I'd been staying out of anti-akrasia discussion mostly because my strategy for getting things done is so different from the common one that it barely seems related, much less relevant. I've been asked, though, so here's what I have to say about it.
One of the main premises in how I go about getting things done is that the process of choosing what to do is a major factor in how easy it is to get yourself to do that thing. Being confident that the goal is one that you want to achieve and that the next step really is the best thing to do to achieve it is important. Without that, you're likely going to have to run on willpower rather than coast on your own drive to reach that goal. (Fear is another common drive, and sometimes an unavoidable one, but I really don't recommend it. On top of the stress of working that way, running on fear means that you risk having a sudden impetus failure if you misjudge what your brain will consider a safe solution to the problem in the immediate sense. Example: My fear of doctors would be better handled by taking active steps to keep my health, not by avoiding dealing with medical issues altogether as I've been doing for the last few years. I'm working on turning that around, and am actually planning to post about it as an open question of instrumental rationality sometime in the next few days.)
I have a few techniques that I use to determine that a particular goal or next step is a good one. These may seem rather basic - it's noteworthy that I didn't really comprehend the concept of 'pursuing goals' when I was younger, and that I've only been building these skills for a few years - but they do seem to be quite effective, if somewhat slower than the normal methods.
My main technique for determining whether a goal is a good one is to think about the goal in many different contexts - not by sitting down and trying to come up with relevant contexts all at once (which seems likely to induce bias, along with being more difficult than my technique), but by thinking about the goal from time to time as I go about my life. When I'm doing this, I'll spend anywhere from a week or two to a few years occasionally asking myself 'how would the situation I'm in be different if I'd already achieved my goal?'. This seems to give a pretty accurate idea of whether the goal is a good one, helps make any flaws obvious, and perhaps most importantly makes the goal feel realistic - it's not some hypothetical thing with little connection with reality, but an actual possible state of the world. The key is to ask the question in as many different situations as possible, to get a very well-defined idea of how the goal will work. For very life-changing goals, where most of my normal situations would be changed beyond recognition after achieving it, the question 'how would I accomplish what I'm currently in the process of accomplishing?' serves the same purpose, though it's a bit harder to answer in many cases and answering it is more likely to involve specific research. The most obvious flaw in this technique is that it doesn't address new situations that will be caused by achieving the goal - for example, if I were to own a house, I would have to maintain the property, which is not something I currently have to deal with, and which this technique won't address. Therefore, it's also advisable to talk to someone who's already achieved the goal, ask them about such issues, and make a point of visualizing them at appropriate times. ('How would this round of goofing off be different if I owned a house? It wouldn't exist; I'd be out cleaning the gutters.')
I use a similar technique for brainstorming ways of achieving goals, and I generally start doing so as soon as a goal I'm contemplating starts to look like it'd be a good thing to do. This is actually a rather important part of deciding whether or not a goal is a good one: Something that would be nice to have, but that would take more effort than it's worth, is not a good goal. The specific technique is to keep the goal in mind, and let the priming effect of being in different situations trigger different ideas of how to handle the goal by thinking about it for a few seconds at a time in many different contexts. Sometimes this results in specific ideas for solutions to specific problems ('I could rent out a room to help pay the bills, or in trade for help with chores and errands'), sometimes it results in topics that need to be researched ('Credit scores: How the heck do they work?', 'hmm, I wonder what science has to say about houses'), and sometimes it results in ideas that help refine the goal and narrow it down to something specific and manageable ('I seem to prefer small spaces to large ones. I should get small house. How small do houses come, anyway? Ooo, microhomes, how cool!').
Both of the above are not just compatible with plenty of goofing off, they're actually improved by it, if the goofing off is of sufficiently high quality - by which I mean that it involves a variety of things, and that I'm comfortable and relaxed rather than guilty or anxious while doing it. The above techniques are hard to use during activities that require focus, and would be distracting in those cases, but goofing off provides an ideal opportunity to switch tasks in the middle of doing something to go research the awesome idea I just had or the piece of information that I just realized is key in making any progress in figuring out how to accomplish something. If the goal is compelling enough, I generally find that I have trouble *not* making little bits of progress on a project every time I think about it - I'm actually having some trouble not wandering off to go research credit scores right now, for example - and if the goal isn't compelling enough to attract my attention even after having used the first technique for a while, it's probably not a goal I actually care about achieving.
You may notice that neither of the techniques has a clear end state in which I explicitly decide that I'm going to pursue a given goal. That's because I rarely discover a need to do so. I find that if I find a goal compelling, I tend to naturally start taking actions toward achieving that goal as I discover them, and eventually it becomes obvious that I'm willing to put effort into achieving a particular goal. (Note: There is a risk of problems with the sunk cost fallacy and similar issues, here. I don't have specific advice for dealing with them, except to scrupulously avoid thinking of yourself as 'the type of person who does [goal or task]' in relation to these things. I find that sufficient, but I'm also very practiced at that in particular, so I don't expect that it will work for everyone.) In instances where that doesn't work - where the early stages require a large enough investment of resources to make me want some extra assurance that I'm not going to lose interest, for example - I follow the stated techniques until I'm confident that I have a good idea of how achieving the goal would affect things, and what kind of work it will take, and then I imagine not pursuing it. If the goal is one that I find compelling, I will have an immediate emotional reaction to the idea of letting it go. This technique isn't perfect, but it's the most useful one I've found so far.
