Choosing Between Multiple Alternatives

As a bit of light relief, here’s a simple method for choosing between multiple possible alternatives. I’ve found this very effective.

Say you have to chose which book to read next, and you have five candidates:

Oliver Twist

The Grapes of Wrath

The Life of Pi

War and Peace

Pride and Prejudice

You do it by repeatedly comparing the first and the last on the list and rejecting one of them.

So you start on the above list by comparing Oliver Twist to Pride and Prejudice. Pride and Prejudice wins, so you delete Oliver Twist and compare again, this time with The Grapes of Wrath

Oliver Twist

The Grapes of Wrath

The Life of Pi

War and Peace

Pride and Prejudice

This time The Grapes of Wrath wins, so the next round is to compare it with War and Peace.

Oliver Twist

The Grapes of Wrath

The Life of Pi

War and Peace

Pride and Prejudice

You decide War and Peace is a bit heavy. So now it’s between Grapes and The Life of Pi.

The Grapes of Wrath wins!

Oliver Twist

The Grapes of Wrath

The Life of Pi

War and Peace

Pride and Prejudice

I’ve used this on a many different types of choice, including which soup to have for lunch, what movie to watch next and - yes - what book to read. Try it out when you’re in a restaurant and can’t choose between the items on the menu. You’ll find it works really well.

But one thing I haven’t been able to work out is how I can apply it to a time management system. Any ideas?

Comments

[Seraphim]

This seems to me identical to an initial FVP scan, except for the order in which you scan the list.

[Mark Forster]

You're right. There are the same number of comparisons in the two methods.

But I find this method much faster than FVP, and would definitely prefer it if I were only choosing the one winner. If I had to rank the whole list in order then FVP would be better.

[AGM]

I like that comparison method and have used it for a few things, including employee rankings. It could help with time management when faced with several potential tasks that seem equally doable.

[ubi]

So if I examine the mechanics of this "First-or-Last" (FoL) procedure and forget the premise – that you are choosing 1 of N things – the virtue lies in the list becoming shorter while staying contiguous.

What if the premise were different for the example above, and the goal is to read all the books (i.e. Get Everything Done [GED])? At each stage, you must read the first or last one remaining on the list. You avoid the less-interesting one each time until there is only one left. So you don't start with the best book, but just the better one of two. Perhaps this limited freedom is sufficient motivation to get started.

If we extend this concept to a general list of tasks, an issue arises: how to handle new and reentered tasks? After having chosen one or more Last tasks from the list, any new entries would result in a noncontiguous list. So a new 'open' list would need to be maintained while using FoL to complete a closed-list backlog. Seems complicated and not flexible enough.

But what if we change the procedure to just consider the First-or-Second (FoS) items? After all, there is nothing special about the first and last items on a menu. Using FoS to choose 1 of N doesn't result in a contiguous list. But it would for GED, if you simply copy the First task to the end if you choose the Second.

An even simpler approach just occurred to me. I'll write it in a separate comment.

[ubi]

Top or Cross (ToC) System

1. Write a list of tasks, one per line. Add more tasks at the bottom whenever they come into your head.

2. Decide whether to take action on the first task (top of list). Yes: do some work on it, cross it out, and reenter if not finished. No: just cross it out, and reenter it unless you think it doesn't need doing/remembering anymore.

That's it. The only penalty for not doing the Top task is having to cross it out and perhaps reenter it at the bottom. The list stays contiguous, unless you violate Step 2 by doing some (emergency?) work on a task further down the list without crossing out all the items above it.

[Mark Forster]

ubi:

In the past I've used a method involving choosing between the first and last tasks.

Whenever the first task is chosen, the next comparison is the next first task with the last task, which of course may now be a different task due to new tasks being entered. The one proviso is that you can't do the same task twice in a row, so if the first task has been re-entered the comparison is with the task immediately before it.

However the rule changes when the last task is chosen. Instead of the next comparison being with the last task it is with the task before the task that has just been worked on. This means that new tasks are "locked out" of the comparison until the first task is chosen again.

For example, say you have the following list:

Email
Paper
Voicemail
Call Joe
Check Diary

The comparison is with Email and Check Diary. You chose Check Diary. You enter another two tasks while working on it and then delete the Check Diary task.

Email
Paper
Voicemail
Call Joe
(Check Diary)
Sharpen Pencils
Re-arrange Paperclips

The next comparison is between Email and Call Joe.

Hope you can follow this!

[Fabio]

Hi Mark, I had a similar problem with my 100 lifetime books list. But I solved it using your random time management method. It has much less overhead and, above all, it does not force me to make choices when I do not want to make choices! ;-)

[Mark Forster]

Fabio:

Doing it the random way is fine, provided you're prepared to stick with the result!

[Cricket]

Sometimes, the value of the random method is it gives you a different perspective. If you really don't want to accept the choice, then you've learned something about that task. Maybe you shouldn't be doing it at all, or you need to break it down or use some other method to make it more manageable.

[Sathya]

Could this also be used to select goals. More correctly select goals that one should let go off pursuing.
I remember reading about Marks goal achievement process, where you decide what you don't want, and narrow down to what you want.
Also reminds of Warren Buffett's 2 list strategy.

[Sathya]

Hi Mark,
The more I read your older articles before Autofocus, I realize you have been inventing and adopting various 'thinking tools' for productivity and goal achievement.
A humble suggestion: Do you think it will be a good idea for you to collate a list of such 'thinking tools' as a book?
I'll definitely buy it :)

[Mark Forster]

Sathya:

Have you read the four I've already written?