Flexible Autofocus
Monday, January 9, 2017 at 7:15 In my post Systematic, Fast and Flexible I said that FVP was systematic and flexible but not fast. Accordingly I spent a bit of time developing a variation on FVP called Fast FVP. Those readers who have tried it out have generally speaking found it works well.
However in the post I also said that Autofocus (AF1) was systematic and fast but not flexible. So I’ve now been looking at how I can produce Flexible Autofocus. I’ve been trying this out over the last few days and actually prefer it to Fast FVP.
Here’s how it works:
1. It uses exactly the same rules as AF1 except for the following.
2. When you finish scanning a page, you no longer proceed automatically to the next page and repeat the page scanning procedure. Instead, you scan forward ignoring the page structure until you come to a task that stands out as ready to be done. You may skip over several pages before this happens.
3. Once you have done the task that stood out, you are “trapped” on its page. You have to carry out the full page scanning procedure on that page in order to be released from it. Once you have been released, you repeat rule 2.
4. AF1’s dismissal rules obviously wouldn’t work with this. I don’t see much point in trying to work out a different form of dismissal, so there is no dismissal in this method.
I’m so far very satisfied with how this works. We’ll see how well it stands up in the long term. But there is one challenge yet to come:
The Systematic Next Hour
Reader Comments (38)
That just might be the case!
Also, as I mentioned before in a forum post, we did have a period of "no dismissal" shenanigans some years past where we found out that AF can actually work without any dismissal, and merely adding a "streamline/clean-up list" task or something similar to the AF list would be sufficient.
SCAN MODE:
1. Scan the list, looking for an item that stands out. Don't pay attention to pages. Keep moving through all the pages til something stands out.
2. Work on that item as long as you want.
3. When done, cross out the item. If it's unfinished, re-enter the item at the end of your list.
PAGE LOCK MODE:
4. Briefly read through each item on the page, pausing enough to let the item register in your mind.
5. Scan through the page, looking for an item that stands out. If you get to the bottom of the page, continue from the top of the same page. If you cycle through every item on the page, and nothing stands out, then continue on the next page with Step 1.
6. If something DOES stand out, then work on that item as long as you want. When done, cross out the item. If it's unfinished, re-enter the item at the end of your list. Then repeat Step 5, starting from the item immediately following the item you just crossed out.
Did I get it right?
<< Did I get it right? >>
I think so. Except that I don't usually bother with Step 4. I just carry on round the page from the task I've just done.
When you reach the oldest page, if you do nothing on that page, dismiss the items on that page. Only dismiss one page per cycle.
Some variations to slow it down:
2-step dismissal. The first time you do nothing on the page, put the items on notice by dismissing them. The second time, dismiss them.
Dismiss only one thing on the page. If glacially-slow, you can add it to the newest page.
Some variations to speed it up:
Any time you do nothing on a page, whether or not it's the oldes, dismiss the entire page.
<< Some variations to speed it up: Any time you do nothing on a page, whether or not it's the oldes, dismiss the entire page. >>
But that would slow it down. It would tend to make you stop on that page and work on something just to prevent dismissal. Without this rule, you can quickly move forward to something that stands out.
<< Any time you do nothing on a page, whether or not it's the oldes, dismiss the entire page. >>
Umm... that variation is called AF1.
I have a nasty feeling that I might have deleted one of your comments by mistake. If I did, it was probably on this thread.
Can you remember what it was if I did in fact do that?
I checked my recent comments and all seem to be there! I don't believe I'd commented on this post yet either. I know there were some double-posts recently that got tidied up though. Thanks tor asking!
I actually am a bit perplexed why people are insisting on adding a dismissal rule to this system!
I thought there was a difference when I wrote the idea. Brain glitch. Trying to think of what it was.
Nuntym: For someone who hated the idea of dismissal, I've come a long way. Now I'm terrified of not having it!
Seraphim: I resisted dismissal for ages. At the time, it was because I didn't want to give up on projects. Now, though, I'd probably do as you say, and work on them just to avoid dismissal. Not the intended result!
