Get Everything Done

Caibre65, have you actually tried AF2ND? Instead of automatic dismissal rules which are rather arbitrary, it makes you pay frequent attention to the oldest items in the list. If they still need doing, you rewrite them right away. If they don't, then you "dismiss" them by making a conscious decision to delete that item and consider the next one.

Old tasks which deserve to be abandoned are pruned efficiently and effectively without any need to review dismissed tasks or worry about something important being accidentally dismissed.

AF2ND is a great system and addresses AF2's known weakness. What's not to like?

Sorry, Calibre65, didn't see your reply until now!

I first discovered Mark Forster a couple years ago and started experimenting with the then-current system, which I think was SuperFocus v2. I found the compulsory nature of column 2 generated substantial resistance for me, so I quickly abandoned it for AF1, which is well-loved for a good reason. It has a great feel to it and flows well -- "like water" as David Allen might say! Unlike most people, I always wanted to implement my list on 3x5 index cards, which just appeals to me for both practical and esthetic reasons.

Unfortunately, I encountered the same problem with AF1 as everyone else does -- it really doesn't handle urgent and important tasks very well. So like Mark and countless others, I embarked on a quest to find or devise the best system with all the benefits of AF1, but more effective at handling urgent and important tasks. After studying the rules of ALL of Mark's systems (including abandoned ones like AF3), I decided to try my hand at modifying AF1 to suit my needs.

Inspired by nuntym's Context Autofocus (CAF), I developed my own hybrid of AF1, AF2, CAF and a little GTD, incorporating original ideas as well. I decided that I wanted to have manual prioritization, so I used stars to mark higher-priority tasks. (I concluded it was a bit much to expect the system to magically know what I consider important or urgent.) I originally developed this system (and wrote up the rules for it) on November 18, 2011, and revised the processing rules again on December 6, 2011 after a couple weeks of experience using it. After a few more months, I decided to add an optional "Top 3" rule on April 5, 2012. My system worked well, but it was more complex than I wanted. I longed for the simplicity of AF1.

I decided to give FV when I discovered it was out. (I was away from the forums for a couple months and missed the announcement, discovered it at least a month later.) I was excited to give try FV, but I quickly found that standard FV didn't meet my needs. I then tried adding prioritization to FV (making Prioritized FV), which was much more effective for me, though much debate ensued in the FV forum over it. I still believe that Prioritized FV is an improvement over standard FV, but many people disagree.

Ultimately, I gave up on FV because I realized that the core processing mechanism (creating a chain and then executing it) was generating substantial resistance for me. I realized that no system based on FV would work for me, so I decided to return to Autofocus-style systems.

After spending some time trying to devise a simpler Autofocus system with less complexity than my old one, I encountered AF2ND and was intrigued. After a few days of consideration, I decided that the approach taken by AF2ND was more promising, so I abandoned my unfinished system and decided to try AF2ND instead. Again, I found the need to prioritize, so I tried one tweak (AF2ND+) that worked okay, and recently tried a different one (AF2NDP) that worked much better, and just settled on yet another approach (AF2NDP2) that may just be the winner -- but it's only been a couple days, so the jury's out on this one.

My experiences with AF2ND and my tweaks to it are chronicled in the AF2ND thread:

http://markforster.squarespace.com/forum/post/1908142

Editing fail. I was trying to say that I was excited to hear more about FV in the months leading up to its release, and when I discovered it was out (a month or more after the fact), I decided to give it a try.

Caibre65, sorry I misspelled your handle last time.

The problem I had with FV was the core mechanism -- working a preselected chain felt like a straitjacket to me and caused me to subconsciously avoid using the system. I discussed this issue at length last year in a thread on the FV forum that you participated in:

http://markforster.squarespace.com/fv-forum/post/1900291

Your AF2/FV hybrid is interesting, and I could see how that could work well. I'll be interested to hear back from you in a month or three to see if it holds up over time.