I find it useful, when using the above techniques, to take time from time to time to consolidate all the gathered information, make sure there are no conflicts, and take note of any obvious gaps in my knowledge. I find that taking time for that flows naturally from certain instances of using the second technique, particularly instances that involve detailed research or asking a person for advice. The latter is often more useful than the former, in my experience - it generally turns into a conversation about the goal as a whole, in which I explain what details of the goal or plan that I've decided on already and why I decided on those particular details (which is useful for noticing flaws in the reasoning behind decisions, among other things), and the other person offers targeted suggestions, points out areas that still need work, and generally acts as a sounding board.
I find the above techniques useful for personal goals, but they tend to take too long to be useful for situations involving others. In those cases, the process needs to be condensed into one or a few conversations. The conversations that I have that are successful at that have a few things in common - they're long, they're very focused, they involve looking at the issue from several angles - but most importantly, they're all about coming to an emotionally and intellectually salient, implementable conclusion. This is actually quite different from normal conversations, where a particular topic might be the subject of three or four exchanges before the topic is changed, or even 'deep' conversations, which rarely involve more than allowing each participant to explain their opinion in detail and offer a few comments on the other opinion or opinions presented. They're also not the same as the kinds of conversations that I've had with supervisors at most of the jobs I've held; those did involve implementable conclusions, but the conclusions had generally been decided by the supervisor before the conversation ever started, and even if not, the goal was to come up with a minimally-acceptable plan as quickly as possible with as little discussion and as few questions as possible, which rarely resulted in a plan that I actually cared about achieving for its own sake.
The most obvious difference between the kind of conversation that I use to develop useful conclusions and normal conversations is the degree of focus. When I first started developing this technique, before my usual conversation partners got used to it, I found myself redirecting the conversation back to the topic at hand almost every few sentences. Somewhat surprisingly, this did not ruin any of my friendships; the advantages of the increased focus were obvious enough quickly enough that this was seen as not just acceptable but preferred. (Sample: One major relationship with daily conversations, a large majority of which were this type; one major relationship with conversations 3-4 days a week, approximately 1/3 - 1/2 of which were this type; one significant relationship with conversations once per 1-2 weeks, approximately 1/2 - 3/4 of which were this type; near-daily group conversations in which I used this technique whenever relevant but not throughout the entirety of any conversation; a handful of conversations with strangers or acquaintances where I was specifically approached or recommended as someone to talk to partly because of this technique. In all three of the relationships, the technique has been commented on and noted as something that is appreciated.) I don't generally do that any more - the people I most often converse with know this technique well enough that I don't have to, and I generally take a softer approach with strangers now that I have a feel for how significant a tangent can be recovered from - but keeping an eye out for conversational drift and making sure that we're making progress on the goal is still an important part of such conversations.
The actual contents of the kinds of conversations that I use to reach salient conclusions are driven by the specific goals of the conversations, but in the case of planning to achieve a particular goal, it's generally focused on answering the same kinds of questions that are answered by my solo techniques: What is the goal? Is it a good goal? What will achieving that goal look like in the real world? What possible side effects do we want to encourage or avoid? How can various sub-goals be achieved? In what ways is one possible solution or approach better than another? Generally, a given conversation is only going to handle one or a few of these questions, if they're really unresolved.
When I'm working with someone on their own projects, those questions are supplemented by questions about what they've done so far, what kinds of things have worked for them in the past, and why they're having trouble with the issue at hand, plus even more focus on applying rationality techniques than I use when discussing projects I'll be working on personally. Ideally, this turns up a particular issue that can be dealt with either by having the person change their expectations (e.g. if they're stymied by the fact that a third party doesn't react as they 'should' to some stimulus) or approach the problem in a different way (e.g. by trying a solution that they hadn't thought of on their own). Even if it doesn't turn up something like that, such a conversation will often give a person enough of a different perspective on the issue that they can start gathering information and looking for solutions on their own.
When I'm dealing with issues of my own productivity, though, I tend to stick more to the core questions and really focus on finding very specific, clear answers. For example, it's not uncommon for a client to approach the company I work for with a request for a specific build (in Second Life or OpenSim, a 3d scene and all the code that makes it do things) and for us to subsequently spend an hour or more just determining what the client's goals are for that build to make sure that what they've requested will actually accomplish what they want. I actually find it almost impossible to take a project seriously if we haven't done that part, now; every time I've let myself be talked into working without it, I've had to re-do the majority of my work to accommodate the client's actual preferences, and nothing kills my motivation to work quite like feeling certain that I'm not actually accomplishing anything.
Once I have a firm enough description of what actually needs to be accomplished, I work out the details using another variation of my solo techniques - not because there's some particular advantage to doing it over several days rather than sitting down and working it out all at once, but just out of personal preference - and do most of the actual work in bursts of caffeine-fueled building, coding, and testing. Most client projects come with hard deadlines, which I find useful not just as a specific goal, but also because the owner of the company I work for is a dear friend and I prefer not to put him through the stress of worrying about whether I'll meet them. (Training oneself to notice and avoid deadline-related stress may be a more feasible option for most people.)
Overall, my method suits my own skills and limitations well: It doesn't require much willpower, nor expect me to be able to focus on demand; it's flexible enough that if I'm having a rough day or suddenly find that I have an emergency to take care of, it's not a problem. It takes advantage of the rather diffuse state of mind that I prefer to spend time in and my particular way of observing the world. It may be so specialized for me that it's not viable for anyone else, but it does at least exist as an alternate way of handling the process of making sure that things get done.