It also depends on how seriously you take dismissal. If it means gone forever, you'll resist. If it means gone until you feel like writing it on the next page, it does not good.
A hibernation list is a good compromise, and it was a big step for me. (Yes, this site was on my hibernation list for a while, which is why I dropped off.) A shorter time might be resisted less, so be more useful. Maybe gone until I recover from a project, or post-vacation backlog. Maybe until I can prove the remaining list isn't too optimistic. Long enough that I get used to not doing it, but not so long that I'll work on it rather than dismiss it. A separate list helps psychologically, and also lets you throw out the book when it's full.
Mark and Nuntym: Some of us have great difficulty weeding. If weeding needs to happen, then the system needs to enforce it.
Now I'm wondering if you're right. Maybe the system doesn't need it. The only benefit is that you can safely ignore dismissed pages, which speeds up the system even more.
(Option: Ignore until date. Ignore the undone tasks on this page until that date.)
What is the benefit of the page-lock? Why not continue scanning the entire list from where you are?
For example,
Select the next three pages you want to work on;
Work on two of them (by circulating around in the usual AF1 manner);
Pick two more so you again have three picked out.
You might always search for pages from the back, to encourage rotating around them, assuming you are moving the selected pages to the front.
Why select three pages at a time? To encourage picking important projects. We tend to pick quick & easy things for "what to do now," while we make more responsible choices about what to do soon or what to do later. It's SMEMA's "psychological distance," or anti-Akrasia.
I am doing something similar with index cards, except my cards tend to be themed by project. When I cross out a task, I rewrite it on the same cad, for that project. Non-project items go on catch-all cards. There is usually only one catch-all card open (all others are full), but project cards start out with only a few tasks on them.
<< Some of us have great difficulty weeding. If weeding needs to happen, then the system needs to enforce it.>>
If weeding is enforced it's no longer weeding, it's dismissal.
The difference between dismissal and weeding is that dismissal is intended to put the squeeze on you to work on tasks that for some reason you don't want to do, while weeding is a matter of keeping the list relevant. As such it doesn't _need_ to happen. It's just tidying up.
<< What is the benefit of the page-lock? Why not continue scanning the entire list from where you are? >>
You could ask that question about AF1. As a general rule it encourages you go deeper into the list, rather than just scan over the surface. The aim of Flexible Autofocus is to improve the flexibility of AF1 while maintaining much of its depth. It's a question of balance. I think FAF gets it just about right.
<< Why select three pages at a time? To encourage picking important projects. >>
Yes, but this is supposed to be _flexible_ Autofocus. If you do as you suggest, you will lose flexibility..
Note that I'm not saying it won't work. I'm just saying you will change the balance between flexibility and depth. As I said in answer to cricket, I think FAF gets it about right.
<< As a general rule it encourages you go deeper into the list, rather than just scan over the surface. The aim of Flexible Autofocus is to improve the flexibility of AF1 while maintaining much of its depth. It's a question of balance. I think FAF gets it just about right. >>
Thanks Mark, this answers a question I was having. I haven't actually tried this system yet, but have been thinking through how it would work. I guessed the key mechanism behind its effectiveness was giving just enough extra focus and depth when you need it. I guessed related tasks would tend to cluster on the same pages, so it's exactly the right place to add some focusing power. In fact, I would guess this might make AutoFocus even more focused, in addition to being more flexible. Is that what you are finding?
Sometimes these subtle effects can have a surprisingly powerful impact on the overall feel and effectiveness of the system. Is that what you are finding?
This gives me a couple of ideas for the system I've been working on myself, for which I just posted a teaser on the FVP forum: http://markforster.squarespace.com/fv-forum/post/2653928
Thanks again!
Thanks Mark!!
I should add that over many years, I have conditioned myself to lack Mark's discipline, and most methods will only take me so far.
Time for my daily dynamic list...
It seems no matter what kind of catch-all I use, it eventually starts weighing me down with a feeling of debt and dread. My new system helps me stay focused but eventually I just need to start over to get away from that feeling of debt.
This triggers a thought... Could this account for the big jump in productivity and engagement that usually occurs when we adopt some new system? Maybe it's not so much the new system, as it is the elimination of the backlog of debt. No-list does this continually.
