"Skip to the END": Handling players bypassing the setup for the main villain

[veekie]

It can be a serious case of confirmation bias for the negative reporting. Ham-handed improvisation is of course bad, and spotted by everyone. If you are combining Improvisation with prior planning, then improvisation is the art of knowing which parts of the plans can be recycled immediately, which parts cannot and should be shelved.

So taking the kidnap scenario, dissected.
First, the premise of the 'kidnapped' daughter. The PCs hear rumors of orc troubles, and then the distressed parent asks them for help. The village in addition has a reward for getting rid of the orcs in general.
The location of said orcs are currently undefined in a large hilly/forested region.

Skill check results usually go for depth of information over breadth, so a good Gather Information check would net you rumors of the following, depending on what they asked about. Remember, Gather Information cannot give you information that the social unit does not possess or is unwilling to tell, it only gives you rumors, how to verify them(that is, a high roll gets you reliable rumors while a low gets you bullshit and brag). I personally go with 1 'fact' per 5 points past the required DC.
The girl
-Her relationship with her parents(poor, also clue to resolution of event).
-Her rumored preference for orcs.(clue to resolution of event)
-Her upcoming engagement to a merchant
-Where she was last seen before her disappearance(begin tracking orcs)
-Favorite places(you can spin these off into minor scenarios where if they look, they can find evidence of her being in contact with the orcs, hidden love letters etc. Also a way to find the orcs)

The orcs
-Their territory and camps in the region
-That they've been there a long time
-Recently sighted NEAR a variety of locations(especially the leader's kid who's out dating with the girl)
-Their general treatment of people(more friendly than usual but territorial, mostly livestock and property theft)
-Their more well known personages(gets you info on orc leaders, which a subsequent investigation can reveal details on)

So at this point, the players have a cluster of potential locations to visit and search, you then lead from these locales to their current encampment. If the PCs derail, then you improvise, keeping in mind that locations and times are fluid. If they sneak into the camp and yoink the girl, you start up a sideplot with the girl trying to escape from her rescuers, etc. If you have a particular location in mind(say a haunted forge), then you can place it in a variety of places. Maybe the girl has been sneaking off to nookie with the leader's kid in the haunted place and its where they find her. Maybe the orcs are camped around it while they try to get the trapped kids out. Maybe the orcs aren't anywhere near it, but the PCs run across it when they are in need of rest and shelter from pursuit. Heck, maybe the merchant the girl is supposed to be engaged to is an evil ghost residing in the forge.

Thats improvisation. You don't skip planning in favor of it, you supplement planning WITH it.

[InnaBinder]

It can be a serious case of confirmation bias for the negative reporting. Ham-handed improvisation is of course bad, and spotted by everyone. If you are combining Improvisation with prior planning, then improvisation is the art of knowing which parts of the plans can be recycled immediately, which parts cannot and should be shelved.

Please clarify what you're saying here about confirmation bias, so that it I do not (mis)interpret it.

Skill check results usually go for depth of information over breadth, so a good Gather Information check would net you rumors of the following, depending on what they asked about. Remember, Gather Information cannot give you information that the social unit does not possess or is unwilling to tell, it only gives you rumors, how to verify them(that is, a high roll gets you reliable rumors while a low gets you bullshit and brag). I personally go with 1 'fact' per 5 points past the required DC.
The girl
-Her relationship with her parents(poor, also clue to resolution of event).
-Her rumored preference for orcs.(clue to resolution of event)
-Her upcoming engagement to a merchant
-Where she was last seen before her disappearance(begin tracking orcs)
-Favorite places(you can spin these off into minor scenarios where if they look, they can find evidence of her being in contact with the orcs, hidden love letters etc. Also a way to find the orcs)

The orcs
-Their territory and camps in the region
-That they've been there a long time
-Recently sighted NEAR a variety of locations(especially the leader's kid who's out dating with the girl)
-Their general treatment of people(more friendly than usual but territorial, mostly livestock and property theft)
-Their more well known personages(gets you info on orc leaders, which a subsequent investigation can reveal details on)

I am having difficulty not reading this as the DM diplomatically telling the players "it's impossible for you to use the skills on your sheet to do what you want them to do."  I find that troubling because article after article, anecdote after anecdote, indicates that good gaming sessions happen when the DM says "Yes" to the players, and bad/railroad-ish gaming sessions happen when the DM frequently says "No" to the players.  If you could clarify how this is saying "Yes" to the players for using Gather Information, rather than saying "No" to the players by politely telling them the Gather Information skill they invested in (or the Bardic Knowledge check, for that matter) doesn't allow them to gather the information in that way, I'd appreciate it.

