But humor me. What does it mean to actually solve it?
To arrive at the correct conclusion, through examining the evidence, identifying what is valid and accurate, and then applying a mixture of deductive, inductive, and abductive reasoning, filtering out who could or could not have committed the crime. If they successfully solve it, then that mixture of those reasoning methods will point them to the person who "really did" do the thing.
At least for me, to "actually solve [a mystery]", there needs to be an answer independent of any desires, preferences, or creations of the people doing the solving. A riddle without an answer isn't a riddle, it's a rhetorical question that someone might take as non-rhetorical. If a "puzzle" lacks a singular solution, it isn't really a
puzzle anymore, and is instead a toy (or possibly a game). At the center of every mystery is exactly that, a puzzle we must pursue by putting the pieces together. To have a mystery continuously shaped by player declarations,
even if those declarations follow rules, would be like having a jigsaw puzzle where the pieces don't actually fit into a pattern until we
decide they fit into a pattern, and then all of a sudden that was the pattern that was "always" there.
It is very important, here, that the conclusion is
and always was correct. To have a final result that did not have any correct answer until
after we declared it is a serious problem for "actually solv[ing]" it. Whether it was declared collaboratively or not is irrelevant; a Borromean linkage is still a linkage even if no two rings are linked to each other.
Also, I'm not actually saying it has to be foreknown. In a Clue game, nobody knows who the real killer is,
but there is in fact a real killer. It isn't decided by anyone (it's done with cards put into sleeves, at least in the versions I played as a child), but it is in fact something specific: a specific weapon, a specific room, a specific killer. The players can acquire evidence, mostly in the negative (observing what cards it
couldn't be, by finding them through play) in order to narrow down what it therefore
must be. There is no possibility in this context that someone could--to quote the Cthulhu Dark rules--"think...it more interesting if you failed". There can be no sudden swerve from solving mystery A to solving mystery B and leaving mystery A as a dangling thread. (Because yes, I did actually read both the Cthulhu Dark rules and the linked play-summary thereof.)
For instance:
The wealthy industrialist, Victor Harrow, is found dead in his locked study, shot once in the chest. There were four people in the house that night: his estranged daughter Clara, his business partner Lewis Grant, the family maid Sofia, and the private investigator Harrow had hired the week before, named Mason Drake. Unknown to the players, the truth is as follows: Lewis Grant killed Harrow because he discovered Harrow was planning to cut him out of the company. Grant entered the study using a spare key he had secretly copied, shot Harrow during an argument, wiped his prints, and left through the servants’ entrance unnoticed. Clara heard the shot but assumed her father had taken his own life. Sofia found the body but panicked and ran without alerting anyone. Mason Drake was outside on a phone call at the time of the murder.
There are fingerprints on the copied key hidden in Grant’s coat pocket. There are financial papers in Harrow’s desk showing he was planning to disinherit Grant. Grant’s shoes have traces of mud from the servants’ entrance path. Sofia can be found later and will nervously reveal that she saw Grant leave through the side door.
All of this is sitting in the GM’s notes, pre-written.
What does it mean to actually solve it?
The PCs examine the area, try to find as much evidence as they can, try to analyze that evidence as effectively as they can, and then, on the basis of their analysis (which may not be correct!), try to determine the correct answer to the mystery. Since there is a fact of the matter--here, Grant really was the killer all along, no matter what collaborations or contributions a person may have put in, regardless of the source e.g. dice-slinging or card-drawing or GM-authoring or
whatever--the players can truly be right or wrong (or right about some things and wrong about others). So, for example, the players might correctly name Lewis Grant the killer (and presumably extract a confession from him in front of police, Poirot-style), without ever locating the murder weapon. Or they might fail, and name Sofia the killer because they missed the key and mud on his shoes, but correctly determined that Sofia had seen the body and failed to report it. Or they could just completely fail to solve it at all.
The "ideal" solve, the whole kit and caboodle, would be more or less as a typical mystery-novel solution works, where the detective lays out the chain of events, specifies the pieces of evidence which narrowed it down to that result and only that result, and conclusively demonstrates Grant's guilt before the police, ensuring that he is duly convicted in a court of law by a jury of his peers etc. etc. and all the innocent suspects go free.
