Using the Right Tool

This post is a story about the the cost of using the wrong tool for the job. The story begins during the last week of October. I was on the Greek island of Kefalonia where we have some land with olive trees. One of my jobs for the week was pruning 37 of the 70+ trees.

Two Novembers ago, a friend started pruning these trees properly to maximize olive production. This means cutting a 30 foot tree down to a 4 to 6 foot stump. Do this to just about any other kind of fruit tree and you kill it. But to a 200 year olive tree, it rejuvenates it.

Within one year of doing this kind of severe pruning, the bare stump now has 200 or so shoots coming out of it from every point, transforming the stump into a bush (see the "before" picture above). The pruner's job is to cut about 190 of these 200 shoots, leaving \~10 as candidates to form the eventual branches (see the "after" picture).

Two years after the severe pruning, the job of pruning the shoots is smaller, with only 50-100 shoots to cut per tree. I'm told that by the third year, the pruning job is de minimis.

During the last week of October this year, my job was to prune 14 "first year" trees, and 23 "second year" trees. It took me 5 days of work to prune the 14, plus 15 of the 23. I ran out of time and asked my friend to prune the remaining 8 second years.

To do this pruning work, I used a hand saw and hand pruners (see the small flash of red at the bottom of the pictures above). The first morning before I headed out to the land to begin the work, my friend offered me his hand axe for the work. I declined.

Four days into the work, I realized that the hand axe would probably work better. So I bought one, and used it to prune the last 5 or 6 trees of the 29 that I completed.

Using the axe to do the same work, I noticed immediately that had I used it the whole time, I could have completed all 37 trees in 2 or 3 days, with less fatigue.

Basically, I had been using the wrong tool for the bulk of the work. It took me 4 days to figure that out.

Contrast that with a business project I'm working on. A different friend brought me in to help a startup with its redesign of its codebase. The work involves some data analytics. The current code uses conjoint analysis — an application of regression.

It took me 5-10 minutes studying their current approach to realize that they could significantly improve their website by switching from conjoint analysis (one mathematical tool) to gaussian elimination (another math tool).

Five minutes for math vs. four days for pruning. Why the difference?

It's because in my long life, I probably have crossed the mythical 10,000 hour plateau doing math and computer science. In contrast, I've spent less than 100 hours pruning olive trees.

This experience reminded me of a recent Quora question along the lines of: "What is the problem with people who learn to code later in life from a book?" The consensus answer tracked my story above.

That is, just like I declined my friend's offer of a hand axe on the first day of my pruning work, newbie coders tend to think they know more than they really do.

When I first read that on Quora, I thought: "Snob coders". But then I thought about the above story and I realized those "snobs" were probably right.