So at this point, the players have a cluster of potential locations to visit and search, you then lead from these locales to their current encampment. If the PCs derail, then you improvise, keeping in mind that locations and times are fluid. If they sneak into the camp and yoink the girl, you start up a sideplot with the girl trying to escape from her rescuers, etc. If you have a particular location in mind(say a haunted forge), then you can place it in a variety of places. Maybe the girl has been sneaking off to nookie with the leader's kid in the haunted place and its where they find her. Maybe the orcs are camped around it while they try to get the trapped kids out. Maybe the orcs aren't anywhere near it, but the PCs run across it when they are in need of rest and shelter from pursuit. Heck, maybe the merchant the girl is supposed to be engaged to is an evil ghost residing in the forge.

The fact that the quoted portion here uses "you then lead" followed immediately by "If the PCs derail" indicates, by my reading, that it's a railroad.  Keeping "locations and times. . .fluid" makes it Shrodinger's Railroad, but a railroad nonetheless. (Making them fluid at this point also merely sets those locations in stone for the future, as well, necessitating good on-the-spot notations that don't disrupt the flow of the session, unless the DM contrives to never use this particular stretch of the game world again).

I did not miss Rejakor's response, but it seems to me that much of the above applies to his commentary as well, and I didn't wish to seem like I was harping on the same things any more than I probably already do.  If you feel any particular point in his reply was not addressed adequately above or in preceding comments, please do say so.

[sirpercival]

I am having difficulty not reading this as the DM diplomatically telling the players "it's impossible for you to use the skills on your sheet to do what you want them to do."  I find that troubling because article after article, anecdote after anecdote, indicates that good gaming sessions happen when the DM says "Yes" to the players, and bad/railroad-ish gaming sessions happen when the DM frequently says "No" to the players.  If you could clarify how this is saying "Yes" to the players for using Gather Information, rather than saying "No" to the players by politely telling them the Gather Information skill they invested in (or the Bardic Knowledge check, for that matter) doesn't allow them to gather the information in that way, I'd appreciate it.

It's the DM's job to determine what information is available for the players to discover.  If you let the players dictate what they find out with a single skill check, you are allowing the players to DM instead of yourself.

[Mooncrow]

fuck it.

[InnaBinder]

How is getting relevant clues to the next stage of the plot "useless", out of curiosity?   

Because "useless" is in quotations, I presume it's a direct quote.  If I used that particular word, could you point me to the passage so that I can clarify for context?  Thanks.

[Mooncrow]

Paraphrased, not quoted, my bad.  But, what else does "Gather Information skill they invested in (or the Bardic Knowledge check, for that matter) doesn't allow them to gather the information in that way" mean in that context?

If there's a single skill being used to break your game, you un-break the skill.

Also, I find it fascinating how many out of context quotes from different game systems you throw around - do you play a lot of systems, or are these just phrases you hear online? 

[InnaBinder]

Paraphrased, not quoted, my bad.  But, what else does "Gather Information skill they invested in (or the Bardic Knowledge check, for that matter) doesn't allow them to gather the information in that way" mean in that context?

If there's a single skill being used to break your game, you un-break the skill.

Also, I find it fascinating how many out of context quotes from different game systems you throw around - do you play a lot of systems, or are these just phrases you hear online?

What I meant - and understood it to mean when I paraphrased other folks in that way - was "that skill does not do what you think it does.  There is no skill to do what you want to do.  Yes, I know what it's called.  Language is odd that way, sometimes."  I apologize for any failure on my part to convey contextual meaning.

Un-breaking a skill beforehand is fine.  Un-breaking it in the middle smacks of Stealth Nerfs and an unprepared DM, to me.  Also, if (non-Epic use) skills are breaking the game, spells are almost certainly going to make it go FUBAR, I'd think?

I'm not sure what you mean by "out of context quotes from different game systems;" I've played a bunch of different systems over the course of more than 30 years* (gulp), but I haven't deliberately been using specific systems other than a sort of generalized 3.X, as the system that I figure has the largest fan base on this site.

*more than 30 years is not intended to be any sort of Appeal to Authority.

[Kajhera]

Well, at a certain basic level, using a skill to Gather Information - shouldn't let them bypass relevant information. If you're worried they'll miss certain plot-relevant details but you want to give them the specific results they are after... it's not extraordinarily strange to put the plot-relevant details as bonus information in the gather information check.

Heck, my DM would throw in a completely random tidbit anytime we went after something, just for fun and occasional relevance...

