Skill Acquisition in Plumbing
Notes from my pre-apprenticeship.
Acquiring any skill follows the same process of:
- identifying key constraints and goals,
- repeating with high focus under varied conditions,
- emphasising errors and failure modes,
- and changing one thing at a time when there is a problem.
Another common principle in skill acquisition is that you must visually track what you’re acting on, not the body part that you are moving. For instance, when learning the piano, you watch the keys; when drilling a screw, you watch the screw. This is because you can only track success/failure of the output and improve if you’re watching the output. However, there are some habits specific to plumbing that take some time to install, especially for someone like me who comes from music, computer programming, and academic/STEM fields. I went home and took note of things almost every day; this is a summary of what helped me most.
Force requirement
Many tasks require twice as much force or aggressive commitment as I think they need, and then some:
- The rivet gun needs a lot of leverage;
- the molten pool needs to be really molten;
- the striker needs a hard scratch before it lights up.
It’s not like on a musical instrument where too much force just breaks the instrument. Moreover, rigidity/stiffness is often more important than flexibility. When bowing a violin, I have to keep my wrist loose and flexible. When using a hammer, I didn’t realise I had to keep my wrist more rigid/locked, because the point is maximum force transmission, not precision. Rather comically, for instance, the correct way to get roof screws to bite the metal is to keep yourself rigid and to use as much of your bodyweight as possible to just force it in. (Obviously, once it’s in, you want to ease up the load, since you don’t want to crush the bottom.) I lost a lot of time at the beginning of the course looking for more “elegant” solutions.
Big adjustments
As a consequence, small adjustments often make little difference. It’s not like in singing, for instance, where a millimetre-scale difference in tongue position can produce a totally different result. If something isn’t working, I need to double or even triple the force, change the tool straight up, or use a completely different angle. This kept causing me to repeat the same mistakes over and over because I thought I was changing something but I wasn’t. For example, when I was struggling to use a rivet gun, I would try a slightly different angle, and if that failed, I might use slightly higher force. If even that failed, I would run out of ideas. But the actual problem wasn’t the variable I was changing, but how much I was changing it: I should have tripled my force, using both hands and my bodyweight, and if that still didn’t work, I should have tried changing the rivet gun. Even if a lot of force is required, things shouldn’t require a convoluted, heroic struggle. Such overcomplicated struggles, even if they work once, are unsustainable—they expend too much cognitive, physical and motivational energy. You should change something. It is like losing weight: fighting against hunger by force every day might work, but it is far more sustainable to eat according to appetite and figure out why your appetite is dysregulated.
Visual demonstrations
Prior to plumbing I’d never learned any skills that require following visual demos. Violin teachers, for instance, tend to just adjust the instrument on your body rather than showing you. As a result, I didn’t realise how useful it can be initially just copy everything more or less exactly as is, from the visuals of the process to the angles to the stance. This often beats just copying the parts I think are important. Almost always, that leads to missing something that seems minor in the posture or movement that was actually load-bearing. For instance, when wrapping Teflon around threads, I didn’t realise I had to keep it tense by holding it on the inside with my thumb, because I was just copying the demo vaguely instead of precisely. When copying demo projects/photos, it is often worth either copying as many little details as possible or asking whether they are relevant. This does not mean that official demonstrations are always ideal. It means you have to improvise starting from something that basically works, even if it’s imperfect. It’s the difference between:
- a conspiracy theorist who thinks the Earth is flat (ignoring the standard method and trying to figure stuff out from scratch),
- and someone who accepts that we have built a shocking amount of infrastructure on the belief that the Earth is roughly round, but then questions measurement errors (accepts the standard method, then figures out what doesn’t work in that method).
Planning and sequencing
Actions in plumbing are often costly or time-consuming to reverse. Lower reversibility means it is more important to spend some time at the beginning chunking the task into smaller parts, and determining which tasks should be performed in what order. It is also paramount to spend some time thinking hard about geometry: what hidden extra lengths might be necessary? Where is length not as important? In contrast, when writing an essay or a computer program, I can often just sort of start chucking things on the page, and if there’s a problem I can instantly move stuff around. And when learning a musical piece, the order is already set for me, and I don’t have to think about sequencing. This sequencing flexibility inherited from other domains often becomes a bad habit in plumbing.