As I said, my current system is AF2NDP2, which I've just devised this week. I'll be equally interested to see if my system holds up in a month or three, but for comparison, I'll offer an overview of how you could use AF2NDP2, so you can contemplate how it might feel compared to your AF2/FV hybrid:

* Any time you feel like it, scan through the list and star the tasks that you want to prioritize. You could do it in the morning as a "what do I definitely want to work on today?" question, or on any other schedule that's convenient for you. AF2NDP2 allows you to add or remove stars at any time for any reason. (They're your priorities, after all; the system doesn't know or care why you want to prioritize something.)

* When selecting a new task to work on, delete the oldest task(s) if they are now unwanted, then rewrite the oldest remaining task at the end of the list (as in AF2ND). Then scan the list for the oldest top-priority task (if any) and rewrite that as well (the AF2NDP2 tweak). (If there are both single-starred and double-starred tasks in the list, rewrite only the oldest double-starred task.) After that, scan back from the end of the list for a task that "stands out" to work on (as in AF2).

That's it, nice and simple. Comparing with your AF2/FV hybrid, you could scan the list and star tasks that you definitely want to do today, in exactly the same way that you're currently dotting them to make your FV chain/mini-AF2 list. (Of course, if you like dots better, feel free to use them instead; I prefer stars.) As you work through the day, new and urgent tasks naturally land at the end of your list anyhow (as with AF2), but the AF2NDP2 tweak continually pulls the top-priority tasks to the end of the list where they'll be considered for action instead of letting them remain buried in the list where they would be forgotten under AF2. Unlike your hybrid, the entire list is available to work on at any time; whether this is a benefit or a drawback is debatable and perhaps a matter of preference.

One of the things I really love about AF2NDP2 is the flexibility of being able to add extra stars if necessary when I have multiple priorities and a lot to juggle at once. As I described in the AF2ND thread, on Saturday night, I had a list that already had 12 single-starred items and 3 double-starred priority items, and I added 7 more single-starred items, 2 more double-starred items and 9 triple-starred items for Sunday's "must-do" tasks. At the end of the day on Sunday, every single one of those "must-do" tasks was finished.

You could start out using AF2NDP2 with the intention of finishing all starred tasks every day (which is close to how your AF2/FV hybrid works), but if that doesn't work out and yesterday's "must-do" tasks didn't get finished, but today you have new "must-do" tasks that are more urgent than yesterday's unfinished tasks, you could just double-star today's "must-do" tasks instead, and AF2NDP2 will focus on today's tasks until they're done, then automatically refocus on yesterday's leftover tasks. That seems rather simple and elegant to me.

Or you could have multiple priority levels at all times -- single star for tasks to focus on for the week or month and double stars for tasks to focus on for the day, perhaps. Then every time you're done with your goals for the day, you'd be refocused toward the weekly/monthly goals.

No, you're quite right -- I'm rewriting two old items each time I select a new one. It doesn't feel like a lot of work to me, but maybe I would be in the minority on that? At any rate, it does a great job of keeping the right tasks on deck, which matters. This AF2NDP2 system would be effortless in an electronic implementation.

It's true that I'm using index cards, but they could just as easily be notebook pages -- I'm using one line per item, not one index card per item. (I've tried that, too unwieldy.) I date each card (which helps me put them back in order if I drop them), but this system doesn't rely on index cards at all. I just happen to like them, and it's fun to tear up a card and throw it away when everything on that card is crossed out. :)

It occurs to me that in an electronic implementation, I would probably want the option to reverse the display order of the entire list so the newest items are always written/rewritten at the start of the list and the end of the list is the oldest, so one could scan forward for the one that "stands out". Scanning forward feels more natural, but writing the list in reverse on paper seems more awkward than scanning backward AF2-style.

That's true, if you have just one top-priority task, it will constantly stay at the end of the list, as the first thing to consider doing when scanning the list backwards AF2-style. And doesn't that fit the idea of "top priority" nicely?

