The Pattern

It’s definitely a pattern with me.

I start a new interest or hobby, and there is almost always a parade of Amazon packages that follow.

It was true when I got into 3D Printing, more than 6 years ago, when one printer led to numerous accessories, and eventually to owning over half a dozen printers (though I’m down to 2 of my own now, plus one my son owns).

It is true of the CNC machine I’m working on, which has needed loads and loads of parts (to be fair, since it’s mostly 3D printed, that’s to be expected), and never seems to be quite “done.”

And most recently, it’s been the pattern with the home espresso machine we got as a family Christmas present last year.

It’s almost like I’m some kind of gear hound…

A Deep Rabbit Hole

The thing is, that when I start with a hobby or pastime, I usually want to get better at it. Some of that certainly involves practice, and I’ve pulled a LOT of shots of espresso over the last couple of months.

But there’s also always that itch, that sense that with just this one more piece of gear, I’ll be able to get the perfect first layer on my 3D print, or a better cut on my CNC, or pull the perfect shot of espresso.

And so, despite the fact that the Breville Barista Express (which we bought at the suggestion of some family friends who have one and love it) comes with everything you need to make espresso with the exception of the coffee beans, I’ve added numerous items to improve my process:

• Not one, but two digital scales for weighing the beans before grinding, and weighing the output as the shot is pulled (as shown below)
• Spring-loaded tamper for consistent tamp pressure
• WDT tool for breaking up clumps in the grounds and evenly distributing them in the basket (at first a home-made one, and later a commercial model)
• Spray bottle for RDT (don’t worry if you don’t know the acronyms…maybe even be glad if you don’t!)
• Puck screens, to help more evenly distribute the water, and protect the shower screen on the espresso machine from direct contact with the grounds
• Bottomless portafilter and larger basket, to more easily diagnose shot quality…and the most expensive:
• Upgraded grinder for better control and consistency compared to the built-in grinder on my machine

My youngest son, who works at a coffee shop, has on several occasions said (in a way only teens can pull off) that I don’t need all this gear to make coffee. And he’s not wrong.

But the pull can be very strong when you’re trying to improve to view some…thing…as the solution to whatever problem you’re trying to solve.

I am fortunate, both in having the resources to indulge my gear habits, but also in being sufficiently frugal that I tend not to spend a lot on gear (certainly not compared to what you CAN spend on espresso machines, grinders, and gear).

And of course being the gear hound that I am, the first thing I did with my new grinder was use my 3D printer to print an adjustable zero indicator, since the burrs can vary somewhat, so the stock indicator isn’t necessarily a true “zero” point.

Focus on Fundamentals

So, what’s the point? Well, as I think about my couple of months of making espresso, and spending a lot of time reading Reddit posts, watching videos from James Hoffmann’s back catalog, and trying different things on my machine, it occurs to me that sometimes I need to remember the fundamentals:

1. To successfully fix a problem you first need to measure it.
2. When troubleshooting, change only one thing (or as few things as you can) at a time.
3. Set a bar for what “good enough” looks like.

Let’s jump into those a bit deeper individually, and look at how they apply not just to coffee making, but to application development as well.

Measure First

The saying “if you can’t measure it, you can’t fix it” is pretty commonly-known. But not always commonly followed. In my case, one of the things I’ve failed to do so far in my espresso making is to document as many of the variables (and there are many) for each shot, so I can get a better picture of which ones seem to have the most impact on getting good results. If I don’t know with confidence that a shot that outputs 36 grams of espresso from 18 grams of ground coffee tastes better than one that outputs 50 grams of coffee, it’s hard to know how to tweak for better flavor.

By the same token, if I don’t have a baseline for what a “normal” page response time looks like in a web application, then it can be much harder to know if or when it gets “too slow” and to figure out why, and how to fix it. The same goes for database queries. And even if I don’t know what “normal” looks like, measuring the current state of any system can provide a useful baseline against which to measure any attempts at optimization.

One Change at a Time

Once we start looking to improve, whether coffee or an app, it’s best to change one thing at a time, or where that’s not possible, change as few things as we can. The trouble with changing multiple things is that even if we get the improvement we’re looking for, it’s difficult or impossible to tell what was responsible for the improvement. So when we want to further optimize, we may end up moving the wrong lever and getting frustrated.

With espresso, there’s a very long list of things that can change the result you get (which is why machines like the Breville Barista Express automate many variables, to at least help you get consistent, if not optimal results):

• Water temperature
• Grind size
• Dose (quantity of ground coffee in the basket)
• Amount of water filtered through the bed of coffee
• Time spent pulling the shot
• Freshness of beans (how long since roasted)
• Variety of beans (origin, single origin, blend)
• Roast level (light, medium, dark)
• Grinder type (blade grinder…just say no, conical burr grinder, flat burr grinder, electric, hand, etc.)
• Puck preparation (tamping, WDT, RDT, puck screen, paper filters)

I’m probably missing a few more on top of these, but you can probably see that it would be really easy to try changing a bunch of things at once, but even if you got a really great shot, then it would be a question of why? Which one of the things you changed were responsible?

What Does Done Look Like?

One thing that’s been true almost from the very start for me is that I’m not looking to be an espresso drinker. I typically make an Americano with some half and half, and a bit of flavored syrup and sweetener. So a double shot of espresso, around 4oz of hot water, 2oz of half and half, and the fixins. And also almost from the very start, the coffee I’m making with my espresso shots tastes very good, and much better than the Keurig pod coffee that had been my habit, more out of convenience than anything else.

So why obsess over improving the espresso I’m making? Why spent time and money in improving something that’s already good?

Partly, because I enjoy pursuing improvement. And that part is OK. That’s what hobbies are about, sometimes.

But part of it, arguably, is because I didn’t go in with an idea of what “good enough” or “done” looks like. And part of that is the fact that I really have no idea, even after weeks of making espresso, what a “good” espresso shot should taste like. If you read up on espresso, you’ll see people rave about shots that are “sweet” or “creamy” or have various tasting notes. I’m certainly not experiencing that. And I’ve yet to make an espresso where I thought “yeah, I’d like to drink that straight up,” though I’ve had some where I wouldn’t hate it.

But because I haven’t set any kind of a measurable goal, it’s difficult to know whether I’m making real progress. And that can lead to frustration.

Again, I’d draw the analogy to app development around performance, or perhaps around feature completeness. If you don’t decide what “done” looks like, you can spend days or weeks chasing performance optimizations for diminishing returns, or continually say yes to new features requested by a customer, with inevitable delays in delivery. And unlike a hobby, the consequences in app development tend to be a little more serious.

A Reformed Approach

So does this mean I’m giving up my gear hound ways?

Probably not.

But thinking through this, I do think there’s room for me to apply the principles above to my hobbies, and be a little more rigorous in how I approach them…at least those hobbies where I’m aiming for specific improvements for which measuring and setting benchmarks will likely yield faster results.

And increased rigor probably also reduces the chances of catastrophic incidents like this one: