I can't understand how curses or mass effect maneuvers (via a spell) would work

[sinker]

Consider that in your second example the GM is under no burden whatsoever to compel. He may do so if it is good for the story, but there is no other reason that he really should compel that (provided that the GM is prioritizing the story rather than "realism" like he should be).

What it comes down to is that were I your GM I would compel it when I intended to compel it, and all of your convincing wouldn't make a difference. A fate point (or tag) on the other hand, might be enough.

[zenten]

See, I'm willing to consider a players reasoning on just about anything, but I don't really see *how* to factor in the spending of a tag/compel.

[admiralducksauce]

That's decided by who wins the social contest between player and GM!   

This is funny and all, but it's also the most accurate explanation I've seen so far.  If the player can convince the GM to Compel the NPC without using your tag, then kudos to you.  You get your Compel and keep your tag.  Good job.

[zenten]

This is funny and all, but it's also the most accurate explanation I've seen so far.  If the player can convince the GM to Compel the NPC without using your tag, then kudos to you.  You get your Compel and keep your tag.  Good job.

Thing is, as the GM I don't *care* if the player spends the tag.  It's not like spending that tag will make things more fun for the group, so I can't figure out how or why to factor that in to my decision to initiate the compel.

[Becq]

This is another one of those areas in which the system leaves it up to the table to decide how to apply the mechanics.  Yes, the GM can simply decide to compel NPCs on their consequences without input from the players.  Yes, the players can suggest that the GM compel the NPCs on their consequences without paying Fate.  And yes, the players can tag or pay Fate to invoke those consequences on their own.

My thoughts run along the following lines:

If it seems clear that a particular injury (as indicated by a consequence) *should* lead to a some difficulty (compel) most or all of the time, then it probably shouldn't cost the players a tag or a Fate point to trigger it.  For example, if an NPC has a Broken Leg and tries to jump across a chasm, then the GM probably ought to run the compel 'for free', whether he thought of it or it was suggested by a player.  The result *might* be that the NPC pays a Fate point to ignore the compel, but the compel should probably happen.

On the opposite extreme, if a consequence *could* lead to a particular difficulty, but only with a considerable amount of unluck, then the GM should probably only run the compel if the players trigger it via a tag or Fate point.  For example, if the same NPC with the broken leg wanted to hobble a few yards around the corner of a building, using the wall as a support, then its reasonable to assume that the action would normally succeed ... unless "something goes wrong".

How to adjudicate this?  Well, at the risk of being vague, I'd say that if a player suggests a compel, and you immediately agree, then run the compel on your own dime.  If you don't think the compel makes sense at all, then just say 'no'.  And if you are pondering whether or not the suggestion is truly compel-worthy, then hold out your hand and wait for a Fate point to materialize (or a tag to be used).  So yes, it's basically a judgement call.

[sinker]

Thing is, as the GM I don't *care* if the player spends the tag.  It's not like spending that tag will make things more fun for the group, so I can't figure out how or why to factor that in to my decision to initiate the compel.

For me this kind of thing is an expenditure of resources. The player wants this to happen enough that he's using resources that could be used elsewhere. If you take his resources and then don't at least give him an equal advantage then he's likely to be sore about it, which can lead to bad blood, which can lead to a distinct lack of fun at the table.

So what I'm trying to say is that a tag or fate point (see I equate the two) spent should at least give an advantage, even if it doesn't trigger the compel, but if the compel is reasonable then it can be a reason to trigger it.

Becq's got some good thoughts above too.