Another possibility is to always keep a "Top 3" of top-priority tasks, and it will keep cycling between those three. I like that idea, but the only problem I see is that medium priorities will be entirely neglected if you *always* keep a Top 3 by design. One possible solution to this would be to have the rules be to rewrite the oldest task for EACH priority level instead of just the top priority level, which would basically cause each priority level to cycle independently and simultaneously, but at the cost of even MORE rewriting. Thoughts?

As for tasks disappearing into the list at light speed, they will, but they'll cycle back around again sooner or later, more frequently if they're high proirity (especially if rewriting each priority level) -- so does it really matter if non-priority tasks float away?

No, I haven't tried just one priority task yet. Have you tried double stars yet?

Hmm, shall we call that "AF2R" for "AF2, Reversing"?

I'm a little confused about bouncing off either end of the list by changing direction -- doesn't this mean you'd be confined to the start of the list, if you don't wrap around back to the end?

Cricket, what do you think of AF2NDP2? Also, I hear what you're saying about medium-priority tasks aging. That's why I'm wondering if AF2NDP3 should be rewriting each priority in turn (instead of just the lowest and highest priorities), but I'm worried that it's overkill and just too much rewriting. Of course, it only differs if you use double/triple stars or more, so maybe it would tend to be a self-regulating problem?

Giving some more thought to the "Top 3" idea, if you were using AF2NDP2 and keeping maybe 20 priority tasks (with a single star) and always keeping a double star on exactly 3 tasks, that would keep those 3 tasks at the end of the list while the rest of the priority tasks would float away just like the non-priority tasks. On the one hand, each time you finish one of those Top 3 tasks, you'd presumably be most likely to promote on of the other 20 priority tasks to take its place, which will help work through them all. On the other hand, if the Top 3 are large projects that won't be done quickly, the medium priorities could languish indefinitely.

I'm thinking more and more that AF2NDP3 should be to rewrite the oldest of every priority level, to keep them all cycling at an appropriate pace. The downside is even more rewriting, but it would allow a "Top 3" of big projects without derailing medium priorities, and anyone who doesn't want to do so much rewriting can always avoid using multiple stars. I could use some feedback here if anyone has thoughts on this...

Mark, what's been your experience with trying AF2NDP2? Did you find the extra rewriting to be as burdensome as suspected, or did it turn out to be a minor thing?

Cricket:

<< Deven, I've totally lost track of the variations of AF2ND...., so can't comment on the specifics. >>

The details are in the AF2ND thread:

http://markforster.squarespace.com/forum/post/1908142

Here's a quick summary:

* AF2ND - nuntym's original tweak to AF2, replacing dismissal with rewriting the oldest task each time a new task is selected.

* AF2ND+ - My first attempt at adding prioritization to AF2ND, involved keeping a separate AF2ND list for priority tasks. This didn't work as well as I hoped (generated resistance for me), and it was complicated. I have abandoned this one.

* AF2NDP - My second attempt, involved rewriting all priority tasks once daily. Effective but cumbersome with more than a couple priority tasks. I have abandoned this one as well.

* AF2NDP2 - My third attempt, involves rewriting the oldest normal-priority task AND the oldest top-priority task each time a task is selected. Very effective, but may tend to neglect medium priorities. I haven't abandoned this one yet; I am comparing with the next one.

* AF2NDP3 - My fourth attempt, involves rewriting the oldest task of EACH priority level (oldest unstarred task, oldest single-starred task, oldest double-starred task, etc.) each time a task is selected. I expect this will handle medium priorities better, but at the cost of even more rewriting. I am trying this one out starting today, but I may decide to return to AF2NDP2.

Note that AF2ND+ and all variations of AF2NDP are effectively identical to AF2ND if no tasks are actually prioritized, and AF2NDP3 is effectively identical to AF2NDP2 if only a single priority level is used.

Cricket:

If your current system is working for you, great! Everyone seems to be different, and no system seems to fit everyone. That being said, if you ever do experiment with AF2NDP2 or AF2NDP3 (my current favorites), I'll certainly be interested to hear about your experience. Given your previous feedback on Prioritized FV, I suspect you might like it. (Unlike other systems, I would suggest throwing everything at the system at once instead of growing a new list from scratch, but to make a pass to assign priorities in the process.)