[Mooncrow]

Paraphrased, not quoted, my bad.  But, what else does "Gather Information skill they invested in (or the Bardic Knowledge check, for that matter) doesn't allow them to gather the information in that way" mean in that context?

If there's a single skill being used to break your game, you un-break the skill.

Also, I find it fascinating how many out of context quotes from different game systems you throw around - do you play a lot of systems, or are these just phrases you hear online?

What I meant - and understood it to mean when I paraphrased other folks in that way - was "that skill does not do what you think it does.  There is no skill to do what you want to do.  Yes, I know what it's called.  Language is odd that way, sometimes."  I apologize for any failure on my part to convey contextual meaning.

Un-breaking a skill beforehand is fine.  Un-breaking it in the middle smacks of Stealth Nerfs and an unprepared DM, to me.  Also, if (non-Epic use) skills are breaking the game, spells are almost certainly going to make it go FUBAR, I'd think?

I'm not sure what you mean by "out of context quotes from different game systems;" I've played a bunch of different systems over the course of more than 30 years* (gulp), but I haven't deliberately been using specific systems other than a sort of generalized 3.X, as the system that I figure has the largest fan base on this site.

*more than 30 years is not intended to be any sort of Appeal to Authority.

So, you also want to make this change in the middle of a game without actually changing anything?  For the record, I said nothing about stealth-nerfing, feel free to tell them that they don't get to gain information for several steps ahead, and that the local innkeeper is no longer going to know the exact room of the dungeon that the orc chief hangs out in.  Or, if you really don't want to make the change mid-game, put up with the encounter skipping until this one is over, and put them into practice next game. 

The idea of always saying yes to players is pretty much a concept from Burning Wheel that has caught on like wildfire throughout the community.  But the thing is, it's not really about always saying yes, the concept is "say yes, or roll the dice"  - if something the player asks to do is unimportant, then go ahead and tell them that they can do it, without a dice roll.  If it is important, figure out what roll they would use to do that thing.  If the is no real relevant roll - the example in the book is where they're getting ready to break into a house and a player ask "did they leave a window unlocked?" BW has the Die of Fate, where they roll a d6 and on a 1, the answer is to the player's benefit. 

So, as you can see, it's not really about always saying yes.  It's about saying yes to the inconsequential, and always letting them try to do something.  But it's not an excuse to break the game. 

[veekie]

Please clarify what you're saying here about confirmation bias, so that it I do not (mis)interpret it.

Confirmation bias is a problem because good improvisation cannot be distinguished from planned plot. When its ham handed improvisation, it can be detected, and at the same time, often for dubious end goals(generally speaking, attempting to preserve a plan that has gone down in shambles). You improvise to facilitate and reuse unused components.

Examples(both done personally, at bad times and at good)

Bad Improvisation - NPC shows up to plot-dump on the players at the end of downtime. Players unexpectedly brutally mock her so she throws down WAY ahead of schedule, and her contingencies to avoid dying go off out of sequence(that is they wind up being cheap as hell). One of the worst improvs I've done.

Good Improvisation - Original scenario is to go down a curved slope(its a crater, objective in the middle) with  a variety of obstacles and a hard to kill flying dragon-skeleton harrying them along the way. The second stretch was greased with a poisonous slime.
So the players decided...why not just create a sled and toboggan down to the goal instead of going down slowly and being harried the whole time by the undead. The entire stretch had to be improvised into a single chase scene, and the nature of obstructions revised(I basically made a 10-item long table of hazards with open ended solutions while I distracted the players with describing how they were going to make the sled).
They wind up sledding down a slope with undead, traps and difficult terrain while a giant flaming dragon skeleton chases them. 

The only true difference is time taken, since planning involves making decisions about the game before the game, while improvisation is performed while in the hot seat. Both scenarios draw directly from your mind and inspirations, but planning has the advantage of being performed in a low pressure environment as well as allowing for checking for conflicts, while improvisation has access to more powerful inspirations(theres no inspiration as narratively powerful as taking a completely unexpected player idea to its conclusion) and is less likely to become entrapped in a dead end.

As to the skill in particular, the skill itself actually does lend to that purpose.

An evening’s time, a few gold pieces for buying drinks and making friends, and a DC 10 Gather Information check get you a general idea of a city’s major news items, assuming there are no obvious reasons why the information would be withheld. The higher your check result, the better the information.

If you want to find out about a specific rumor, or a specific item, or obtain a map, or do something else along those lines, the DC for the check is 15 to 25, or even higher.

Note that it works primarily on general and public knowledge, just as knowledge is supposed to apply mainly to factual and theoretical knowledge. If theres any reason why the information wouldn't be given out, the check might not work at all.

