. Do we not trust him because he doesn't have to tell us or do we not trust him because he is untrustworthy. Your example above is an example of poor DMing, and of course an abuse of DM Fiat.

In the above case, what if the fact that you are going to get caught is necessary to this "narrative".

I'll keep this under the understanding that we are talking about reasonable DMs and reasonable players.

Let's say that I'm DMing an adventure in which I need the players to be captured. Let's say I have a perfectly valid reason for the players to be captured. In D&D I can go about it multiple ways. All of them according to the rules without any DM fiat. The DM Fiat came into effect when I decided that they needed to get captured. However, if the rules don't cover it I'd have to make something up that allows me to capture them, once again using DM Fiat. What if originally the adventure did not require me to capture the characters but halfway through the adventure I figure that situation X would make more sense if the characters were captured? Am I allowed to change the script to make the adventure make more sense or make it more enjoyable?

I imagine that in the "agreement" game I could do something like this also. If it requires the player to agree to it, then either the player trusts me enough that he'll allow himself to be captured or the game has some clause that allows me to "break" the rules. Because in this specific adventure, if the players are not captured there is no adventure. If the game does not allow that then the DM is left to decide how to accomplish that.

In both those cases the fact remains that the DM will have to make some judgment calls that might not be covered by the rules, in some cases without the players consent. Does that make the "arbitrary decision" any less valid?

Okay, but again, I think then we're going far beyond the scope of the GM fiat issue. Nobody is arguing that a game can solve a faulty social contract, but I was assuming that we were talking about a functioning group as a basis for assessing GM fiat.

Since we are talking about "reasonable" group, the questions anwer still remains. Is DM Fiat Ok? Yes.

Right, and I'm not arguing that. My posts were initially in response to the claim that GM fiat is part-and-parcel of all roleplaying, and I didn't agree, and had examples from existing games to share.

However, even in those games the DM is going to have to improvise. If he is forced to improvise he is using DM Fiat. In the game you are using as an example both the DM and player have to agree to the outcome. If you are playing with a reasonable DM the answer to DM fiat is still yes. Hopefully the player is not an asshat himself and agrees to the game as brought forward by the DM.

I would argue, though, that the more latitude you give the GM, the more prone the game is going to be to abuse, even form the well-intentioned. The simple fact that the DMG, DMG2, the latest Dungeoncraft article from Monte Cook, and a host of DM'ing advice written over the years specifically address the issue of fairness is demonstrative of the enormous game-ruining potential of DM fiat.

I disagree. A good or bad DM can always improve and there is good reason to have all those sources of information. DM Fiat does not pose a greater risk of "enormous game-ruining potential" than using the Mystic-Thurge in a game.

Sure, you can say, "Well, don't play with crappy DMs, then." This is easier said than done, as the many tales of cat-piss men you can find on gaming fora will attest to.

But I'll still say it. However, I would amend that to be "Don't play in crappy games." Even bad DMs need the practice to get better and a "reasonable" group should be able to help a bad DM become better, even good.

I don't have much sympathy for people that choose to stay in crappy games and then come to places like this and cry about it.

Ergo, I, personally, really like how a lot of new RPGs handle things with more checks and balances, as well as explicit procedures for intents and outcomes. I think it minimizes the dependence on the skill level of the GM. To me, that's pretty cool. It's gets the system-handling out of the way so we can focus on the game being played.

I've already found a game like that, D&D.

 

[IcyCool]

Sure he can. That's the point I've been trying to make with guards example. The DM can subvert your successful Hide/Move Silently roll by simply saying, "You manage to be whisper quiet and hide in the shadows... but a guard decides to check on the king's fish tank. He walks into the room, sees you, and starts shouting for backup."

Well, no. The DM sets up a difficulty for the task, and if the player succeeds, the DM describes what happens next. If the DM decides to screw the player out of that success, he isn't really following the rules of D&D anymore, is he? You claim the same for BW, the only difference being that since the player knew the situation prior to rolling, he's far more aware of the DM screwing him. You seem to want to insist that the DM can screw the players in D&D, but not in BW. I'm pointing out that he can do it in both. It's just far more obvious when he does it in BW. And that if he does, it doesn't really have much to do with the system (or GM fiat), and a whole lot to do with the DM being a prick.

I understand you are a fan of BW. Please understand that I am not saying it is a bad system. But you seem to be thinking that the GM randomly changing rolls and "screwing" the players is built into D&D, which is false. And then you base arguements off of that. If this is not what you are doing, then I've misunderstood you.