As for what to focus on, I personally like the balance of having both priority and non-priority tasks flowing through the system, and I think I may prefer the multi-level priorities flowing through in AF2NDP3 better than just top-level in AF2NDP2, despite the extra rewriting. When I was using AF2ND+ originally, I felt "trapped" in the priority list and never saw the non-priority tasks. The nice thing about AF2ND and all AF2NDP variants is that everything in the list is guaranteed to cycle through and get at least nominal attention, sooner (if prioritized) or later (if not).

As for tasks having "fixed priorities", all of my prioritizing tweaks have always allowed reprioritizing (adding or removing stars) at any time, for any reason. My current work list has 11 single-starred tasks and 7 double-starred task. If I had a must-do task for Friday, I would add it to my list now as a double-starred task to keep it on the radar, and change it to a triple-starred task at the end of the day on Thursday or the start of the day on Friday, making it a top priority at that time. (Or with AF2NDP3, I could make it a triple-starred task now and just let it remain at the end of my list until Friday.)

As for frequent review of the entire list, all of these AF2ND variants force incremental review of the entire list, but if that's not frequent enough, you can certainly choose to review the list at any time (daily, weekly, whenever) and reprioritize tasks as you go. Probably a good idea to do that once in a while, but the systems don't require it.

As far as rewriting goes, even if you rewrite a task without thinking about it, it will still be there in your line of sight as you scan backwards from the end of the list for a task that "stands out", so hopefully you're giving it a little thought during that scan, at least subconsciously...

To give an idea of the cycle time, my current list had 59 tasks at the start of the day, and when I counted earlier today, there were 11 single-starred tasks and 7 double-starred tasks. As I rewrite tasks for AF2NDP3, I am rewriting an unstarred task last rewritten on January 25 (almost a month ago), a single-starred task last rewritten on February 13 (a week ago today) and a double-starred task last rewritten on February 19 (yesterday). So I'm cycling through the full list in less than a month, the single-starred tasks in a week, and the double-starred tasks in a day. I'm happy with that workflow, personally.

Cricket,

Glad to hear you're finding the discussion useful. Sometimes I post and get no response and wonder if I wasted my time writing the post! (Still, it often helps document the history of my experiments, which I find useful even if nobody else is interested...)

In a way, your week page reminds me of my AF2ND+ tweak, where I had a separate AF2ND list for the priority tasks and only processed the non-priority AF2ND list when nothing stood out on the priority AF2ND list. The problem I had with this was that it felt like the non-priority AF2ND list was stagnating and I felt "trapped" on the priority AF2ND list. Ultimately, I decided I would rather do more rewriting and see at least the illusion of progress on the non-priority tasks on my list, so I abandoned AF2ND+ and started on the AF2NDP series of experiments. So far, I'm happiest with AF2NDP3, despite how much rewriting is involved.

I'm curious about your thoughts on incremental review. For myself, I tried using GTD years ago and it fell apart for me because the weekly review process was too heavyweight -- the system worked okay for a while until motivation waned and I started resisting the weekly review process because it felt too burdensome. I've concluded that I need a system that incorporates incremental review rather than relying on a heavyweight review process that I have to set aside time for. When life gets busy, taking an hour to plan might be wise, but it's still hard to do.

Deadlines are a bit of a dilemma to me, because I hesitate to complicate the system to incorporate deadlines into the workflow of the system, yet I don't want to miss deadlines either. For now, I've been adding parenthetical notes after the task such as "(by Wed.)" or "(by 3/1/13)", then prioritizing the tasks appropriately to ensure the deadlines stay on the radar. (As the deadline gets closer, I will increase the priority.) I also use calendar reminders in my email program (Thunderbird) at work for timed reminders, such as weekly or monthly tasks that I can add to my list when the alarm goes off.