Tool geometry
Each tool works best at a certain angle and with a certain geometry. Because we often use the same tools, a tremendous amount of time can be saved by getting this right from the start:
- Aviation snips, for instance, often work better when kept nearly flat to the sheet, and should not be closed to the tip.
- They also suck at sharp turns; it is better wherever possible to curve when switching directions. This is because the snips need a certain amount of room in the opposite direction to not get stuck, which 90-degree angles do not provide. For especially short cuts at 90 degrees, a slight curve is often acceptable as a compromise.
- A hand saw is designed to work best when its entire length is used to cut.
- Junctions should be oriented such that water coming in from the side does not fight against the direction of the main pipe.
- Rivets must fit into the holes in both sheets and sit flush to the sheet before being driven in.
I also found the wide variety of wrench-like tools confusing:
- The terms wrench and spanner are often interchangeable.
- The ring spanner is the one with the closed loop. Because it grips all or most sides of the nut or bolt head, it produces a stronger grip.
- The open-end spanner has a U-shaped jaw that allows you to manoeuvre better in awkward locations or where you can’t go all the way around. Often, you use an adjustable spanner/wrench, also known as a shifter, which is more versatile but more likely to slip.
- Both the Stillson pipe wrench and the Footprint pipe wrench will chew or mark the surface with serrated jaws, and so are no good for finished chrome or plastic.
- The Stillson jaw is finely adjusted with a knurled screw wheel. A Footprint has a two-part, scissor-like action: pressure makes it grip harder, and releasing pressure makes it quick to lift and reposition.
- Using two Footprints in opposite directions is a good way to loosen something that’s really stuck. Otherwise, there’s nothing counterholding the motion on whatever your target is attached to, and the whole assembly may rotate pointlessly instead.
- Ordinary pliers have relatively limited jaw opening. Multigrips are the adjustable pliers that we use for a variety of tasks. Like pipe wrenches, these are also serrated.
- The duck-bill clamps are useful for folding, weathering (turning up), and clamping, although other tools such as multigrips may be used for this purpose depending on the geometry. (Confusingly, these differ from duck-bill pliers.)
Tool geometry often needs to be memorised by rote. Although it is often useful to try to understand why things are a certain way, with these features, there is basically no deeper reason than they’re designed that way, and that’s how they work. Much time can be wasted trying to “understand” something on a deeper level that doesn’t have a deeper level. For example:
- Drilling and riveting work better when the tool is straight and aligned with the desired direction of the fastener.
- When using the folding machine, it is important to ensure that flaps do not get stuck on the sides of the machine or bump up against other flaps.
- The circular saw must be running before the blade touches the wood;
- but in contrast, you start the drill after you’ve fully committed with your bodyweight into the target.
It is also important to ask: what needs to move against what, and in what fashion? No matter how much you adhere to the Correct Method™, you will keep making ridiculous errors if you do not ask this question:
- It is no use trying to clamp the folding machine down to make a safety fold if the metal is past the clamp, which is surprisingly narrow.
- It is no use trying to seal a flange into a Pittsburgh seam by hammering in the direction that the flange comes out, unless you’re manually holding it in from the other end.
- It is no use hammering something only vaguely in the direction that you want it. For instance, in the case of the Pittsburgh seam, once the seam is partially bent over the flange, change the angle of the hammer and/or the metal to hit it more diagonally.
- If you need to fold a section that doesn’t form a straight line across the entire piece of sheet metal, you should cut slots and use duck-bill clamps instead to isolate the section instead. The folding machine will obviously act on too much of the sheet.
- It is no use trying to drill somewhere where the drill chuck will get caught on the wall or some other obstacle, or make you unable to aim straight. Pull out the drill bit a little, or use a longer drill bit. Make sure it’s really fastened. Then try drilling.
