Why ask a question and then provide the right answer.
It's a common rhetorical device. I understand the logic you're using and am trying to point out a flaw in it.
OAs are immediate interrupts, so they occur just before the action that trigger them. An OA is limited to the melee weapons reach being a melee basic or melee at-will attack.
And OA is /also/ limmitted to occuring in adjacent squares, and the Reach quality of a weapon very clearly doesn't alter that. Threatening reach does. Polearm gamble does not. While it's very tempting to assume that Polearm Gamble must function like a 3e AoO with a reach weapon, there's nothing to indicate that it does. OAs are not immediate interrupts, either. They do interrupt the provoking action, but not everything that the rules say about immediate intterupts aply to them.
Typical triggers are limited to leaving an adjacent square. This triggers on entering an adjacent square.
Exactly. Why do you think that leaving a square and entering a square are the same thing? Obviously, they are in the sense that it's hard to do one without doing the other, but why, would PG be phrased the way it is if the two were meant to carry the exact same meaning and implications?
There's no term like 'threatening' in 4e, except, of course, for threatening reach. Threatening reach works the way you seem to want Polearm Gamble to work (and the way I'd certainly prefer it to work), so you have to wonder: why doesn't it give limitted threatening reach instead of a novel trigger for a conventional AoO? Maybe, because it's not meant to work anything like threatening reach.
If the target is not valid for a melee basic attack or melee at-will during the time of the trigger then the attack fails, just as any other attack would fail if the target moved out of the way before it was resolved.
I understand the reasoning, it's just reasoning that aplies more to threatening reach or 3e AoOs than to PG.
The odd choice of words in PG only makes sense if the intent is to avoid giving PCs access to 3e style AoOs against non-adjacent targets. Given the fighter's ability to stop an enemy cold with an OA, that's an understandible design decision. It wasn't elaborated upon to a great degree, but then 4e tries (often, too hard) to be concise.
I'd prefer PG to say something like:
Effect: When you wield a pole-arm, you gain threatening reach. However, if you use your threatening reach to take an OA against a non-adjacent enemy, you grant that enemy combat advantage while it is adjacent to you, until the start of your next turn.
Clear and concise. The OA happens in the non-adjacent square, which must be in your reach. Yes, a fighter can stop someone from moving adjacent to him with that OA, partially negating the combat advantage drawback, but shifting (and, of course, teleporting and forced movement) still don't provoke.
Or, if it must deny any hint of threatening reach:
When you wield a pole-arm, you can make an opportunity attack against an enemy who has moved into a square adjacent to you from a non-adjacent square. However, you grant that enemy combat advantage while it is adjacent to you, until the end of it's turn.
At least this makes it a little clearer that the OA happens in the adjacent square (triggering event in the perfect tense), but it does sound a bit clumsy.