My priority levels are normally set based on relative importance/urgency, which does have a time component, but I wouldn't say they're primarily time-based -- except for "must-do" tasks for the day, which I elevate to a top-level priority above everything else that's on the list. Since I'm generally finding that I have a number of medium priorities with single stars and high priorities with double stars, these top-priority "must-do" tasks tend to end up with triple stars, but they are always cleared ASAP, leaving me with single and double stars again.

To give you an example, I have a weekly timesheet due Monday mornings, but I want to start getting them submitted at the end of the week so it's not hanging over my head at the start of the week. Therefore, I put in a weekly reminder in Thunderbird, which popped up as soon as I logged in this morning. At that time, I put "timesheet" on my list as a triple-starred task. Since it curently happens to be the only task at that priority level, it will remain at the end of the list every time I rewrite the tasks for AF2NDP3 (until it's done or I add other triple-starred tasks), and therefore it will be the first task I see each time I scan backwards for a task to work on. That keeps it very much on the radar!

Cricket,

I felt trapped because everything I was considering was a priority task and even if I might have wanted to work on a non-priority task for a change of pace, it was never being considered because it was off on the other list. I don't mind so much if the old tasks are stagnating, as long as it's because I'm choosing not to work on them. Having them disappear from sight entirely was a different matter.

How often do I pause to reprioritize? Occasionally, not frequently. Basically, if something jumps out at me as needing to be reprioritized, I'll do so. But I'm more likely to do that incrementally as the tasks flow through the rewriting process, rather than scanning the list to update priorities. That being said, perhaps I should do a bit more scanning, at least for approaching deadlines, perhaps daily or weekly?

Massive review processes don't work well for me. GTD fell apart for me because the weekly review was too much for me, not to mention the massive initial review the system called for. (I spent many hours on an initial review and was nowhere close to done.) I need a system that incorporates incremental review naturally so it doesn't fall apart when I don't have time or energy for periodic larger reviews. That doesn't mean I'm opposed to periodic reviews, just that systems that require such reviews (like GTD) just aren't a good fit for me.

Compared with your weekly reviews, my process is more like this: Didn't work on it because I didn't need to? Leave unstarred, it'll come around again eventually. Didn't work on it but should have? Keep prioritized so it remains on the radar.

My priority levels are fairly simple. Unstarred tasks are "whenever" or "someday/maybe" low-priority tasks, not urgent and probably less important. Single-starred tasks are "nice to do" medium-priority tasks, probably somewhat important but still not urgent. Double-starred tasks are "should" high-priority tasks, usually very important, possibly somewhat urgent. Triple-starred tasks are "drop everything" or "must do today" top-priority tasks, usually very urgent and very important. Generally, I only have a few triple-starred tasks (often none), a handful of double-starred tasks, a fair number of single-starred tasks, and lots of unstarred tasks.

Thinking about this further, and having wanted to experiment with the "Top 3" idea, I've decided to incorporate that to make AF2NDP4:

http://markforster.squarespace.com/forum/post/1908142?currentPage=2#post2078625

I'm going to experiment with using AF2NDP4 starting today. The only difference from AF2NDP3 is the addition of an optional "Top 3" priority level to focus on task completion more. For my own choice of priority levels, I'll use triple stars for the "Top 3" tasks and (when necessary) use quadruple stars for "drop everything" or "must do today" tasks. Of course, this entails even MORE rewriting, but I'm okay with that. (Someone who dislikes rewriting could use AF2NDP4 with single or double stars for the "Top 3" level if they prefer.)

Note that I've been very happy with AF2NDP3. I'm just curious if the "Top 3" idea would improve it further or not.

As for your milestone chart, it's an interesting idea, and it seems like a nice thing to have, but that falls in the category of extensive planning work that probably wouldn't be a good fit for me. Glad to hear it works for you though!

As for required weekly tasks, I don't have a great solution for periodic tasks and deadlines; I've been relying on calendar entries to trigger adding high-priority tasks to my list.