I started a discussion on this subject over at spielfrieks, and it generated mixed opinions -- on the one hand, Rick Thornquist, Mik Svellov, and Eric Hautemont (president of Days of Wonder, the publisher) all said that they've seen games where the correct accuser didn't win. On the other hand, Greg Alexnevikus (sp?) said he thinks the scoring system is broken. So, we have eminent people falling down on both sides on this one.
One response that interested me was a simple variant proposed by Matthew, who suggested adding a rule that when someone makes an accusation, everyone else circles the person they think is the culprit, and if the person making the accusation is correct, everyone who also had that culprit gets 1 point. I think this could nicely solve the problem zaiga sees with the scoring system.
-Jeff
This means that, yes, revelations play a part in strategy, but, no, they do not provide an alternative way of winning, you still have to find out who is the culprit so that you can actually end the game.
In a game where the object is to determine who the killer is, I would say that "finding out who is the culprit so that you can actually end the game" a perfectly good game-end condition.
I think the confusion here stems from imposing an attempt at an alternate win condition onto this scoring system, which may or may not be intended.
Also, there's one more thing that's not being considered. All of these arguements assume that all players are very smart- so smart in fact that even if they themselves have not determined who the killer is, they know for a fact whish revelations are correct and which are incorrect, they know for a fact how much information each other player has gleaned from their questioning, and they know for a fact weather or not an accusation made will be correct and will end the game.
In practice I presume these facts will not be so obvious. Suppose a player believes they know the killer, looks at the scores and determines that a correct accusation will make him the winner, and therfore accuses someone. Now suppose they had overlooked something, or misunderstood a piece of information. The accusation turns out to be wrong, and that information alone may clue someone else in on the correct murderer.
Let me liken this to Puerto Rico for a moment... the first few times I played PR, we thought the scoring was open information. This led to long-ish games, with people counting and re-counting each other's score to determine if they should try and end the game or prolong it. When we found that the scoring was in fact supposed to be hidden, the game became very different, as we then had to make an educated guess about our score relative to other players'. And lo and behold, sometmes we were wrong. Sometimes people would end the game only to find that they had lost by 1 VP.
This analogy may not be perfect when applied to Mystery of the Abbey, as if you are correct in your accusation, then you know weather other peoples' revelations are right or not. The point is that when you hide some of the information, the game gets a lot more complex. In PR it's possible to track how many points each opponent has receieved from shipping, but somehow it still doesn't always happen.
- Seth