Other than the GUmshoe system which seems to hand the players the answers on a plate, what other alternatives exist to the problem of players failing vital skill/stat rolls to pick up necessary clues?

A cop-out method for lazy GMs (such as I) is to distribute clues but not have a solution in mind.  The players will figure something out.  Maybe invent a new clue or two on the fly that demolishes their original theory and then validate then replacement theory as correct.  It's dirty trickery but if you can pull it off with a straight face the players will eat it up.

Primary: Don't make the roll to find the clue but to determine the level of detail about the clue.

Secondary: Don't make one clue the only possible link. Make spiderwebs of clues around important facts so that there is overlap and multiple way to get the info.
=

Sometimes on vital roll failures, I'll place false clues or clues which are not directly relevant or useful.

Quote from: jrients;328268A cop-out method for lazy GMs (such as I) is to distribute clues but not have a solution in mind.  The players will figure something out.  Maybe invent a new clue or two on the fly that demolishes their original theory and then validate then replacement theory as correct.  It's dirty trickery but if you can pull it off with a straight face the players will eat it up.

First of all doing this sort of thing myself, but it has problems.

For one thing, players aren't stupid and in particular if your player are also GM their own right, they know about these tricks too.

Secondly it kind of puts the whole pacing of the adventure on the GM's shoulders. Essentially the right answer, key to solving the mystery, becomes "when the GM is statisfied that the players have done enough". It devalues the players contribution to the game down to performing tricks to amuse the GM.

Again, I am not above doing this and it is probably okay if the investigation is just a sideshow to the main part of the adventure, but for an investigation heavy game I don't think it is a good approach.

Not that I have a good approach to suggest.

I make many of the clues interactions with people. You can't just "miss" a person. You can miss that something's up, but still get relevant information.

That and I don't assume that because the crime (or whatever) is over things stop happening. The suspect might try whatever he did again. Or someone in the know might try and retaliate. Or whatever he stole wakes up and gets out of hand. Stuff happens, and if the players don't figure out why this session, that's hardly got to be the end of the action.

Quote from: Ghost Whistler;328265Other than the GUmshoe system which seems to hand the players the answers on a plate, what other alternatives exist to the problem of players failing vital skill/stat rolls to pick up necessary clues?

I allow the PCs to fail to solve the mystery. Completely solves the problem of worrying about them missing clues.

A friend of mine came up with a distinction between "clues" and "evidence" in investigative games, which I now use all the time when I'm planning investigative scenarios.

"Clues" are bits and pieces of information about the situation at hand. Once the PCs accumulate sufficient clues, they should have a fair idea of what's going on. Importantly, that's sufficient clues, not all the clues. You don't need to interview every single janitor at NASA to work out that the Moon landings happened, for example. Investigative games bog down when the PCs get stuck and can't obtain enough clues to work out what is happening. This is frustrating for the players, because they're confused and can't get any further, and frustrating for the GM, because nothing is less fun than watching players walking around in circles. The solution is to make sure that clues are abundant, and aren't too tricky to obtain.

Pieces of "evidence" are special clues: they don't just tell you what's going on, they are actually useful in attaining the investigator's desired result. It's perfectly possible for PCs to know exactly what's going on, but be unable to do anything about it - this is especially common in games with a strong supernatural element, where the PCs can't exactly go to the police and tell them that Sir Rockham Sockham is killing his enemies via voodoo curses. For example, in a police procedural "clues" will let the PCs work out who the murderer is, but only "evidence" will be admissible in court. In a ghost-hunting horror game, the "clues" might let the PCs work out that the mansion is being haunted, but only the "evidence" will let them work out ways to lay the ghost to rest.

When I run investigative games, the challenge isn't so much in assembling the clues as it is in collecting enough evidence to get the result the PCs want out of the investigation; in general, I make the evidence harder to get to than the vanilla clues. In my experience, it's less frustrating for players if they find out what's going on but don't manage to work out how to get the solution they want; players accept failure more easily if they at least know the context and understand why they failed. Furthermore, once players have assembled enough clues to understand what's going on, this generally gives them pointers towards where they might go to find the evidence they need to bring the case to a successful close, so they're at least not bumbling around in the dark; if they fail to get the evidence but were able to at least have a stab at obtaining it, that's less OOC frustrating than never knowing what the hell they were trying to achieve in the first place.

My problem with the Gumshoe system is that it makes no distinction between these two types of information. In my investigative games, it's quite possible for the players to get to the end of the investigation (they're usually time-critical) with a fair understanding of what's going on but without the evidence they need to enact the optimal solutions. There are, of course, always fallback plans, but they're usually fraught with risk. (For example, in a Call of Cthulhu game the optimal solution might be to stop the cult summoning their dark god in the first place, and the relevant "evidence" might be pieces of information the PCs could use to disrupt the ritual or spoil the cult's plans. Fighting the sinister demon once it's summoned is of course a possibility, but you don't want it to be your plan A...)

Quote from: Greentongue;328275Primary: Don't make the roll to find the clue but to determine the level of detail about the clue.

Secondary: Don't make one clue the only possible link. Make spiderwebs of clues around important facts so that there is overlap and multiple way to get the info.
=

But they still have to find the clue to examine it for detail, surely. Do you just say 'you find some DNA', for example?