Reducing context-switching
The time spent moving around, switching tools, re-measuring, and so on adds up very quickly and leads to mistakes. For instance, instead of going back and forth between cutting pipes and hanging them up, it is much faster to write down all the necessary lengths, cut them all at once, and then hang them. When making flashings, get all the pieces of sheet metal first, mark them out, cut them, and then start riveting, drilling or folding.
Positioning/strength
Working against gravity is harder:
- It is often easier to hammer, braze, or drill from above;
- it is easier to use a copper cutter closer to your body so your extended arms aren’t dragging you down (torque);
- it is often helpful to push down on a flaring tool with your whole body for leverage.
Ridiculously, I scored in the high 40s in VCE Physics and still had to be reminded of these basic applications. This is the problem with modern teaching: it incorrectly segments people into “visual” and “abstract” learners, which is insane, because the abstraction (mathematical formulae) should deepen an understanding of the referent reality (objects moving around), not replace it with disembodied, decontextualised symbols. My physics classes should have begun with visual demonstrations of moving objects; instead, there were approximately zero. Little things like:
- With scissors or snips, you get the most leverage squeezing from the far end (longer distance from the pivot/fulcrum), and the most force is near the pivot/fulcrum (closer) where there is the least movement. Think of it like this: a larger hand movement with moderate force produces a small blade movement with large force.
- If you’re holding up a slope with a piece of timber, but it keeps getting knocked over, nail it into another piece of timber into a T shape. It’s like when you spread your legs apart on the bus to stop falling over: the wider base lets your body move further sideways before your weight passes outside your feet and gravity starts pulling you over.
Change the geometry of the problem to favour you. Clamp it. Brace it. Put timber under it. Move it closer. You want less stuff to keep track of—as few moving parts as possible. Most of what people call “strength” is actually about skill rather than muscle size. Within the middle range, muscle size actually has shockingly little to do with strength/power. Bruce Lee was tiny but had immense power. Working out is only helpful if you’re training the correct skills—muscle size is nice to have but often unnecessary, and a surprising number of gym exercises are unhelpful in plumbing. For instance, I found active hangs and table rows to be great for rivet guns, saws, and snips, as they train grip strength and pulling endurance. I routinely saw more muscular guys struggle with snips and rivet guns because they didn’t understand leverage, or didn’t properly train grip strength/endurance. Again, it’s all about sustainability. No matter how strong you are, it’s better to reduce as much friction as possible throughout the day. Use your bodyweight for the rivet gun or flaring tool. Put a stand below your slanted sheets so that you can rivet them together without the bottom one drooping lower. Change your position so you can drill straight instead of accepting an awkward tilt.
The blessing and curse of the human mind is that it saves energy by reducing attention on routine things that don’t appear to matter much. The problem is is that if you start doing things absent-mindedly, you get screwed the second the angle changes. You have to remain mindful. Mindfulness is awareness of awareness—that is, control over attention. Much of modern “meditation” is a rehabilitation effort for phone-addled people who can no longer sustain awareness or control of their attention. Unlike older meditative traditions, modern meditation is simply marketed as “stress relief” with no deeper goal. However, simply being mindful is useless; you must know what to be mindful of and keep your attention there. What needs to be done here? How can I make this task easier? What are common failure modes? Too often, I would find myself riveting or drilling at an uncomfortable angle with my arms too far away and with insufficient support. Why? Because I was letting too many shallow thoughts drift into my mind instead of focusing on what was important. Without conscious control, your attention drifts to whatever the monkey brain finds most urgent.
Measuring, marking, and Keeping It Straight™
Measuring and marking make up much of plumbing. As for tools:
- A HB pencil is ideal: any lighter and you can’t see it, but any thicker and you lose precision. It’s also light enough that you can rub it out with your gloves instead of using an eraser.
- Flat, straight steel rulers are most convenient: with thicker rulers, it is harder to mark precisely because the measurement is higher up.
