The variables

Temperature, speed, angle, pressure, milk type, cup size, pour height, the
vortex, the noise the pitcher makes, the texture of the milk, whether you
moved the cup, when you stopped steaming, how full the cup was when you
started.

Espresso has three variables. Milk has all of these.

The one that matters most

The angle of the steam wand to the milk surface. Tip it wrong and the milk just
heats. Tip it right and it spins and incorporates air and the bubbles stay
small and the texture becomes silk.

Why I do not make cappuccinos

I make espresso and I put hot milk in a cup. The result is a cappuccino in the
sense that it has espresso and milk, and nothing else in the sense that it does
not taste like one.

A real cappuccino requires the milk to be steamed and textured and poured in a
way that keeps the texture. I do not have the hand for it and I have made peace
with that.

What people tell me

“It just takes practice.” This is true. “You just have to relax.” This is also
true. “It is easy once you get it.” This is not true. It is hard the first time
and hard the fiftieth time, and then one day you are not thinking about it and
you are doing it right.

The steaming that changed everything

A friend showed me that the sweet spot for the wand is half-submerged, not a
quarter-submerged, and the angle is 45 degrees, not straight down. The
temperature matters less than the spin.

One cappuccino with that angle and I tasted what was possible. I have made
maybe two cappuccinos that good since then.

What I do instead

Espresso drinks where milk is poured, not steamed. A macchiato is espresso
with a dot of foamed milk, which does not require texture. A cortado is
espresso with milk in equal parts, which requires neither temperature nor
texture to work.

The cappuccino is a goal. Everything else is the path.

A technique that requires decades is beautiful to watch and painful to
practice. A technique that requires days is the one that will actually get
used.

The one that ran too fast

30 millilitres in 18 seconds. Watery, sour, wrong. I thought the machine was
broken until I backed up and realized—I had changed the dose, lower, and lower
dose means less resistance means faster flow.

Lesson: Dose affects flow more than people admit. Consistency in dose is
consistency in flow.

The one that choked

12 millilitres in 28 seconds, the espresso stopped flowing. I panicked and
thought the machine was clogged. It was not. I had changed the grind too fine
and the water had nowhere to go.

Lesson: There is a fine line between “extracts perfectly” and “does not flow
at all” and the line is a notch on the grinder. Adjust by half-notches when
you are close.

The one that tasted like caramel

A three-day-old roast, dialed in perfectly, tamped level, pulled at exactly 28
seconds to 2 ounces. It tasted like caramel and nuts and nothing else was
there to distract.

Lesson: There is a day in the life of beans when everything comes together,
and that day moves as the bean ages. Tasting is how you find the day, not the
manual.

The one I almost discarded

Pulled it, tasted sour, almost poured it out. Then I tasted it again ten
seconds later and the temperature had changed and it tasted perfect.

Lesson: An espresso cools fast. Taste it in a warm cup and wait fifteen
seconds before deciding.

The one that broke the pattern

All my shots tasted the same—pleasant, fine, nothing wrong. Then a friend with
a better grinder pulled a shot in my machine and it tasted a different kind of
good. I used the same beans, same dose, same grind number, same tamp.

The grinder was different.

Lesson: You cannot taste what the machine does not grind. The grinder is the
limit, and when you hit the limit, the limit is the only thing worth buying.

The pattern that emerged

Every shot teaches something only if you are changing one variable at a time.
The shot that runs too fast and the tamp was also wrong and the beans are two
days old is not a shot, it is noise.

The shot where everything was the same except the grind is a sentence in a
language the espresso machine speaks.

Listen to that language and the machine will teach you everything else.

What the grinder does

It cuts beans into pieces that are all roughly the same size. A lot of the
pieces will be finer than you want and a lot will be coarser than you want, but
the average will be the size you set.

The better the grinder, the smaller the spread. A machine that costs $50 cuts
everything into a range from powder to chunks. A machine that costs $300 cuts
everything into a tighter range. A machine that costs $1000 cuts into a very
tight range.

The effect on the cup

A tight range of particles extracts more evenly. Even extraction tastes clean.
Uneven extraction tastes muddy or sour depending on which way the unevenness
goes.

Which grinder to start with

If you drink espresso three times a week, a grinder that costs $150 is the
minimum. It will be loud and slow but the grind will be consistent enough to
learn from.

