[TrevorJ]

Some great suggestions here, but I find the best method for me is to just crank some music, down a couple swigs of coffee and dive in. About 95% of the time the issue isn't about getting stuck partway through, or approaching the problem wrong, it's about getting through the first ten minutes of work. Resistance crumbles after that and motivation and focus kick in to full gear.

[gwern]

http://lesswrong.com/lw/3kv/working_hurts_less_than_procrast...

[einhverfr]

You know, if a task isn't particularly urgent, I have no problem with procrastinating for a couple days on it (and working on other things in the mean time) as long as I keep coming back to it.

This resembles the Feynman method of how to be a genius, but sometime if a task is hard, and seems insurmountable, sometimes all you need is some time for your mind to grasp the task. I have occasionally tackled a problem, given up as it seemed hard, come back a couple days later and discovered a trivial solution, completing in half an hour what I figured would take me 3. So I call this "productive procrastination" and it's a good thing. Another function of my complex to-do list is that it gives me a chance to think about tasks before I start (again, I don't worth through it sequentially). The difficulty is in figuring out when to procrastinate and when not to. There are a lot of things you never want to procrastinate in doing (if you take a long time to bill your customers, they will usually reciprocate when it comes to timeliness of payment, for example). But there are a lot of cases where procrastination can be helpful. So the challenge is when to and when not to.

[diminish]

The 4 things suggested are exactly what I discovered to work for me; but I guess it depends on your personality and your particular case.

[PerryCox]

I'm the same way, and once I've actually started working it's hard for me to shut it down and stop working.

[onlyup]

Agreed, but it's still so hard to do. This is the biggest obstacle I face.