Just Say Yes is another concept that had been oversimplified. Mooncrow already described the generalities so I won't repeat that part, but the other key factor is when you roll the dice, you don't just succeed or fail, because that just means the players would roll again until they succeed. You roll, and whether it succeeds or fails, you get results, possibly misleading or even harmful results, but still, results.

Additionally, unless its down to the climatic scene, it is a better idea to have a roll contribute or take away from a goal, rather than outright resolve it all or fail it all. This gives more room for the group to apply their capabilities, and also greater space for expansion into subplots.

[Rejakor]

The base motivation behind 'say yes' is to stop DMs saying 'no you can't do that' and replace it with 'yes you can TRY to do that'.  That keeps games moving, even if it leads to outcomes the DM wasn't expecting.

It's a way to stop yourself railroading, basically.  And it's a fact of human psychology that the more you say 'yes' the more interested and excited people get.  In improv drama, it's called 'accepting offers', i.e. when someone says 'i'm the mailman' you immediately accept that that person is a mailman and converse with them as someone who would be talking to the mailman.  And when you start talking to them as a police officer pulling them over for setting fire to mailboxes, they accept that they as the mailman have been doing this, or at least that you are a police officer who thinks that they did.  The principles the same, but in tabletop we have dice that we roll as well, so you accept that someone can try instead of accepting that they automatically do whatever they say they are doing.

Not letting people automatically do anything they want (fly, get information from people those people don't have, take six turns in succession) if they succeed on a roll isn't not-railroading.  Railroading is when someone wants to do something reasonable, and makes the roll, and then you don't let them do it for bullshit reasons.  The reasons are the important bit.  If it's part of the world and they didn't know about it, it could look like railroading, although later they'll find out that it isn't.  But if you do it just because you have an encounter behind door A, and they chose to scale cliff B, then that is railroading.  They shouldn't get to just jump into the bottomless chasm and expect because they rolled a 20 on jump that they land perfectly fine.  It's not railroading to have someone go splat if they do that.  And it's not railroading to not have the townsfolk know where the frequently moved orc camp is (although they might know some forester who might know and the PCs could track him down.. I still don't get why bypassing the haunted forge encounter is a big deal).  Essentially, setting up the world, as long as you're not doing it to specifically block reasonable ideas, is perfectly fine.  The thing that's wrong is when you change the world or the DCs of the check or whatever specifically to force the PCs to do one thing.

[saethone]

I read through the first couple pages but if the conversation has changed between page 3-6 sorry I didn't catch i

To me its not a question that can be answered "Always do this" because its very dependant on whats at the end.

For example, if you have a BBEG who insanely outlevels the players, he may kill them for amusement, or he may treat them like treat a fly - just swat it away and go on with his day, or he may even try to recruit them. Likewise, a beast won't have a thought process and will just see dinner.

That's not even accounting for "ends" that don't involve a BBEG.

I guess, in general, I just roll with whatever would logically happen in the world. If the players aren't smart enough to pick up the clues that this guy can eat you alive, then it may be a TPK. Its never occurred that way in my group, but if it did I doubt they'd be mad. The times characters have died in our games, the players admitted it fell back to a mistake they made. nobody has ever blamed it on the GM or bad luck or anything, so it seems to work, so long as they understand fair is fair, and the GM plays NPCs as the NPCs would act in that situation.

[InnaBinder]

The base motivation behind 'say yes' is to stop DMs saying 'no you can't do that' and replace it with 'yes you can TRY to do that'.  That keeps games moving, even if it leads to outcomes the DM wasn't expecting.

It's a way to stop yourself railroading, basically.  And it's a fact of human psychology that the more you say 'yes' the more interested and excited people get.  In improv drama, it's called 'accepting offers', i.e. when someone says 'i'm the mailman' you immediately accept that that person is a mailman and converse with them as someone who would be talking to the mailman.  And when you start talking to them as a police officer pulling them over for setting fire to mailboxes, they accept that they as the mailman have been doing this, or at least that you are a police officer who thinks that they did.  The principles the same, but in tabletop we have dice that we roll as well, so you accept that someone can try instead of accepting that they automatically do whatever they say they are doing.

