Bluff, Sense Motive, and constructs

[Heliomance]

How would you adjudicate a player attempting to use Sense Motive on a construct, if they didn't know it was a construct? I've a homebrew class which has the ability to make constructs that look exactly like humans, no way to tell the difference without magic. But the thing is, it's not human. It doesn't have emotions, its expression is entirely under its concious control. This means that those little tells of when someone is lying that Sense Motive lets you read simply aren't there. Making all Sense Motive checks auto fail seems a bit harsh, but I'm not sure what to do. Advice?

[Greenish]

How would you adjudicate a player attempting to use Sense Motive on a construct, if they didn't know it was a construct? I've a homebrew class which has the ability to make constructs that look exactly like humans, no way to tell the difference without magic. But the thing is, it's not human. It doesn't have emotions, its expression is entirely under its concious control.

Too much Philip K. ****?

[Edit]: Sometimes, that filter really annoys me.

[awa]

sense motive should let you detect its a construct i believe you can do that even with shape-changing spells

[kamikasei]

If the construct can manipulate their body language to consciously display what should be unconscious emotional reactions, give them a bonus to Bluff and let that cover it.

If the construct simply doesn't have any body language it doesn't deliberately choose to use (and isn't necessarily adept at using the "correct" gestures), let that absence of affect be the first thing anyone using Sense Motive picks up on.

Too much Philip K. Dick?

[Edit]: Sometimes, that filter really annoys me.

We can work around it. We have the technology.

[Shhalahr Windrider]

This means that those little tells of when someone is lying that Sense Motive lets you read simply aren't there.

And the absence of those Tells are going to be trouble, too. No real person is that much a blank. The construct is going to have to make Bluff and/or Disguise checks just to pass for truly human anyway. If the construct succeeds on that, then you may be able to justify a +4-ish bonus on Bluff checks for actual lying.

[Zaydos]

Read Do Androids Dream of Electric Sheep by Phillip K. -filtered- mentioned above. They use a test based on asking questions and measuring how the android's pupils dilate because it is just such a little bit slower than a real human's natural reaction. What you're talking about would be much more obvious so a Sense Motive check ought to let them figure it out.

[Greenish]

We can work around it. We have the technology.

Yeah, it just happens to be a major infraction, and I'm not too keen on those.

Though the forum rules also maintain that words that also have a benign meaning are not filtered. I'll remember that the next time I'll need to **** a crossbow.

[Zeofar]

Aye, like the people have been saying, you should be able to tell if someone isn't acting like a normal person. IF they are actually trying to look human, what would make sense is an opposed Disguise/Spot check, and if the construct wins, they get a large bonus to Bluff. If they lose, the character knows something is off, AND the construct gets no bonus. Arguably, Sense Motive should be able to tell that something is wrobut if I recang, ll correctly, pretending to be something you aren't is primarily handled by Disguise and opposed by Spot. You could always just make an exception and allow this Disguise to be seen by Spot or Sense Motive.

Alternatively, what I'm also seeing here is that the constructs are acting totally naturally, but they are simply lying at the same time. A giant flat bonus to Sense Motive works here, but it might as well be an auto-fail if you put it high enough.

How would you adjudicate a player attempting to use Sense Motive on a construct, if they didn't know it was a construct? I've a homebrew class which has the ability to make constructs that look exactly like humans, no way to tell the difference without magic. But the thing is, it's not human. It doesn't have emotions, its expression is entirely under its concious control. This means that those little tells of when someone is lying that Sense Motive lets you read simply aren't there. Making all Sense Motive checks auto fail seems a bit harsh, but I'm not sure what to do. Advice?

These statements come close to incongruous. Unless the construct has perfect knowledge of human psychology, and is actively using it to replicate human behavior, these can't both be true.

[Shhalahr Windrider]

Yeah, it just happens to be a major infraction, and I'm not too keen on those.

It’s only an infraction if circumvented to post obscene content. Confirmed by Roland.

These statements come close to incongruous. Unless the construct has perfect knowledge of human psychology, and is actively using it to replicate human behavior, these can't both be true.

Even then, it you can expect lag time between deciding what the proper reaction is and actual exhibition of that reaction.