- Foldable rulers are a pain. They are annoying to straighten, and often don’t stay straight. Get a flat, sturdy foldable ruler, ideally where the circle in the middle doesn’t interfere with marking.
Some tips on marking:
- Try to measure from the same datum wherever possible: if you need to measure 150 mm to the right, then another 100 mm to the right of that, measure 250 mm from the datum rather than 100 mm from the first marking. Measurement errors add up.
- When trying to mark long lengths, make sure to physically hold down one end of the ruler so that it doesn’t move while you adjust the other.
- If possible, to compensate for real-world deviation, when marking something out to fit some other object, it is better to use that object as a direct reference rather than going by the book.
- When marking a straight line, it is often possible to avoid marking out the whole line: when using the guillotine or folding machine, you only really need a small mark on either side, as the machine provides the straight line for you.
- When trimming sheet metal to fit against an existing angle, you can usually sight the angle and cut a roughly matching line rather than marking precise start and end points.
Some tips on procedure:
- It is helpful to mark out the lengths on the pieces you have cut.
- You may also mark out things like the waste side, purpose, and anything else that helps to reduce the load on your working memory.
- There is no harm in taking longer to double-check your markings and making sure their purpose is really obvious.
- For example, if you’re cutting sheet metal with a guillotine, mark with big arrows the ends you want to cut through, and make sure they’re in the right place.
- Confirm orientation, direction, location, and so on.
Unless you have a perfect photographic memory, all of this saves a tremendous amount of time, and you won’t have to constantly redo your work. Finally, some tips on Keeping It Straight™:
- Implicit references are the neglected hidden variable. For instance, when folding 90 degrees on the folding machine, instead of overthinking it, you can use the machine bed or any other level surface as a reference.
- When trying to drill multiple holes in a straight line, you may be able to just eyeball a short distance from a common reference line, such as a fold or some other visible feature.
- When checking whether multiple objects are in a straight line, the first and last object are the implicit reference. Everything must be on a straight line in between.
- Often, straightness is more about visuals than function. In such cases, millimetre-grade precision may be unnecessary. Just choose an implicit reference and eyeball it—if you can’t notice any major issue, neither will anyone else.
Improvisation
Most “overthinking” is actually a problem of not knowing which constraints are important. Many measurements don’t have to be millimetre-precise. It’s okay if there’s one too many rivets. The solution is to ask: what needs to be satisfied? Where does the water have to flow? Where does the timber have to fit in? And then even if it doesn’t go perfectly according to plan, you can improvise and find a way to satisfy this constraint:
- For instance, if you need a plank of wood to fasten your flashings to, but only have a bunch of small pieces, joining the small pieces may still very well work.
- Or if a flap of sheet metal is colliding with another flap while folding, you may be able to just snip a bit of it off.
- A bad fold can also often be fixed with a hammer, or by folding in reverse.
- A slightly off fit can be hammered together.
A good chef can repair a dish by adding more salt or water or herbs or spices. A good musician can pretend that his mistake wasn’t actually a mistake by improvising a different melody. An expert isn’t someone who makes no mistakes—this is impossible—but someone who has multiple contingencies for each one. Once you have decided on an improvisation, you must commit. If you do something halfway and predictably fail, you have learned nothing about whether it would have worked.
Creativity does not come out of nowhere. It is what happens when you have seen many ways of doing things in various scenarios, and pattern-matching exploratively. If you know you can tear paper straight by first cutting a slot into a crease, you might also try snipping a bit of corrugated sheet metal and manually tearing off the rest. This is useful when you have no shears. The more books you’ve read, the more creative a writer you become. The more songs you’ve listened to, the more creative a musician you become. A novice plumber is terrible at improvising because he hasn’t yet seen enough. He must start with the standard method that definitely works, and build from there.