Not letting people automatically do anything they want (fly, get information from people those people don't have, take six turns in succession) if they succeed on a roll isn't not-railroading.  Railroading is when someone wants to do something reasonable, and makes the roll, and then you don't let them do it for bullshit reasons.  The reasons are the important bit.  If it's part of the world and they didn't know about it, it could look like railroading, although later they'll find out that it isn't.  But if you do it just because you have an encounter behind door A, and they chose to scale cliff B, then that is railroading.  They shouldn't get to just jump into the bottomless chasm and expect because they rolled a 20 on jump that they land perfectly fine.  It's not railroading to have someone go splat if they do that.  And it's not railroading to not have the townsfolk know where the frequently moved orc camp is (although they might know some forester who might know and the PCs could track him down.. I still don't get why bypassing the haunted forge encounter is a big deal).  Essentially, setting up the world, as long as you're not doing it to specifically block reasonable ideas, is perfectly fine.  The thing that's wrong is when you change the world or the DCs of the check or whatever specifically to force the PCs to do one thing.

When presented with a cliff, or a similarly obvious, physical encounter, the players will understand more or less intrinsically what is and isn't possible to do with their skill check beforehand; it's built into the description of the thing.  Social skills, excepting odd corner-cases, don't work that way; there's no obvious 'difficulty meter' of any sort laid out for them (certainly not with the info we've sketched on the example encounter so far) to indicate what a reasonable chance of success entails.  That means it's a natural psychological reaction on the part of the players, when they roll exceptionally well and get no meaningful result, to assume that the DM is simply arbitrarily negating their use of the skill, and, by inference, their skill choices.  Arbitrarily negating PC choices is deprotagonizing, and as such, should be avoided in general per an old thread regarding a paladin and a Drow slaver - even when the deprotagonizing result is based on perception.

Reasons why the haunted weapons shop may be important (I have to qualify with "may" because this example has been an organic, off-the-cuff one rather than scripted) include:

• Provided the encounters helpful to allow the party, or some significant fraction thereof, to level before encountering Smells-Like-Farts.  Whether or not the party has access to Haste or Greater Magic Fang could be significant mitigating factors in the difficulty of said encounter.
• Provided the party access to magic weapons and/or materials and instructions for upgrading their own weapons (it is a weapons shop, after all), which, again, can greatly influence the nature of their encounter with Smells-Like-Farts.
• Provided evidence hidden from the townspeople (or not obtained via Gather Information) to indicate Smells-Like-Angst and the daughter have been busy trying to find out where half-orcs come from, influencing how they approach or even consider their goal of rescuing her.

No doubt this list is not exhaustive.  Could they obtain these goals through other means?  Sure, but for some (like leveling to make the encounter more balanced) they've voluntarily avoided those things in the rush, and for others (like more info on the relationship between orc and daughter) finding a way to provide that extra information to a group that's not picked up the available breadcrumbs means either ad-libbing another encounter or asking for time for a rewrite to insert said encounter.

I read through the first couple pages but if the conversation has changed between page 3-6 sorry I didn't catch i

To me its not a question that can be answered "Always do this" because its very dependant on whats at the end.

For example, if you have a BBEG who insanely outlevels the players, he may kill them for amusement, or he may treat them like treat a fly - just swat it away and go on with his day, or he may even try to recruit them. Likewise, a beast won't have a thought process and will just see dinner.

That's not even accounting for "ends" that don't involve a BBEG.

I guess, in general, I just roll with whatever would logically happen in the world. If the players aren't smart enough to pick up the clues that this guy can eat you alive, then it may be a TPK. Its never occurred that way in my group, but if it did I doubt they'd be mad. The times characters have died in our games, the players admitted it fell back to a mistake they made. nobody has ever blamed it on the GM or bad luck or anything, so it seems to work, so long as they understand fair is fair, and the GM plays NPCs as the NPCs would act in that situation.

Rejakor himself (among others here and offsite) has opined that when my "players weren't smart enough to pick up the clues that this guy can eat you alive" and it led to a TPK, I was in the wrong, I did a bad job as DM, and I need to stop doing that, in a thread on the old boards regarding a single BBEG that was mathematically within the group's CR.  So, while I'm genuinely happy for you that you've never had players blame you, as DM, for their characters dying, that's not my experience, and therefore not one I'm able to relate to, especially.

[Rejakor]

Pretty sure that's a misquote.

What I would more likely have said was that reading your players is an essential DM skill, and if they ignore the warning signs and go barreling ahead into the dragon's mouth, either you have Retarded Players syndrome or fucked up somehow.  I do consider it part of the DM's job to set up the world in such a way that it takes a modicum of effort for the players to actually commit suicide.  It is a matter of extent.  If a shady guy mentions an ultrapowerful dragon in the hills and the PCs go running off to investigate, and then they meet a little girl, she asks them for [magical item], they say no, she turns into an ultra powerful dragon and kills them, that's a failure as a DM.  If you post clues but they're subtle and easy for the party to miss, and they do something reasonable and get OHKO'd for it, that's a failure as a DM.  If you give the party information that this guy is probably hardcore, and then they forge ahead anyway, and you write up a CR+1 minions encounter and then a CR+2 boss encounter for them and they die in the fight, that's not a DM failure(probably).

