March 17, 2011

Decision trees and self-challenges: how my laptop’s recent battery failure is a great excuse to think

March 17, 2011 - Categories: analysis, decision, geek, kaizen

The battery on my Lenovo X61 tablet refuses to hold a charge, and there seems to be no way to fix it. The battery is no longer covered under warranty, so I’ll need to replace it on my own if I choose to. An easy algorithm for decision-making is be to postpone spending money until I can demonstrate really good benefits for doing so. (Or in this case, nine business days before I really need a new battery.) Because I’m curious about the way I might think about other choices, I’m going to think through some of the strategies I use to make decisions. =)

Decision trees

I like breaking things down into decision trees, similar to the technique described by Ken Watanabe in his kid-friendly book Problem Solving 101. It’s useful to figure out what the options are and what their costs and benefits might be. I realized that I actually have two independent choices: what to do with the battery, and what to do with the laptop. Here is my current decision tree.

I’m probably going to go with choice A.2 for the short-term choice, and we’ll see how my savings work out for choice B. We’re saving for a fair bit of travel this year, so B.2 is more likely than B.1. Fortunately, I work with two laptops, so it’s fine. My basic choice is good. Here’s another technique I use to examine that more closely:

Estimating option value

Hmm. Well, I can still use my battery-less X61T for drawing, writing, and coding. I’ll need to properly hibernate it before transferring locations, or leave it mostly in one place. I just won’t use it out and about as much. I don’t spend that much time in cafes, so it’s really more the shift between the kitchen and the living room or the basement.

So, what’s the estimated gap between the expected value of a fully-functional laptop and a battery-less one? In my case, probably not as much as it would be for other people, because I’ve got my work laptop in addition to this. The upper bound on value for me must be $5/day – definitely can’t be more than that, and is probably nowhere near that number. The cost is probably just a few more extra minutes starting up and shutting down, and a little less flexibility, which doesn’t translate into a large cost because I can use that time for something else. It might even be a net benefit if it encourages me to use a sketchbook during our upcoming trip. =) Worst-case scenario is that it might cost me an hour of work if I forget to save something, but that’s just about discipline.

The value gap might be bigger for J-, but we’ll see if she can handle it. It’s going to be a big gap if we sell this, but then it’s okay to get a new battery closer to that day. Besides, I usually run my laptops into the ground anyway. This one was an exception. I replaced my Eee after a little less than a year), but that was mainly because J-’s need for a computer coincided neatly with my curiosity about tablet PCs.

Setting up challenges

Another way to find out if I’m sufficiently interested in something is to ask questions and set myself challenges. For example, if I want to double-check the potential benefits of the fancy new X220 tablet compared to, say, the lower-prices X220 laptop or my current X61T, I can ask the following questions:

  1. Will I draw often enough to make the tablet worthwhile?
  2. Will I need more than 3 GB often enough to make the upgrade to a 64-bit OS worth the hassle?
  3. Will I run into CPU processing limits often enough to make sense to switch?
  4. Will I need the battery life often enough to make the extended battery life worthwhile?

Answers 2 and 3 seem to be “no” at the moment. VMs would be a good use of additional memory and processing power, but I’ve been doing fine with two computers. If I can cope with a battery-less life, the answer to 4 is probably not significant, unless I find myself going to way more conferences and meetings (and if my scanner proves unwieldy). The answer to question 1 is the most interesting.

I’ve taken lots of sketchnotes, but I’ve done fewer illustrations than I’d hoped I’d draw with the X61T. The workflow isn’t as smooth as keyboard + Cintiq, but it’s (semi-portably) fun. I haven’t figured out how to stop GIMP and Inkscape from jittering so much, although MyPaint and OneNote make beautifully smooth lines. I tend to do my sketchnotes plugged in, but I have a few sketchnotes from meetings where power outlets were few and far between. If I use paper notes for the portable sketches (maybe index cards or notebooks?), then I’ll get a better idea of the incremental value of A.1 or B.1. I can set myself an arbitrary threshold – maybe fill a notebook full of out-and-about sketches and notes – and reconsider my decision when I’ve achieved it. Result: Better drawing skills, a habit of drawing, and an idea of how much I might benefit from the infinitely scrollable paper and the multiplicity of colours on a digital canvas.

I’ve exaggerated the level of thought I usually go to for something like this. There’s room in my “dream/opportunity/kaizen” fund for a new battery if it turns out I absolutely must have one. But it’s fun to think through the techniques I might use to decide something, and writing it down now for something that isn’t critical may help me remember it later when I need to decide something more major. And who knows, it might get you thinking about something… =)

(I might end up getting a lot of value out of not having a battery for this notebook. Look, a blog post, and more reasons to draw/sketch on paper! Stay tuned for progress.)