Speed
Often, instructors provide crude instructions such as “don’t overthink!” or “go faster!”, but it is easy to obey too literally. Thinking too little is also catastrophic. Thinking hard is not the problem; being stalled because you do not understand the important constraints is the problem. Similarly, going faster should not mean rushing. It is about removing unnecessary time sinks that don’t affect the quality of the output: context switches, bad tools, poor planning. Sometimes I would see students finish before me and feel like I was falling behind, but then I would notice that they compromised far too much quality to get there. I eventually became much faster at sheet metal, for instance, because I initially took that extra time to try to get things right, even if I looked ridiculously slow.
Not caring about what other students think, say or do is an excellent skill to have. An inability to do this can sometimes be due to insecurity that results from not having anything else going on in your life. Counterintuitively, it is psychologically helpful to have multiple other skills or projects that you’re working on so you feel less pressure for plumbing to work out. This way, you can feel comfortable being slower initially and outpacing everyone in the long run. You’ll also be less tempted to get sidetracked by the constant discussions about “do you have an apprenticeship yet?”, “are you looking?”, “how was your trial day?”, “what do you think about Peter?”, and other apparently “work-related” discussions that are often more about insecurity and impatience. Because these discussions appear work-related, it is far easier to get caught in them instead of cutting sheet metal or cleaning up. Very rarely does anyone talk about anything useful; there is little to miss out on unless it’s the teacher talking. It is important to remember that most people rush into things out of desperation and end up with worse outcomes, even if they look better initially—rivets that are not quite straight, useless degrees, mediocre relationships, overpriced stocks, overspecialised apprenticeships, and so on. If you know your plan is better in the long run, it doesn’t matter if everyone else hits “milestones” before you do.
Coda
Many of these points seem like total common sense to someone with prior experience in hand skills, but to me it was like walking around in a permanent haze of confusion, which is why it took me almost two weeks to figure out something as basic as cutting sheet metal with aviation snips. The problem is that even once I got really fast with sheet metal, I kept failing to apply the same insights to other skills because old habits were so deeply ingrained. There was basically no fix for this other than repetition.
I eventually outpaced most of the class because I showed up, paid attention, looked for patterns in my mistakes, tried not to repeat errors, and brutally tracked failures/successes. This minimum bar is now apparently impossible due to TikTok addiction. People keep saying that phones destroy our attention spans, but it goes deeper than this. Just like sitting down all day atrophies your muscles, phone addiction destroys our capacity for cognitive effort. Instead of memorising facts, we can Google them; this means we never end up developing a conceptual network in our heads. Instead of doing mental maths, we use our calculators. Having to pull out a calculator for minor additions wastes time. Phones cripple our ability to pay attention to boring stuff that requires time and effort. We skim just the “important parts”, but the boring parts are the necessary connective tissue. They ruin our sleep, which further dents cognitive capacity. And notifications can distract you at any time, because people no longer understand that phone calls can wait, emails can wait, text messages can wait, and no one will die.
Most of what I summarised here came from noting the instructors’ corrections and looking for common patterns, not from some mystical trade genius, of which I very obviously had none. I would have been useless without the intermittently excellent remarks of the hopelessly spread-thin instructors; it was up to me to piece them together. This also supports my idea that “natural talent” isn’t that important beyond initial acceleration: I started as the absolute slowest, and the only students that I couldn’t catch up to were the ones who had far more prior work experience than me. Nor is there such a thing as a “self-made man”; contrary to modern narcissism, no one learns anything without the help of others.
There is no magic point in time where everything clicks into place at once. People who acquire skills faster than others do so because they never stop finding ways to accumulate improvements over time. Improvement never becomes effortless, but it’s emotionally easier because your outputs are now basically usable rather than catastrophically embarrassing, and familiarity means it’s much easier to diagnose and repair errors. Moreover, there is often little functional difference between someone who can cut something 95% straight and 98% straight. Being able to cook well enough is useful, but there are diminishing returns. Being able to cook like Gordon Ramsay is a fun goal, but is not 10 times more useful for ordinary purposes. There is something to be said about aiming for the stars (the standard) but being happy you landed on the moon (your actual worth).