I've never had players angry at me for dying, and I run brutal horrible games that often have permanent injuries, deaths, and rarely (I swear, far less than it logically seems there should be) TPKs.  I signpost well enough that the PCs usually know it's insanely dangerous and it's not 'hur you died' but instead bad rolls, bad luck, or bad tactics in an extremely dangerous situation that kills the PCs, which is usually fairly obvious so instead of 'grr evil dm' I get 'let's roll a new party and go level up and get revenge'.

That means it's a natural psychological reaction on the part of the players, when they roll exceptionally well and get no meaningful result, to assume that the DM is simply arbitrarily negating their use of the skill, and, by inference, their skill choices.  Arbitrarily negating PC choices is deprotagonizing, and as such, should be avoided in general per an old thread regarding a paladin and a Drow slaver - even when the deprotagonizing result is based on perception.

So you run social checks as specific result?  Player:'I roll a gather information check to find the orc camp did I find it' DM: 'yes/no'

That's not really how the PHB says they work.  PHB says you can gather information on a topic, and if the topic is too specific you might get information about things related to it but not the thing itself.  So you might get that 'that orc boy and that girl have been sneaking off into the woods TO THE SOUTH on a regular basis' when you ask for 'orc camp location'  Or you might get the location of a guy who knows about orcs and he's spying on the orcs and thinks they have a breeding program to breed out humans (he is a psychotic bigot) starting by seducing(not stealing) X girl etc etc and you drop clues that she is in LURRRRRVE as they investigate.  You don't need to write up townspeople encounters.  If you have trouble adlibbing NPCs, you should have some notable ones noted down for exactly this kind of thing.

Reasons why the haunted weapons shop may be important (I have to qualify with "may" because this example has been an organic, off-the-cuff one rather than scripted) include:

    Provided the encounters helpful to allow the party, or some significant fraction thereof, to level before encountering Smells-Like-Farts.  Whether or not the party has access to Haste or Greater Magic Fang could be significant mitigating factors in the difficulty of said encounter.
    Provided the party access to magic weapons and/or materials and instructions for upgrading their own weapons (it is a weapons shop, after all), which, again, can greatly influence the nature of their encounter with Smells-Like-Farts.
    Provided evidence hidden from the townspeople (or not obtained via Gather Information) to indicate Smells-Like-Angst and the daughter have been busy trying to find out where half-orcs come from, influencing how they approach or even consider their goal of rescuing her.

DM controls difficulty of the encounter.  If you have outright said that the orc camp has a 5th level shaman and a 4th level fighter and smells-like-farts is a 3rd level rogue 3rd level swashbuckler and there are 40 orcs and 3 ogres in the camp, a) what  b) why and c)argh.  Reduce all the levels by 1 except for the shaman and make the shaman distracted (turns up later in the fight) and some of the orcs out on patrol (say 20).  Players have no way of knowing the level of melee characters.  If they go HEY just go 'customized stat array' or 'you don't get to see the sheets of the enemies, just trust me they are correct'.

Party therefore shouldn't need xp to fight orcs.  If you need 5 minutes to re-write the stats, that's fine, but if you're bad at altering stats on the fly, i'd keep several sets of stats for various things like 'they didn't go to the weapon shop' scenarios.

Get special weapon to kill Smells-Like-Farts - this is a real reason to need to go to the weapon shop.  If smells-like-farts is invulnerable or something.  For plot reasons.  If so, though, to avoid railroading them to go to the weapons shop, have alternative plot ways to get past his invuln.  The family sword the paladin's girlfriend has?  Totally is a heirloom of the original smells-like-fart killer.  Smells like Farts for some reason has weaker attacks than you planned for him, party runs away, hears rumour about crazy super weapon forger guy who died and became a ghost /trapping his orc super murdering swords in his forge hint hint/.  Someone who someone else cares about dies and it turns out THE POWER WAS INSIDE YOU ALL ALONG.  Watch the party smother him in earth, chain him up, and drag him around the place like loot.  Drop hints about him being invuln, let the party see him being invuln, they charge in anyway, they die, new characters.  etc etc.  there's lots of ways around plot-devices like that.

Evidence - evidence should be in many places.  Consider the Three Clue Rule.  That's Three Clue MINIMUM.  If you rely on players going to some specific place and finding some specific thing, then either they're not going to find it, or you're going to railroad them there.