[Greenish]

It’s only an infraction if circumvented to post obscene content. Confirmed by Roland.

Ah, good to know.

[Telonius]

I'd set some sort of a static DC (maybe 35-ish - 10 more than a Sense Enchantment) to realize that there are no "tells." If you reach that DC, you realize that whatever this guy is, he has a totally uncanny control of his own body. The emotion he's projecting is the one he intends to project, and is not necessarily what he's feeling. You also realize that your normal methods of detecting emotion are not going to be effective at sensing the target's motives.

Basically, when you hit that DC, you're at a poker table and just realized you can't tell who the sucker is.

[Zaydos]

Sense Enchantment is 15 for Dominated, and 25 for Charmed. This is actually closer to Dominated with the limited range of actions.

[Heliomance]

I'd set some sort of a static DC (maybe 35-ish - 10 more than a Sense Enchantment) to realize that there are no "tells." If you reach that DC, you realize that whatever this guy is, he has a totally uncanny control of his own body. The emotion he's projecting is the one he intends to project, and is not necessarily what he's feeling. You also realize that your normal methods of detecting emotion are not going to be effective at sensing the target's motives.

Basically, when you hit that DC, you're at a poker table and just realized you can't tell who the sucker is.

I quite like this, actually, though I may set the DC a bit lower, so it's actually obtainable on a lucky roll without being really high level or incredibly focused.

[kamikasei]

I'd say that only makes sense if the construct is capable of consciously mimicking unconscious reactions in real time. That's quite a feat. If the construct instead has to choose to exhibit any given emotional reaction, that should be very easy to spot unless it can think and act much, much faster than a human.

[Heliomance]

Well, the purpose of the ability is to make constructs that can pass for human. I just thought later that that would give it the ability to lie rather well.

[kamikasei]

It also implies that it's terrifyingly smart. Maybe it would be simpler for it to have its own unconscious responses designed to mimic humans'.

[Heliomance]

Fair enough. Just treat it like normal then?

[awa]

it depends how you want to do it, it could be very hard to catch in a lie but you could make a check to realize its a construct. you still wouldn't know if it was lying but you would no something was off about it.

[Lord Vukodlak]

I don't think sense motive could identify a construct, it could tell you something is off that the person's responses are cold, mechanical almost artificial. This could make the person believe they talking to a construct or a Vulcan.

[awa]

i was working off the fact that say alter self and such give bonuses to disguise checks

[Shhalahr Windrider]

i was working off the fact that say alter self and such give bonuses to disguise checks

Disguise is opposed by Spot.

[Malbordeus]

hmm. i think you need to make bluff checks to "act in character" ergo, as a living being. maybe disguise vs spot with a huge bonus for looking exactly like a living creature?

the one i like with sensing enchantments is where it goes on to say if you got 15 or more, but not as high as 25, you know something is up with the individual, but might not know exactly what.

its an interesting dilema. I once had the oposite situation where I was DMing and the party were trying to bluff a golem into believing they were its master...

[137beth]

Sense Enchantment is 15 for Dominated, and 25 for Charmed. This is actually closer to Dominated with the limited range of actions.

I'd say it should be more difficult than dominated. If you consider a dominated NPC, specifically if you have seen them un-dominated, you would notice something is off. But for a construct, there is no "normal" state to compare them to. Still, probably easier than charmed...
So DC 20?

[Zeofar]

Hmm... How about this: The construct gets a static bonus to disguise checks to appear human, and to ALL bluff checks equal to the creator's Bluff/Disguise ranks and/or ECL for creating the thing. To create something that can pass as human socially, then you need to know HOW to pass as human socially. Plus, if you are good at magicks, it can do so better in terms of "programming."

[LibraryOgre]

Yeah, it just happens to be a major infraction, and I'm not too keen on those.

Though the forum rules also maintain that words that also have a benign meaning are not filtered. I'll remember that the next time I'll need to **** a crossbow.

The Mod Wonder: For reference, context is important. You will not be infracted for Phillip K. **** or **** a crossbow, anymore than you will be for the infamous "chink in the armor" that caused laughter a while back. For a touch of irony, however, my simple attempt to bypass the filter and write the words still got blocked.

