He could be, but he isn't because we are establishing that the GM is using realism as the or one of the main criteria for the decisions. He is basing that decision on what he considers to be realistic. So realism is the thing driving it.
Yes, I understand that. My point is that though realism is a goal, it's still up to the GM. It's the GM's choice, not realism's choice. So whatever consequences there may be, they are the responsibility of the GM. They've made a choice. They should own it.
Look if I am a player and the GM makes a call based on realism, and it negatively impacts my character in some way, I am not going to be bothered the way I would if he just made the call out of a desire to negatively impact my character. In the former, the GM is giving me a game experience I was looking for. One where the world feels like a believable place, and one byproduct of that is some of his calls on realism aren't going to go my way. That doesn't mean he is morally culpable. That is a baked in understanding if the GM is prioritizing realism: there is a risk that what the GM finds most realistic, might go against what the players are trying to do. And that can be important in player driven play of the type that we are describing.
Again, all of this is fine if it matches player-GM expectation. It's only an issue if there is a misalignment there. Which, if we're talking about player-driven play (I'm gonna use that term here instead of sandbox), then it would seem that in these instances that you're talking about, it's that realism is prioritized over player-driven play.
Especially when an equally realistic option that also supports player-driven play is likely.
If I have knowledge the players do not it comes down to a few of things.
- Is it possible for the characters know this? Some things are just unknowable, either lost to time or a secret so closely guarded that it cannot be discovered.
- If it is possible for the characters to know something, is it potentially general knowledge? This is where appropriate checks come in because it's uncertain whether or not someone would know something. There's also a mix, the player may recognize a symbol for a specific deity but the fact that the symbol has been slightly altered may be rare knowledge.
- If it's not general knowledge could the information be uncovered and have the characters had opportunities to uncover the information? Did they successfully pursue those opportunities to uncover more information?
But is the GM deciding how the NPC reacting? Of course. That's how D&D and similar games work, the GM is making decisions all the time even if that isn't true of other systems. It's not inherently better or worse, it's just my preference for games I play.
I'll address each of your points and how they relate to my overall point.
- Who decides what is unknowable or lost to time or a secret guarded so closely that it cannot be discovered? The GM.
- Yes, there are ways to offer the information to the players, or at least the chance to learn the relevant information. Who decides if there are visual cues like an interaction with another NPC or a religious symbol? The GM.
- Yes, ability checks may help here. Who decides to call for such a check? Who sets the DC? Who decides what information is gained on a successful roll? The GM.
These are all the decisions of the GM. To try and avoid responsibility for all these choices and instead attribute it to "realism" is just wrong. The GM is responsible for so much of it.
Now, if this works for you, again... it's not a problem. But the game does not have to be played this way.
Personally, I want to know how the players will deal with the un-bribeable guard. I don't want to wonder how they'll even find out he's un-bribeable. I'll just share the information. I'm generous with information because I recognize the players' understanding of the situation in play is entirely dependent on my ability to describe it to them. So I share a lot. Why not? What's to be gained? A long drawn-out interaction with a guard? Why? Let's get to the good stuff.
Well this is the sticking point. In the guard example, in my version, the players can learn the information by investigating the guard schedule. But it seems to me that in your version, the GM just shares it with them without them putting in any effort. That seems...odd to me, because you'll have to contrive a route for them to learn the information other than investigating. (Maybe the guard is someone they knew as a kid? They play in the same poker game?).
Indeed, it seems opposed to player driven play to me, because the players don't have to play the game to get the information they need.
Play what game? Twenty questions? Is this what you tend to focus on when you play?
This is the kind of "conventional wisdom" that after many, many years I've come to eschew. There is no reason to hold back every shred of information as if it's a major discovery. You can indeed just narrate things like "you can tell just by the way he looks that this guy isn't going to take a bribe" or "you can just tell this guy is lying".
This is the difference between a principle that says something like "when the outcome of an action is uncertain, roll the dice" and "only roll dice if there are interesting consequences for failure".