If you drink espresso once a day, a grinder that costs $300 to $400 is where
you start to enjoy the coffee and stop fighting the machine.

If you drink espresso more than once a day, you probably already have a grinder
and you know this better than I do.

The grinder I use

A grinder that costs $500, and it is the second-best machine in my kitchen. The
espresso machine is the first because it is a ritual. The grinder is the one
that makes the espresso taste good.

The one I used to have

A grinder that cost $100. It made bitter espresso because the fines (the tiny
particles) were too much, and the fines came from the blade cutting unevenly.

One day the espresso suddenly tasted better. I did not change anything. I
cleaned the chute and the grinder died. I bought the $500 grinder and
everything changed.

The question that matters

“Will I change the grind throughout the day to match the beans?” If yes, you
need a grinder good enough that changing the grind is fast. If no, a cheaper
grinder is fine.

The grinder that takes three minutes to dial in between shots is a grinder you
will not use, and the espresso you do not pull is the worst espresso you can
make.

What I was taught

Tamp down. Use 30 pounds of pressure. Tamp level. Tamp with a twist. Tamp
three times if you are fancy.

What actually matters

• Consistency, so the shot tastes the same way twice
• Pressure perpendicular to the puck, so the water flows through evenly
• The pressure is less important than the evenness

The shot that failed because the tamp was 25 pounds instead of 30 is usually
fine. The shot where the tamp was tilted to one side will taste wrong every
time.

The argument about pressure

The people who say pressure does not matter are halfway right. The people who
say it does matter are also halfway right. The thing they are arguing about is
not the pressure, it is the consistency.

A machine that backs off the pressure at the start of the pull—like some modern
machines do—changes the argument. A machine that holds steady pressure
throughout changes it again.

How I tamp

1. Grind into the portafilter
2. Distribute the grounds with my pinky, knocking the side of the portafilter
   to settle them
3. Tamp straight down, 30 pounds, check the level
4. A small twist to break the seal
5. Lock and pull

The twist breaks the seal, which some people say is unimportant. The twist is
part of my ritual, and the ritual is part of the consistency, so the twist stays
even if it does not matter.

The tamp that changed everything

Using a level tamp instead of a conical one. The level tamp is wider and
lighter, and it is much harder to tilt. The first shot with the level tamp
tasted better than anything I had made in a month. The reason was that I was
not tilting anymore.

The pressure did not change. The consistency of the pressure changed.

The variable that matters is the one you stop noticing because you have done
it the same way so many times it is automatic.

The protocol

Pull the first shot with the existing grind. Check the time and the flow. If it
runs too fast, grind finer. If it runs too slow, grind coarser. Change by a
small amount. Pull again.

What I actually do

Pull a shot, taste it, think about what changed, adjust, pull again. The tasting
is the only part that matters. The time is a lie you tell yourself—what matters
is the taste and the flow, and flow is observable in the group head.

The shots between dialing in

Every shot between adjustments is data. An espresso machine that runs 28
millilitres in 28 seconds is running at the tempo that brings out sweetness.
One that runs in 22 seconds is running too fast. One that runs in 35 seconds is
running too slow.

The grind that gets you 28 seconds with this water, this roast, this machine,
this pressure—that is the grind.

What changes the grind

• Fresh beans, every day for two weeks
• New beans, which need a new grind
• The humidity in the room
• How much you tamped yesterday’s machine

The grind is not a number, it is a moment. The moment you change anything that
touches the flow, the dial moves.

The protocol I actually follow

1. Grind
2. Dose into the portafilter, pile it loosely
3. Tamp: 30 pounds, two more pounds of twist
4. Lock and pull
5. Watch the flow and taste the result
6. Adjust by half a notch

The tamp is the thing I change by accident more often than on purpose, which is
why the shots are inconsistent, which is why I keep a scale and weigh my
doses—not because consistency is the point, but because I cannot taste
consistency if I do not know what changed.

The one rule that holds

Changing one variable is harder than it sounds. You change grind and you change
tamp pressure and you change temperature because the one you changed first did
not work.

The shot that teaches you is the one where you changed exactly one thing, it
failed, and you changed it back. That is data. Everything else is noise.