Eating Frogs or Playing Vicious Rock-Paper-Scissors

My blog post titles have been getting weirder and weirder. This one’ll make sense by the end, I swear.

At recommendation by Kenzi at the Center for Applied Rationality, I’ve been reading the book Eat That Frog by Brian Tracy right now, which is based around the following premises:

“You will never be caught up.”
“There is never enough time to do everything, but there is always enough time to do the most important thing.”

However, this only works if you actually have the focusing ability to turn [intending to do the most important thing] into not just doing it but finishing it. And you need to be able to trust yourself to do that. This post explores a particular kind of failure mode that occurs if you don’t have that kind of trust.

The titular frogs refer to the big, unappetizing things that will make a big impact if you finally suck it up and do them—a metaphorical interpretation of the following quotation: (whose ‘author’ will go unnamed here because it’s almost certainly erroneously attributed)

“If it’s your job to eat a frog, it’s best to do it first thing in the morning. And If it’s your job to eat two frogs, it’s best to eat the biggest one first.”

Because when you get to the end of the day, it’s too easy to just say “I’ll eat the frog tomorrow.” The book contains a bunch of tips for how to get yourself to actually eat the biggest, ugliest frogs, as promptly as possible, on a day-to-day basis.

One of them is the ABCDE delegating technique. It recommends, when prioritizing, to go through your list of things to do and mark each one with one of these letters.

  • A: something you must do, that will have serious negative consequences if you don’t do it. Not necessarily today, but sooner than you’d like to think about.
  • B: something you really “should” do. It’ll have some negative consequences if you don’t, but they won’t be as bad as in A. For many people, processing email is in here.
  • C: something that it would be nice if you did them but it doesn’t really make a big impact on your career goals or whatever. Either because it’s busywork or because it’s unrelated to work.
  • D: delegate — something that’s worth doing, but not worth doing yourself.
  • E: eliminate — something that’s not worth doing, when you consider the full opportunity cost of it.

Brian Tracy describes how this relates to frog-eating: you want to do your A tasks first. Don’t work on a B task if you still have an A task left undone. Over time, you’ll develop the habit of working on things that really matter, and become 2-3× as productive as your peers (he says).

I tried simulating how this would go, for me or for someone else. It seemed really likely that, in fact, what would happen is that the person would sit down to do some A task, but it would be hard and aversive and so they would (kind of implicitly following the maxim of “don’t do a B task while As are undone”) escape the discomfort by doing C tasks instead.

If someone observes themselves as behaving this way, then the decision to work on B tasks most of the time suddenly makes sense—if you have to choose between B and C, pick B! And since the choice labeled “A” consistently produces C-like behaviour, you know not to pick it.

But what’s going on here?

Vicious Rock-Paper-Scissors

Awhile ago, I started modelling some of my own irrational behaviour with the popular game Rock-Paper-Scissors, because it seems to capture the weird cyclical (i.e. non-transitive) nature of my preferences. In theory, a rational agent who prefers A over B and B over C would always prefer A over C. Some economic theories also posit that humans are well-described as rational agents. *snrk*

The particular behaviour that first illustrated this structure to me was when I was up late one night, in theory working on a paper but actually on Facebook. I had to get up early the next morning, so why was I procrastinating at this point in the night?

This case study doesn’t follow the ABC thing, but has a similar irrational nature. I had three main options, and as before, you can model me as having two decision-makers—my near-mode brain who wanted to meet physiological needs and/or get dopamine hits, and my far-mode brain, who wanted to meet my long-term goals.

  1. Go to bed. This is what my near-mode brain really wanted to do. Exceeeepppt, I had to…
  2. Work on the paper. This is what my far-mode brain knew I had to do before I could sleep. But it was uncomfortable work, so instead I was more inclined to…
  3. Go on Facebook. This was preferred by my near-mode brain over option 2, although not over option 1 (given just those two choices, both modes would have been aligned on wanting sleep more).

So there I found myself, in a situation where neither part of me was getting what I wanted. My body would much rather have been sleeping than facebooking, but my mind would much rather have been working than facebooking. Yet there I was.

Activity Far’s rankings Near’s rankings
Work 1 3
Facebook 2 2
Sleep 3* 1

*Far-mode values getting lots of sleep, but has the explicit rule of “I can’t sleep until the work is done,” hence it ranks last.