[InnaBinder]

Where do all these hints come from when they skip large swaths of plot?  Are you recommending OOC chatter (spoils immersion) or that they conveniently run into three different folks (3 clue rule) who all randomly happen to have cogent and different information for the party on Smells-Like-Farts and his lair (spoils verisimilitude) or what?

[veekie]

^^
The hints come exactly from their attempts to bypass plot. So they gather information on the orcs and instead of getting GPS coordinates to the orc camp, they get to learn a variety of things like where the orcs have been recently sighted, or the names of a number of people who know where the orcs might be. You wrap the clues WITHIN these information packages, when they roll gather information to find the orc camp, they might get a list of two locations where the orcs are known to reside, as well as the name of a gossipy hunter at the outskirts of town who might have seen them.

Then you have three possible outcomes, the two locations and the hunter.
Say the first location is the forge, where they might search the place(and find a bit of an orc's diary), talk to a ghost(who could tell them when the orcs turn up at the forge lately), track the tribe's movement or even use knowledges to figure out what they might use the forge for(if they are using the place to make magical gear, you can then reveal where they might be collecting components, as well as what they are making, etc). These are all places you can drop further clues.

Say the second location is a cave-riddled hill with wildlife, where they might clear the place out(so the ruckus attracts the tribe), identify the location's significance(holy ground or just a hunting place, maybe their winter cache even) or conceal themselves and wait for the orcs to turn up. Again, during this time they can pick up clues.

And if they went with the hunter, he'd just blab the latest gossip to them while guiding them towards the camp, possibly stopping at one of the above two locations for the night.

That's not really how the PHB says they work.  PHB says you can gather information on a topic, and if the topic is too specific you might get information about things related to it but not the thing itself.  So you might get that 'that orc boy and that girl have been sneaking off into the woods TO THE SOUTH on a regular basis' when you ask for 'orc camp location'  Or you might get the location of a guy who knows about orcs and he's spying on the orcs and thinks they have a breeding program to breed out humans (he is a psychotic bigot) starting by seducing(not stealing) X girl etc etc and you drop clues that she is in LURRRRRVE as they investigate.  You don't need to write up townspeople encounters.  If you have trouble adlibbing NPCs, you should have some notable ones noted down for exactly this kind of thing.

Definitely agree on this. While you don't give them the answer outright, a good roll should put them some distance towards a resolution, I personally find 2-3 successful attempts(whatever the approach, as long as its plausible) to be sufficient for buildup and clue waving, with 2 leads being usually sufficient for them to work it out on their own without significant effort, and the third as a failsafe to put them at the right spot.
Failing too leads them towards resolution, but not on their terms. A bad attempt at Gather Info may put them in line of some misleading information, which leads to them blundering around the orcs hunting grounds(and then being exposed to snares for prey etc)

DM controls difficulty of the encounter.  If you have outright said that the orc camp has a 5th level shaman and a 4th level fighter and smells-like-farts is a 3rd level rogue 3rd level swashbuckler and there are 40 orcs and 3 ogres in the camp, a) what  b) why and c)argh.  Reduce all the levels by 1 except for the shaman and make the shaman distracted (turns up later in the fight) and some of the orcs out on patrol (say 20).  Players have no way of knowing the level of melee characters.  If they go HEY just go 'customized stat array' or 'you don't get to see the sheets of the enemies, just trust me they are correct'.

This though, can be a bit difficult without practice. That said unless they're grossly out of sequence(like skipping a 2 level campaign arc), they'd do fine with +/- 2 to major statistics(easily explained by moving from an elite to common stat array, or some debilitating condition), and either temporary hp or starting wounded.

Now for specific locations, they can largely be repurposed, but you can use the above changes to discovery rolls to put them there as an intermediate step to finding the orcs. Assuming a two step process(Find and pursue first lead which leads to locale and second lead), you can intersect them easily.

[Rejakor]

They are only skipping large swathes of plot because you hand them the location of the orc camp with a single gather info roll.  There is no reason that has to happen.

I covered the gather info roll in my last post.

Three Clue Rule means if you expect the PCs to search a room for clues, plant 3 clues that all point to the same place.  Not the same clue.  Just 3 different clues at least.  This is because while it's obvious to the DM what the answer is, the PCs don't have the 'god view' of what's going on and dark purple lipstick on a handkerchief is not automatically connected to the purple lipped temptress who had them kicked out of the casino last night (3 gaming sessions ago).

