Bah! I thought I was done with this thread! But I find myself drawn back by what you say here. First let's confirm that Deception doesn't use "might".
There has been some quibbling over whether Deception impinges on "thinks" from PHB 185. My own thought experiment to try and understand that goes like this
- A NPC puts forward a piece of misinformation - "I saw two halflings with green-checkered neckerchiefs pass by, day before yesterday"
- The PCs now unavoidably have that notion in mind - two halflings with green-checkered neckerchiefs - and for the sake of argument this is not something they already had in mind (if you feel PCs regularly have such halflings in mind, substitute an alternative of your choosing)
- The PCs do not know it is misinformation, and in this case for whatever reason they don't raise any doubts
- Let's say that something depends on this particular piece of misinformation, so it is not inconsequential if the PCs accept it
The player-characters live in the game world. And at the same time, some DMs are not actors. So the clues that might give away falsehood that we imagine are present in the game world, are not reliably present in the real world. Just as the longsword we might swing in the game world is not present (usually) in the real world. What I am getting at is that for various reasons it can happen that the players don't decide to challenge the misinformation. Does the NPC automatically succeed in deceiving them?
There are several concerns that this raises for me.
- The NPC determined something the PCs think - the two halflings, their neckerchiefs [EDIT So this is to say that deception can come into play in connection with an action that should be covered by PHB 185, were that a rule]
- A check seems well-justified even though it is going to be about something the PCs will think - e.g. realising it is, or has the potential to be, misinformation [EDIT So this is to say that ruling the NPC's deception automatically successful, doesn't seem right, ergo no prior certainty]
- Deception doesn't quibble with us, there's no might (I realised too, that specific-beats-general doesn't turn on "might") [EDIT If it turns out to be helpful, we can unpack that further]
It reads to me that the game designers didn't want social skills to override player agency, but they were comfortable with social (and other skills) determining what player-characters
think. So whatever theory I must have, must be one that doesn't enforce PHB 185 in whole, or at all times. But to make an exception - even in just this one respect - requires a different theory from what has been propounded.
To me, the simplest theory is to agree that the designers don't intend social skills to impinge player agency, but PHB 185 isn't a rule. It's not their means of getting there. We know the PHB contains many non-rules. A simple example being story-text at the top of each race. Another being text that creates context but falls short of rules - such as the text at the start of character creation. The PHB also contains advice for players, such as about creating a character of a given class. Look at Monk, are they saying it is a rule that you
must think about your connection to a monastery? All through the PHB are statements that were they taken as rules would make the game impossible to play.
Hence I feel drawn to agreeing that overall the designers have shied away from letting ability checks override player agency, but they also do not rule it out mechanically. That's because all through the game some things do override player agency and they empower a DM to extend that list. All of that isn't because of PHB 185. It can't be, unless one says that PHB 185 is only partly true - in just the particular and special ways one claims it to be. That's always a problematic path to take.
So the theory propounded is wrong in one respect. PHB 185 is a guideline: there is no prior establishment of certainty regarding any special group of ability checks. It's true that ability checks shouldn't override player agency, but not for that reason, and its not an absolute truth of the game system that they cannot. Rather it is RAI: a matter of how they expect the game system to be used. [EDIT So the theory I put forward looks closely similar to
@Charlaquin's, except that it omits the assumption of prior certainty. I say that a DM is empowered, expected and endorsed by RAW - both taken holistically, and giving individual consideration to elements that bear on it - to decide that something attempted requires an ability check because in their unfettered judgement it is uncertain. Including for social interaction.]