At first I understood this to be a kind of cyclical preference structure, but what’s really going on is a stalemate—a war of attrition—between these two internal decision-making components. I asserted consciously to myself that I can’t sleep yet, but my unconscious mind fought back and refused to work.

What a pickle. I call this vicious rock paper scissors because it’s kind of a three-beat vicious cycle where work beats sleep (says Far), but Facebook beats work (says Near) and then even though sleep beats Facebook for Near and also Far values time spent sleeping more than time on Facebook… Far insists that sleep isn’t an option, so the system gets stuck at Facebook.

There are a variety of approaches to take to this, including spending a few minutes trying to impart to your sleepy brain that it really wants to do the work first, or using external tools to eliminate the temptation in some way. But I want to highlight the one that goes “sleep now and finish it in the morning.” Why doesn’t that tend to feel like a tenable option? Possibly because you don’t trust yourself to actually get up and start working at the right time.

Trusting yourself to do the work

Similarly, there’s a fundamental component that’s required for the Eat That Frog / Do-the-As-first approach to be tenable at all:

You need to be able to trust that you’ll actually do the most important thing—not just for a bit but to completion—rather than flinching and trying to do some very-unimportant thing instead.

It seems that Eat That Frog gets this on some levels, as just 14 pages later there’s a section called Work All the Time You Work, pointing out that you don’t want to be spending your intended-work-time not-working such that you have to spend your intended-leisure-time working. But I think this particular structure could use special attention, hence this post.

Trust doesn’t emerge from nowhere—you have a model of yourself built up of years of observing your own behaviour. There are two ways to shift this. One works best with sudden intensity; the other is very gradual. You don’t want an in between, where you just sort of try to do things differently now.

Either say, “Oh, I was doing this silly thing because I didn’t realize what was actually going on, but now I’ve read this blog post and I do” and then dive in the next day with a massive undertaking to demonstrate that you have a new approach now.

Or, if you’ve doubts that that’ll work, then start from wherever you are, and use a success spiral. Rather than trying to tackle the biggest thing immediately, try something just outside of your comfort zone, and then something a little bit further, to build up the trust over time. If the one-shot deal fails, this plan makes a decent backup. But don’t plan to fail!

Changing the game

If you catch yourself doing Bs to avoid Cs, then that makes sense in the world where As aren’t a realistic option. But you might want to also spend some time making As be an option.

If you’ve got other examples of this weird behaviour pattern, I’d love to hear them in the comments!

If you found this thought-provoking, I invite you to subscribe:    
About Malcolm

Constantly consciously expanding the boundaries of thoughtspace and actionspace. Creator of Complice, a system for achieving your important goals.

Personal Website



6 Comments

Huey Kwik » 21 Sep 2014 » Reply

The categorization into A and B tasks is interesting. This seems like another way of trying to assess the cost of delay of not doing a given task.

For me, the gem of the article was the idea of building trust in yourself as someone who gets big things done.

Robert Best » 22 Sep 2014 » Reply

I definitely find myself procrastinating with B & Cs, though I don’t have any nice examples to share. I just wanted to let you know, in some way, that I had read this and found value in it. I’d prefer not to clutter your comments with things like “Great work! Thanks!”… so if there is a better way to send you these simple signals, then let me know!

Also, this ABC… labeling reminded me of this:
http://www.dougengelbart.org/about/abc-model.html

    Malcolm » 22 Sep 2014 » Reply

    Haha, I think the best way for you to send me the signal that you got value of my post is to share it with others—either people you know personally that you think would benefit from it, or just broadly on facebook.

    Let’s see… yeah, totally different model Doug has, but also interesting!

Tessa Alexanian » 22 Sep 2014 » Reply

I read this post late last night when I was supposed to be doing work. There’s something rather beautiful about how people are likely to be reading this post while participating in vicious rock-paper-scissors themselves. Certainly drives the point home…

Qiaochu Yuan » 24 Sep 2014 » Reply

“Vicious rock paper scissors” already has a name – it’s called Condorcet’s paradox, and it’s great (by which I mean terrible). Nice job rediscovering it!

    Malcolm » 24 Sep 2014 » Reply

    Oh, good point! I think the part about the multiple voters being subcomponents of one person is new though… I wonder if maybe there’s a third voter I’ve failed to model.

Have your say!

Have your say!

Message

Name *

Email *