EDIT:  swordsage'd by veekie goddamnit veekie stop being awesome

[InnaBinder]

Then you have three possible outcomes, the two locations and the hunter.
Say the first location is the forge, where they might search the place(and find a bit of an orc's diary), talk to a ghost(who could tell them when the orcs turn up at the forge lately), track the tribe's movement or even use knowledges to figure out what they might use the forge for(if they are using the place to make magical gear, you can then reveal where they might be collecting components, as well as what they are making, etc). These are all places you can drop further clues.

Say the second location is a cave-riddled hill with wildlife, where they might clear the place out(so the ruckus attracts the tribe), identify the location's significance(holy ground or just a hunting place, maybe their winter cache even) or conceal themselves and wait for the orcs to turn up. Again, during this time they can pick up clues.

And if they went with the hunter, he'd just blab the latest gossip to them while guiding them towards the camp, possibly stopping at one of the above two locations for the night.

All of these read, to me, as the players NOT skipping to the end.  'They skip the forge and the hill by 'porting or flying or other available, level-appropriate means' is more or less the entire premise of the thread, boiled into a specific example.  If, as a DM, I'm then forced to have the players stumble upon a halfling caravan that JUST SO HAPPENS to have information relevant to their current quest, that's forcefeeding them the plot.  If I rearrange the map on the fly such that they don't avoid the hills/haunted weapon shop/ranger with info regardless of their obvious intent to do that very thing, I'm railroading them.

While you don't give them the answer outright, a good roll should put them some distance towards a resolution, I personally find 2-3 successful attempts(whatever the approach, as long as its plausible) to be sufficient for buildup and clue waving, with 2 leads being usually sufficient for them to work it out on their own without significant effort, and the third as a failsafe to put them at the right spot.
Failing too leads them towards resolution, but not on their terms. A bad attempt at Gather Info may put them in line of some misleading information, which leads to them blundering around the orcs hunting grounds(and then being exposed to snares for prey etc)

Unless you're advocating that I roll the Gather Info checks for the players in secret (and keep accurate record of their modifiers), they'll see the results of their rolls.  You may well know players who won't simply ignore information gleaned from bad rolls, but I've rarely seen them.  They'll instead be more likely to either try again or have one or more of the other party members attempt the same Gather Info, and heed the person who came across as most amiable to the townsfolk by rolling highest.  If they find themselves so pressed for time by the nature of their quest that they can't take additional time, they'll behave as if the information they received was faulty or deliberately deceptive, and ignore it.  Yes, it's metagaming; it's also metagaming to ignore the fact that a player rolled badly and pretend like they'd obtained solid information.

[sirpercival]

Inna, it sounds like you're allowing one skill (Gather Info) to break your game.  Why is it that the PCs have to determine the exact location of the endgame with a series of GI rolls?  Who has that information to give them?

[Rejakor]

Why is it metagaming to give the players information based on their bad gather information roll?  How do the characters know the data is bad/superstition as opposed to good leads?

Also I was assuming this was a lower level adventure - there are just as easily ways to have gather info lead to clues for high level parties.  Divination etc are a lot more cryptic than people assume.  If they ask for 'the location of the orc camp', what they'll get is a psychotic anti-orc bigot, a haunted weapon shop with an orcish diary, and a cave riddled with mystic orc runes.  All of which have the location of the orc camp, none of which ARE the orc camp.

The hunter knows the way, but it's a long way, and during the flight it gets dark and he wants to stop (so does the wizard) oh hey a abandoned shop of some kind hmmmmmm.... etc etc... can't teleport where you haven't seen.. good luck getting the hunter to describe the camouflaged location well enough in all those hills and mountains to actually scry it properly.

But hey, the PC's manage to do all these things without you explicitly just letting them do it (gather info = yeah let me mark it on your map, scry = autosee, teleport = autotravel)... they don't wait for the guy with information for the 'adventurers' hired by X's father rushing down the street, they completely ignore the rumours of the magic weapon forge the orcs use to make powerful items (PCs ignore powerful items?), and turn up at the camp.  They fight orcs.  They knock down Orc the Kid, into negatives.  Out comes girly girl and goes 'NOOOO ORC THE KID' 'PLEASE DON'T KILL HIM, I'LL DO ANYTHING' etc etc.

If the PCs CHOOSE to skip to the end, despite anything you do, they should suffer the consequences thereof, in this case, nearly fucking up and killing someone for no reason (or killing someone for no reason anyway, if they cut down some orcs on the way in).

EDIT:  That said, 'despite anything you do' isn't the same as 'one gather info check success'.  You are the DM, it's your job to FIND ways to give them clues.  PCs ignoring clues?  Fine.  PCs not finding clues because you didn't put them into the game, or put them somewhere the PCs didn't go and then gave up?  Less fine.