Let us start with U.S.-Pakistani relations, after all that is one of the crucial aspects of the story we are trying to tell.
In a post in the past I conceptualised Pakistan's security sector as fragmented, to explain for oft-seen "sinking boat (= 'inevitably and fatally sabotaged') operations," perverse outcomes, and stabs in the heart, even in the midst of what sometimes looks like brinksmanship that is more or less managed for now. I described invisible factions within this abstract notion of a security sector as Faction A (pro-change, genuinely playing along with U.S. and Western demands, out of the conviction that this might be the common good, thus seen as mistaken or even traitors by many), Faction B (Realpolitikers, believing in their own capacity to calculate what Pakistan's "real" interests supposedly are), and Faction C (Islamists, loyal not necessarily to the Pakistani state, thus seen as saboteurs by Faction B at times, and seen as common enemy for the West and Pakistan by Faction A).
One of my conclusions was that if this view has any validity, then Faction A may not be very strong in fact, and ultimately it would always be Faction B that would be the strongest in Pakistan. After all the dominant position of a Realpolitiker faction is what normally characterises the security sector in any country.
To illustrate how useful it can be to consider the interplay between these factions (otherwise abstract notions): in the wake of the Swat valley and South Waziristani operations many seemed to implicitly voice the belief that Faction A (or its hand) has strengthened as the extremists alienated many in Pakistan, including many presumably within Faction B, too, and that even Faction B's attitudes may have been partly transformed as a result. Especially given how the U.S. was ready to aggressively go after the Pakistani Taliban of Baitullah Mehsud from summer 2009, for which some Pakistani good will was hoped in return. Then came U.S. complaints that the Afghan Taliban, as well as FATA forces ready to wage cross-border warfare into Afghanistan, are still left alone, and that north Waziristan and especially Baluchistan remain safe havens (the latter especially, as it is off-limits to even the otherwise informally tolerated U.S. drone strikes). A conclusion implicit in this was that neither the balance of power between the factions, nor the attitudes of Faction B may have changed that much. The game in Pakistan could still be about keeping the Afghan Taliban in play. A rationale for this might be found by some in Pakistan in the following premises: Al-Qaida may be finally weakening in the region as a result of "hammers" applied against it from two directions in the borderland (such a perception is lingering around since a while now); the Pakistani Taliban could be put out of play if it is very necessary, at the cost of going crazy hard on the Pakistani state's own population, again, if need be; the "cross-border" groups can be managed (that is partly what makes them cross the border); and that the strategically reformed Taliban could work very differently in Afghanistan the next time.
Facilitating this game, if there is such a game, are the perverse dynamics in the way Pakistani public perceptions are shaping up (h/t for the link to Paula Broadwell @ KoW), with people, including civil servants and others, believing, since a long time now, that the U.S. (and Indian meddling) is exclusively to blame for everything anyway - which gives a useful blank cheque to those willing to use it.
Spoiling the game, if there is such a game, beyond the disruption caused by any Faction C in any shape or form, is the Pak-Af borderland conglomerate. This is a conceptual and at the same time practical challenge, a strategic one. The U.S. is deploying Max Leverage against it, in the face of its own constraints regarding how much it can do. Going for it, but going for it temporarily only, thereafter half-heartedly hoping for some useful substitution by Afghan forces within Afghanistan. Punching (as big as possible) and leaving, as Patrick Porter put it. No doubt the Taliban will be offering a measured response. This is a twisting of arms, to a degree, but while the U.S. is optimistically hoping to get sufficient leverage out of enhancing an imperfect hammer-and-anvil effect on the Quetta-Shura-Taliban, for the latter the "anvil" might resemble a cushion rather, even if Islamist networks supportive of them did take some beating in Pakistani military offensives in the past. Thus the Taliban may opt to fight for the asymmetrical goal of maintaining just enough leverage on their part. This excellent paper (pdf) highlights how Kandahar is a key pressure point they want to focus on, and that the U.S. could do something about meaningfully fighting back there. But any change in outcome could be only temporary, given the Baluchistani safe haven, which is a part of why "Afghanistan strategy" is something that does not exist, as I stated before.
But things could change in this mix in the end. Outlined above is a short-term scenario, largely a ceteris paribus sort of scenario, which works if its premises are correct. It is envisioned as one of the possible futures of a region where usually there are a huge number of parallel endeavours, both overt and covert, to alter or shape whatever it is that could come ceteris paribus - from northern Afghanistan to Baluchistan, from Iran to China et cetera.