How much campaign lore should players be expected to remember?

Two different memories, one campaign

A lot of table friction over "forgotten" lore actually comes from conflating two separate things: what the player remembers, and what their character would know. A player who forgets a merchant's name after three sessions is completely normal — but that merchant's character has known the party for months, and would greet them by name without hesitation. When the GM treats the player's memory as the limit of what the character knows, ordinary forgetting turns into an awkward stall at the table.

The practical fix is simple: if the character would obviously know it, just tell the player. "The dockworker who owes you a favour" is a character reminding the party who they're talking to, not a hint the player failed to earn. Reserve actual memory tests — rolls, deliberate withheld information — for moments where forgetting is the point, not as the default response to a name slipping a player's mind.

What's reasonable to expect, and what isn't

A rough calibration for a table meeting weekly. Long gaps between sessions lower all of this further.

• Reasonable: the current goal What the party is actively working toward right now. If nobody at the table can state this in a sentence, that's a session-start problem to fix, not a player-memory problem.
• Reasonable: whoever is active this arc The two or three NPCs and factions actually driving the current situation. Players will hold this if it's reinforced session to session.
• Reasonable: roughly what happened last time The broad shape of the last session, especially with a short recap. Exact dialogue and minor details are not reasonable to expect back.
• Unreasonable: names mentioned once, sessions ago A minor NPC or location named a single time ten sessions back and never reinforced since. If it matters again, reintroduce it — don't expect it recalled cold.
• Unreasonable: chronological order of unrelated events Players track cause and effect for things that mattered to them, not a timeline of everything that happened. Don't test the order of events that weren't consequential at the time.

Normal forgetting, overload, or disengagement

Not every forgotten detail means the same thing. Telling these apart changes what's actually worth fixing.

• Normal forgetting An isolated detail slips, a quick reminder fixes it, and the table moves on. This is the default and doesn't need a system-level response — it needs a one-sentence restatement.
• Information overload Forgetting happens immediately after the information was given, and it's widespread across the table. This usually means too much was introduced at once without a hook to attach it to — the fix is trimming what you introduce per scene, not testing recall harder.
• Genuine disengagement Forgetting is one symptom of a broader pattern: missed basic developments, no questions asked, no visible investment in the current goal. This is worth an honest table conversation, not a better recap framework — a memory system can't fix disengagement.

Calibrating expectations to your table

• You can name the two or three NPCs/factions actually active in the current arc, and so can most of the table.
• When a character would obviously know something, you restate it instead of waiting to see if the player remembers.
• Forgotten details get a quick, in-character reminder rather than being treated as a quiz.
• You've lowered expectations appropriately after any long gap since the last session.
• When forgetting is widespread right after you introduce something, you're trimming what's introduced, not repeating it louder.
• You can tell the difference between an isolated forgotten name and a wider pattern worth an honest conversation.