Agh. Sorry to keep going off on this tangent, Croth.

My apologies, if you aren't enjoying debating this topic, I can stop.

 

[buzz]

So for example if I want to try to sommersault over my enemy I have a DC of 25. In other games I might have a percentage chance. In the "agreement" game the DM and I agree that if I make the roll I make the sommersault. Up to this point there is no difference between the three systems. BTW, in D&D this is already covered by the rules.

If all that's at stake is, "Do I somersault over my enemy?", then, yes, there's not much difference. Once you're using the full-on combat rules in BW, you are literally resolving whether you hit, whether you disarm, how severe a wound, etc.

However, stakes in conflict-based games are often wider in scope, especially when we're not just talking about a physical feat. E.g., in the sci-fi version of BW, Burning Empires, you can be playing out an entire military action between two armies, blow-by-blow. However, what you're resolving usually isn't, "Do I kill more of his troops than he kills of mine?" You're more likely resolving, "Do we take over the prison complex?" or "Do we get the enemy to surrender the Ultimax Device?"

In a smaller-scale fight in BW, you might be resolving whether the Black Knight surrenders the princess. By winning the fight (blow-by-blow, wound-by-wound), you're working towards those stakes. In the equivalent D&D battle, you can whittle down his hit points all you want, but short of killing him so that he has no say in the matter, it's the DM who ultimately decides if your overall goal comes to pass.

What if originally the adventure did not require me to capture the characters but halfway through the adventure I figure that situation X would make more sense if the characters were captured? Am I allowed to change the script to make the adventure make more sense or make it more enjoyable?

Well, in D&D, yes. I wouldn't be all that nuts about being forced into a course of action in order to serve the story, but I can't, by the rules, do anything to stop you.

I imagine that in the "agreement" game I could do something like this also. If it requires the player to agree to it, then either the player trusts me enough that he'll allow himself to be captured or the game has some clause that allows me to "break" the rules. Because in this specific adventure, if the players are not captured there is no adventure. If the game does not allow that then the DM is left to decide how to accomplish that.

If the adventure presupposes that the PCs have been captured, you'd probably be better off simply starting after the fact. "Okay, you're all sitting in a jail cell when..."

Some might call this fiat; I'd call it scene-framing or situation-building. For me, it would only be fiat if I'm presented with a false choice during the adventure, i.e., it seems like I can avoid being captured, but the DM is going to simply make that outcome occur no matter what I do. That doesn't seem very fun to me.

However, even in those games the DM is going to have to improvise.

I don't consider improvisation and fiat the same thing, though. Take the jumping on a roc example. Even if the DM has to make an ad-hoc ruling because there are no rules for roc-jumping, as long as the player has input, it's not fiat, by my definition.

Fiat: "You fail." or "You succeed."
Improv: "Hmm... Since there are no rules, how about you make a DC10 Jump to determine if you're making a controlled descent. Then, give me a melee touch attack, with a -4 if you don't make the Jump check. If you biff that, well, you're falling."

In the former example, I have no say whatsoever. In the latter, it's ultimately me and my character's stats that make the determination. For me, that's a lot more fun.

DM Fiat does not pose a greater risk of "enormous game-ruining potential" than using the Mystic-Thurge in a game.

We'll have to agree to disagree, then.

What if the rules consisted solely of: "Roll a d20. If it's high, something good will happen. Otherwise, something bad will happen. The DM makes the determination." That game is going to live or die by the DM. If you don't like what the DM declares, you're totally hosed. Start adding on rules that guide the determination and share the input around the table, and your hose-potential starts to drop.

 

[Brimshack]

Is it necessarily robbing a player if you throw a random variation at them? I guess to me it depends on whether or not the random event has some probability assigned to it. Simply deciding to change the situation after a key roll would strike me as unfair, but giving a plan-foiling twist a chance to occur seems quite reasonable.

Say a rogue does sneak past a group of sleeping gaurds with a successful move silently roll. The posibility that one of them gets up to take a piss is real enough, and it adds a certain flavor to the game. Too often I have seen players design very complex plans which rest on the assumption that all the enemy will be exactly where their assigned posts are at any given moment, and that no-one will be out of place whatsoever. But gaurds aren't (usually) robots; they are (usually) humanoids and that means they often do unexpected things. A little random variation in enemy behavior helps to keep the best laid plans in perspective.