Hope you are all right, I thought a lot about you. I am happy to see a new system and first at all an AF1 which was for me one of the best system you have invented with AF4. I am going to give it a try and see how it work
What I guess is that may be my list will again be huge and huge. This was a real lack of OF1. May be I could suggest an idea which would be to close the list at the end of a time. May be for me a week and then extract which a yellow highliner what will be still availlable at the end of the week. This would be a kind of automatical dismissal and elimination of out to date item.
Then each week I could strat with a new list knowing that some most important items are still somewhere in my list so I could refer to them when i would need it
It is just an idea.
Thanks for your good wishes. I'm still holding up ok, but I've got a lot of treatment still to go.
I think the best thing to do to keep the list a reasonable size is to aggressively weed it, but use whatever method you find works best for you.
Your systematic no-list system just might be GED! A forum conversation led to me to take another look at the GED book, and what I see in Chapter Seven (the rotating timeboxes) sure looks like a systematic no-list to me. It's not a Systematic Next Hour, but close enough?
GED is not intended to be a "No List" system. In effect it is a "Lots of Lists" system. For instance in the example given in Chapter 7 (page 95 in the print edition) the categories are:
Email
Letters
Phone
Filing
Accounts
Client Follow Up
Prospecting
Publicity
New Premises
Each of those will have either a checklist or an in-box (in some cases both). I was quite clear in specifying this. "For each project you will be working off a checklist".
How would you adapt this to working as a "No List" system?
P.S.
It's interesting to see how communications have moved on in the eighteen years since the book was written. Email is no longer my primary method of communication. I've sent one letter so far this year (with much grumbling at the inconvenience) and only seem to receive them from doctors. My house phone has very little traffic and I use my smartphone for everything except making phone calls!
All you would need is a handful of categories that cover everything, and rotate around them in GED's escalating timebox fashion. If a category has one clear priority, then it is the goal for that timebox. In no-list fashion, you can make one of those temporary throw-away lists (I forget what you called them) for any category. When a category needs a new goal, you can select from its no-list list.
OK.... Are you intending to try it out as a no-list system? - because I'd be very interested in how it works for you.
I am currently using GED's escalating timeboxes as one of my tools, but it is mix-and-matched with several other things that work for me. I don't throw away any of my lists, so I doubt mine qualifies as a no-list system. No-list was never great for me anyway, but I imagine those who like no-list could make good use of GED's framework to keep hammering on projects in a tight loop.
I came across this version of AF recently.
I'm slightly confused by the second instruction:
"2. When you finish scanning a page, you no longer proceed automatically to the next page and repeat the page scanning procedure. Instead, you scan forward ignoring the page structure until you come to a task that stands out as ready to be done. You may skip over several pages before this happens."
As per AF1 instructions should this read "you no longer proceed automatically to the top of the same page to repeat the page scanning procedure"? Essentially it seems this version of AF does away with the quick first scan and the slower second scan and instead you just scan down the whole list ignoring the page structure until you come across a task you want to do?
Aside from this and the fact there is no dismissal procedure, there are no other differences with AF1 that I can see? So is the intention therefore that any urgent tasks are more readily available to be selected? Having tried this out I'm not sure whether this improves this problem which was encountered in AF1 (hence the development of SF) as you could be stuck in page-lock mode which is not really any different from AF1?
Also can you or perhaps nuntym explain why the dismissal rules won't work here as it seems so similar to AF1 that I'm struggling to find any differences!
I read in another post that you were working on another system recently - I would be interested if there have been any developments there. I appreciate all your hard work in developing these.
I don't understand from your comment what you think the instructions are saying, so I can't answer your questions.
If you still can't understand my instructions after re-reading them, then I suggest you read Seraphim's summary (6th comment down) which is basically correct, plus my brief reply to it (7th comment down).
If you understand FAF correctly, it should be obvious why AF1's dismissal rules won't work with it.
FAF never established itself as one of the "standard" methods, whereas AF4 is still I believe used by quite a few people - albeit sometimes with their own modifications.