I think the experience of players seeing what is meant to be a secret map is not that uncommon. I've had it happen back when I used to run somewhat map-heavy Rolemaster.
I don't think I've ever changed a map in response. (I could be wrong - I'm recalling play from over two decades ago - but the recollections are stated sincerely.) Sometimes I haven't worried - I've never run a game, as best I can recall, where the map is everything. But I'm pretty sure there must have been occasions when I have asked players to politely ignore what they've inadvertently seen. Of course it's not ideal to metagame reveal a secret map (or any other secret) if part of the goal of play is for the players to discover it via play. But it can happen, and if it does I don't really see the point of debating whose "fault" it is.
Often players - at least those I play with - will engage in the sort of decision-making process I've just described themselves, without any GM request to that effect. They will consider what other information - ie information that was not just a metagame reveal - that they as their PCs had that would have influenced their decision about where to go, and will act on that My players will self-enforce "no metagaming" in other situations, too - eg declare actions that they believe make sense from their PCs' perspectives which are less total than the players' perspectives. In non-gamist play I don't see this as a big deal, because there is no "self-hindering" involved (given that the play is not gamist and so not aimed at beating the dungeon/whatever). If a player wants to do this it's no skin off my nose as GM.
This is part of why I'm so relaxed about having maps on open display, such as in my most recent session - because we're not playing a very gamist game, and to the extent that there are "victory conditions" they're not really map-based, and so the players have no real incentive to study the map for metagame clues like unique vs repeated numbers. The action will be brought via GM control over scene-framing, and the point of the map is to provide an expository aid and a constraint on/context for action declaration within a framed scene.
Robin Laws wrote about this sort of thing in an essay that is part of the Over the Edge rulebook (p 193 of my 20th anniversary edition):
If a player wants to self-enforce such a rule then it's even simpler!
I don't think I've ever changed a map in response. (I could be wrong - I'm recalling play from over two decades ago - but the recollections are stated sincerely.) Sometimes I haven't worried - I've never run a game, as best I can recall, where the map is everything. But I'm pretty sure there must have been occasions when I have asked players to politely ignore what they've inadvertently seen. Of course it's not ideal to metagame reveal a secret map (or any other secret) if part of the goal of play is for the players to discover it via play. But it can happen, and if it does I don't really see the point of debating whose "fault" it is.
Often players - at least those I play with - will engage in the sort of decision-making process I've just described themselves, without any GM request to that effect. They will consider what other information - ie information that was not just a metagame reveal - that they as their PCs had that would have influenced their decision about where to go, and will act on that My players will self-enforce "no metagaming" in other situations, too - eg declare actions that they believe make sense from their PCs' perspectives which are less total than the players' perspectives. In non-gamist play I don't see this as a big deal, because there is no "self-hindering" involved (given that the play is not gamist and so not aimed at beating the dungeon/whatever). If a player wants to do this it's no skin off my nose as GM.
This is part of why I'm so relaxed about having maps on open display, such as in my most recent session - because we're not playing a very gamist game, and to the extent that there are "victory conditions" they're not really map-based, and so the players have no real incentive to study the map for metagame clues like unique vs repeated numbers. The action will be brought via GM control over scene-framing, and the point of the map is to provide an expository aid and a constraint on/context for action declaration within a framed scene.
Robin Laws wrote about this sort of thing in an essay that is part of the Over the Edge rulebook (p 193 of my 20th anniversary edition):
When viewing role-playing as an art-form [ie analogous to improvised theatre, as per earlier material in Laws' essay), rather than a game [ie aimed primarily at achieving victory conditions], it ecomes less important to keep from the players things their characters wouldn't know. When characters separate, you can "cut" back and forth between scenes involving different characters. . . .
The price of this is allowing players access to information known to PCs other than their own. But it's simple enough to rule out of play any actions they attempt based on forbidden knowledge.
The price of this is allowing players access to information known to PCs other than their own. But it's simple enough to rule out of play any actions they attempt based on forbidden knowledge.
If a player wants to self-enforce such a rule then it's even simpler!