Now, how to do it? It seems to me that at least part of the discussion over DM cheating a player out of a roll falls on the question of just how fuzzy is the GM thinking. I've seen a lot of GMs, and I know I've been guilty of this in the past, tell a player to make a roll and then decide what the results are afterwards. This is particularly tempting with social rolls such as Gather Information and Diplomacy. The player rolls a gather information, just looking for local gossip, and say gets aresult of 13. So the GM decides that's enough to learn that there is a plot to kill him but no more. If the player rolled a 19, would the GM give him more? Or would he decide that such an excellent result would mean he learned exactly the exact same amount of information? Trying to get into a castle, your Paladin rolls a Diplomacy check and gets a 12, so the GM decides that's good enough, they let you in. Were you to have rolled a 21, the same GM might have decided that was so excellent that the gaurds decide to let you in. ...the problem is the GM is deciding after the roll, and that increases the odds that the roll will simply confirm his own preferences with respect to what he wants to happen in that situation.

What I have been forcing myself to do lately is sketch out the results in shorthand, or announce the possibilities, if there is no call for secrecy, even if it takes a minute or so before making the player roll the dice. In the gather information and diplomacy examples above, I would scribble down a different result for DCs of 10, 15, 20, and 25. The result of 25 would usually be the optimum result central to the plot line, and if the player managed a 30 I would add something extra (which would then be made up). This wouldn't be shown to the player, but it's essential to me that I decide the results in advance of the roll, just so confirmation bias doesn't lead me to assign the same result for a broad range of die rolls. I think a lot of GM cheating the dice comes from just making the judgement calls after the dice is rolled.

When it comes to a completely random variation such as a gaurd being somewhere or doing something random and inconvenient, I would probably assign about a 10%-20% chance for something along those lines, perhaps even announcing it to the players, and roll it. This way the decision is above board and will cause less suspicion if a disaster comes up, plus it effectively reminds the players that they do live in a world where people go take a piss, suffer insomnia, or wake up for no apparent reason. The players will need to take such things into account, and if they fail to do so, I don't think it is necessarily robbing them to throw a twist that way from time to time.

 

[buzz]

If the DM decides to screw the player out of that success, he isn't really following the rules of D&D anymore, is he?
...
But you seem to be thinking that the GM randomly changing rolls and "screwing" the players is built into D&D, which is false. And then you base arguements off of that. If this is not what you are doing, then I've misunderstood you.

See, that's just it. In the examples I gave, the DM isn't altering the die rolls at all. He can run your Hide and Move Silently checks 100% by the book, and you can succeed at them all day long... but that doesn't stop him from having something happen that makes those successes fruitless.

And I'm not saying that D&D is built to screw players. I'm just pointing out that all you generally resolve in D&D is a specific task. "Did I move to point X without making noise?" "Did I open this lock?" "Did I jump 20 feet?" "Did I change his disposition to 'helpful'?" What each of those actions mean in terms of your intent is only accomplished through the accretion of successful tasks and some negotiation with the DM. Your intent is almost never included in the resolution mechanic.

If you want to resolve "Did I traverse the courtyard without alerting any guards?" or "Did I convince the guildsman to lend me his passkey?" or "Did my high-jump impress the milkman's daughter?", you need to go beyond the D&D rules, generally into "Will the DM let me do this?" territory. In each case, the DM can say:

* "You made your roll, but a guard suddenly runs into the courtyard on his way to the latrine and sees you."
* "You made your Diplomacy roll, so he's now 'helpful'... but not that helpful. Forget it."
* "You made the Jump check, but the young lady happened to be looking the other way at the time. Sorry."

Now, this is perfectly okay. I'm just trying to convey that BW, and similar games, do this differently. The intent is hardwired into the roll. Sure, you can argue that a GM could conceivably hose a player by messing with the stakes... but the player doesn't have to accept the stakes.

To me, it's a very different play experience.

My apologies, if you aren't enjoying debating this topic, I can stop.

As long as it's not bothering Crothian. I just felt bad about hijacking the thread.

If you ever get a chance to play something like BW or TSoY, I recommend it. It'll probably do a better job than I am of explaining all this.

 

[D'karr]

See, that's just it. In the examples I gave, the DM isn't altering the die rolls at all. He can run your Hide and Move Silently checks 100% by the book, and you can succeed at them all day long... but that doesn't stop him from having something happen that makes those successes fruitless.

But like has already been said that is not DM Fiat, that is the DM being an asshat.

 

[diaglo]

That might be it, I don't always know the reasons why things work like they do. I just know how they work.

written like a true OD&D(1974) convert.