Random Amendments for the Random Method

Here are some possible ways in which the Random Method could be improved.

The main problem with it is that the longer the list gets the longer the average time before individual tasks gets actioned.

To counteract this there needs to be either a restriction on the number of tasks entered, or some sort of prioritizing system to ensure that the tasks that need dealing with frequently get despatched quickly.

The easiest way to restrict the number of tasks on the list is to start a new list at the beginning of each day. The aim should be to finish all the tasks each day.

The alternative approach, as I’ve just said, is to allow the list to grow but to develop some sort of prioritizing system.

Here’s a possible one, which is automatic and adjusts to the nature of the random process itself:

1. When you finish working on a task and there is still work left to do, re-enter the task in the normal way at the end of the list, but add another copy of the task.
2.  When you finish working on a task and no work is left to do, cross it out - and if it is a recurrent task re-enter it. Do not delete any extra copies added under rule 1.
3. When you are directed to a task by the Randomizer and the task is “empty”, i.e. there is no work to do, delete it. If it is a recurrent task and there are no other copies of the task in the list, re-enter it. If there are other copies then don’t re-enter it.

Using these rules the speed at which the Random Method picks tasks will automatically adjust to the amount of work that is needed to keep the tasks up-to-date.

At least that’s what I hope will happen. I haven’t tested it out yet!

Anyone want to give it a try?

Comments

[Dan H]

Mark, I will definitely give this a try. One question about the first rule under the alterative approach: when you re-enter the task, but add another copy, do you mean that you re-enter the task but without crossing it out at its earlier position on the list?

[nuntym]

Another possible approach is to use AF4's structure. An additional advantage of this is that you can incorporate urgent items.

This is all from the top of my head so I haven't tested it but here goes:

1. Once you have a list of appropriate length, close the list with a horizontal line. You may add new tasks below this line anytime.

2. Use the maximum number for your RNG as

(number of items above the line OR the number of lines in a page (whichever is smaller)) + (1/2 of the previous number)

3. Use the RNG. If the number is below the number of backlog items or the number of lines in a page (whichever is smaller), use the random method on the items above the line. If the number generated is greater than the above number, go to the list below the line and choose the item that stands out.

4. When all items above the line are worked on, close the list again with new horizontal line at the end, but cross out the items you are not sure you will do.

5. Go back to step 2.

[Mark Forster]

Dan H:

<< do you mean that you re-enter the task but without crossing it out at its earlier position on the list? >>

No, I mean that I would cross the task out, re-enter it at the end of the list and then write the same task again on the next line. Bear in mind that I haven't actually tried this out yet.

[Simon]

How about randomly delete tasks at the end of the day. Decide on an appropriate number of tasks for your list, for example 20. If at the end of the day your list has 30 tasks, then randomly delete 10 tasks to bring the list length back down to 20.

If the deleted tasks are genuinely important, you will recall them the next day and start adding some of them back to the list. If the deleted tasks really aren't that important then you'll just forget about them.

[Dan H]

Mark, just an update from me, having used your random ammendment to the random method for the past couple of days. I think it definitely works at ensuring that started-but-unfinished tasks get worked on more frequently. But this comes with the (obvious) cost of slowing one's pace through the list. It also isn't clear to me that adding additional copies of an unfinished task is better than simply marking that task upon re-entry and doing work on it with every pass (or every other pass, say) through the list. The advantage of additional copies, as you suggested in your post, is that it retains the element of randomness in the selection of unfinished tasks. The disadvantage, I've found, is that it tends to expand the total number of tasks/pages composing the list (counting copies) past the point that the list feels manageable.

[Mark Forster]

Dan H:

The sad truth is that there's no way round the fact that when you increase some items' priority you decrease the priority of everything else. That applies whatever method you are using.