To the point, if you have a construct that can be taken for living by a reasonable person, I'd let the character make a Sense Motive or Bluff check, but unless he succeeded exceptionally well, he gets information only that the construct wishes to give. You might establish a separate DC for the person to notice that it is a construct, but I'd make it high-ish (my suggestion would be 20 + half HD + Charisma; "hunch" with the complexity of the construct taken into account). The reference to Bladerunner/Do Androids Dream of Electric Sheep is apt; Deckard's test is essentially a take 20 on the Sense Motive check to determine if this is a replicant.

[kamikasei]

The reference to Bladerunner/Do Androids Dream of Electric Sheep is apt; Deckard's test is essentially a take 20 on the Sense Motive check to determine if this is a replicant.

This comparison seems to work against the point you're making. (Well, actually there seem to be a few different points floating around...) As far as I understand it, replicants can't consciously control their physiological responses. They can't decide to send false cues to back up a lie they're telling. Their body language is outside of their control, it just differs from humans' in a detectable way. So if the OP's construct is like a replicant, it might take a spot or sense motive check to detect the mismatch and identify it as a construct, but it wouldn't have any advantage in telling lies.

As I see it, the possibilities are roughly:
1) The construct only displays such emotion as it consciously chooses to imitate. This means it looks quite stilted and artificial, falling in to the Uncanny Valley.
2) The construct mimics unconscious responses through an unconscious process of its own, which doesn't quite perfectly replicate a human's. It's less obvious that it's a construct, but you can still tell if you're good enough. It doesn't have any advantage on bluff.
3) The construct has unconscious responses as in (2), but rather than just not quite matching humans', it's designed to give false responses under certain circumstances. It gets a bonus on bluff checks to lie because it doesn't behave any differently than if it were telling the truth. On the other hand, it might have a hard time convincing people it's telling the truth when it is, because it doesn't have any involuntary signals of stress or arousal.
4) The construct can consciously choose to display emotion as in (1), but quickly enough to fool a human in real time. This gets you your bluff bonus, but implies other things about the construct that may not be part of the idea.

[LibraryOgre]

This comparison seems to work against the point you're making. (Well, actually there seem to be a few different points floating around...) As far as I understand it, replicants can't consciously control their physiological responses. They can't decide to send false cues to back up a lie they're telling. Their body language is outside of their control, it just differs from humans' in a detectable way. So if the OP's construct is like a replicant, it might take a spot or sense motive check to detect the mismatch and identify it as a construct, but it wouldn't have any advantage in telling lies.

Replicants cannot... the constructs he mentioned specifically can. For a replicant, you're using "hunch" to guess that something's wrong in a social situation... namely, these replicants are just a bit off from human responses. The constructs, however, can consciously control these things, and so can give off a false positive... develop a tell when they want you to think they're lying, or whatever.

[Zaydos]

Actually in Do Androids Dream of Electric Sheep they specifically could develop an artificial tell; they had no empathy (they did have emotions) and their reactions were, just like he has stated the constructs, completely under their conscious control. The test for replicants was based on how consciously controlled actions have a lag not present in unconscious actions. Some of the replicants would show their lack of empathy by their answers, but the test was actually based mostly on reaction time. In the book it was actually being called out because it might produce false positives with... I forget the book term for them, but people with mental problems and because replicants were getting better at hiding their reactions.

[Psyx]

Have the construct make Bluff checks.

If the player passes the SM checks, then the construct is going to kind of freak them out a little, as it will be probably giving all the signals of a sociopath or someone registering on the autism spectrum.

[LibraryOgre]

Actually in Do Androids Dream of Electric Sheep they specifically could develop an artificial tell; they had no empathy (they did have emotions) and their reactions were, just like he has stated the constructs, completely under their conscious control. The test for replicants was based on how consciously controlled actions have a lag not present in unconscious actions. Some of the replicants would show their lack of empathy by their answers, but the test was actually based mostly on reaction time. In the book it was actually being called out because it might produce false positives with... I forget the book term for them, but people with mental problems and because replicants were getting better at hiding their reactions.

I stand corrected; I was basing it on a half-remembered bit of Bladerunner froma few years ago. Thanks, Zaydos.