Quote from: Warthur;328285My problem with the Gumshoe system is that it makes no distinction between these two types of information. In my investigative games, it's quite possible for the players to get to the end of the investigation (they're usually time-critical) with a fair understanding of what's going on but without the evidence they need to enact the optimal solutions. There are, of course, always fallback plans, but they're usually fraught with risk. (For example, in a Call of Cthulhu game the optimal solution might be to stop the cult summoning their dark god in the first place, and the relevant "evidence" might be pieces of information the PCs could use to disrupt the ritual or spoil the cult's plans. Fighting the sinister demon once it's summoned is of course a possibility, but you don't want it to be your plan A...)

Actually, GUMSHOE does make this distinction, specifically. The core clues will get you to the end of the adventure, but not necessarily provide you with the extra information you might need to reduce risk or get a solution. Point-spend clues can give you this information in the form of special benefits.  What you call "evidence" is a subset of special benefits. In Trail, in particular, investigators can find themselves in deep trouble without the tools they need if they brashly follow the core clues otherwise unprepared.

As to the main question, my suggestion is, take a leaf from GUMSHOE and ensure that all essential clues do not require a roll, rather than fudging die rolls or trying to shoehorn the clue in another way.

Quote from: JimLotFP;328284I allow the PCs to fail to solve the mystery. Completely solves the problem of worrying about them missing clues.

Bingo.

Gumshoe fails for me in that it is a system designed to force players to answer a single riddle.  No matter what happens in the game all portions of the riddle will be collected regardless of player choices.

"Solve this mystery" is not the game for that RPG.  The game is "tell a story where your characters happen to be mystery solvers, maybe you can mystery/riddle which leads to the climax of the story".  Solving the mystery is not what the players are being asked to do.  Players act like they are solving a mystery, not put in the position where they must actually solve it. (i.e. hunt down clues, follow player thought up suppositions, test their theories, determine false leads, -essentially all the legwork)

But what if the mystery isn't solved by the players?  What if they don't gather the clues necessary for them to come to a conclusion?  In a convention game or tournament game the answer is Game Over.  Or traditional RPG the game keeps going and the player either eventually find all the clues and figure out the riddle/mystery or choose another goal.  

This probably sounds piss poor for those who want to tell a story.  IMO it's extraordinary for those who actually want to solve mysteries - one of the essential elements of pretty much every traditional RPG.

I'd suggest a combination of Greentongue and JimLotoFP.

Key clues should not be dependant on roll of the dice and if the players stuff up, let them fail.

s I have previously noted on this topic I run a Murder Mystery company that puts on MMs for guests at hotels and stuff.
Most of the key points to directing the investigators have been pointed out ie
Make a lot of clues people based hard to miss people
Don't  have just one clue to unlock the whole thing
Be prepared to give the players information from sources they choose to investigate (ie don't be too closed minded about your clues)
Make finding the clues easy but give extra detail for good rolls.
Use NPCs to do some of the grunt work.

Examples help to make things clear.
There is a murder in a flee-pit hotel room. The PCs are cops.

The GM thinks they will
i) isolate the crime room and bag up any items that look useful
ii) Ask the other guests if they saw anything
iii) As the hotel clerk about the victim

In fact the PCs walk in and find a couples of clues then head off missing out all of ii and iii and a lot of i.
As A GM when they are leaving I would have the CSI/SOCO guys arriving. If the PCs say nothing to them as the PCs leave the head SOCO will say "we will run over this place for anything else".
If they don't ask any people then have a beat cop say 'do you want me to interview the folks in the hotel?

You can then present a base level of all the information to the PCs in a dossier the next session whilst they high tail it to the Blue Moon Piano bar on the basis of a single match book.

The dossier should hint at things without spoon feeding. Subject A seemed suscpicious, etc and the CSI report should be straight evidence with no hint of interpretation but with a hint of additional studies that can be done. Yes they could look in more detail at that bllod splatter int eh bathroom and test to ensure it was the victims not the killers.
If they do nothing you can always release a fully study later on.

If the players aren't cops you can stil hit some of these points.

Sometimes player do the damnest things. I have a murder game once and a PC asked to get all the victims phone records fot the last 4 months and get a list of anyone they had called more than 10 times. The backstory was that the victim was a member of a witches coven and had been killed by a religious fantantic who's plan was to kill all 13 members, the PCs were supposed to guess by victim 4 or 5 through painstaking interviews and the like. Instead they ended up with a list of 20 people and when the 2nd murder occured they put them all in protective custody and looked for links. As a GM you have to adapt to that stuff.

Quote from: JimLotFP;328284I allow the PCs to fail to solve the mystery. Completely solves the problem of worrying about them missing clues.

Along these lines, you can adopt a story along the  lines of the Friday the 13th TV show (totally unrelated to the movies).  They would fail a fair amount.  They wouldn't even know of an evil artifact until they spot an odd death in the newspaper.  Then, 1-3 more people would die before they wrap it up.  
 
If PCs fail to figure things out, more bad things happen in a day or two  which create more opportunities to figure things out.  
 
I also use the web of clues approach so a missed bit of info doesn't bring them to a full stop, just not as on top of things as they could be.
 
In Call of Cthulhu, I sometimes make things weird an complicated.  I don't expect them to peice it together entirely.  Even if they kill the cultists, they are likely to still have unanswered questions.  That's ok as the Mythos is not supposed to make perfect sense.

Run the mystery as a tactical game so that process can be handled piecemeal. Abstract the investigation space as map, establish movement rules based on investigative skills, abstract the solution as a pawn to be moved on the map maybe, and play it a turn at a time, counting turns as time spent. When done, narrate the result by interpreting the board state (who's where, what tasks have been accomplished or failed, how long did it take to get there, etc.)