Why I Never Pull the Trier
Pulling the trier — drawing a sample from the drum at intervals to check color and smell, gauging where the roast stands — is close to a standard move in roasting.
It’s so common that almost no one stops to ask: what does the act itself, pulling beans out of the drum, actually do to that batch?
It does three things at once. None of them good.
One: it disrupts the environment inside the drum.
Good airflow works by steady negative pressure — a slight, constant suction that carries off smoke and off-notes while letting heat exchange happen undisturbed. That undisturbed state is where a stable roast finds its quality.
Pull the trier, and a gap suddenly opens in the drum wall. The negative pressure breaks instantly; room-temperature air is pulled straight into the drum. Pressure and temperature both collapse in the same moment — the balance you’ve been carefully holding is punctured.
Two: it damages the bean itself, irreversibly.
A bean approaching first crack is building heat and pressure internally; its cell walls are like a sponge slowly unfolding from a compressed state. Pull it out into cold air at that moment, and the fragile, still-opening structure collapses on the spot — and a structure that has collapsed can never unfold the same way again. The beans that get sampled are, in effect, damaged mid-journey.
And what you get back rarely justifies it: most of the time, pulling the trier means glancing at the color and tipping the sample back in — there isn’t much information in that glance.
A bean has traveled a long way under a farmer’s care. It deserves to be treated with some respect in the last few minutes before it opens.
Three, and the least considered: even setting the first two aside, the information the trier gives is always late.
The logic is simple. The bean you pull out shows you its state at the instant it left the drum. By the time you bring it close, look at it, smell it, and form a judgment, a second or two has already passed. By the time you adjust airflow or heat in response, the moment you just read is no longer the present moment. There’s an inherent lag between sampling and reacting — what you’re reading is always the instant just past, never the instant happening now.
So what do I rely on instead of the trier? A signal with zero lag: smell.
Coffee is an intensely aromatic substance. Through the entire roast, it is constantly expressing its real-time state through aroma. Where the batch stands, how far it is from crack, whether to pull it now or wait — it’s all written in the aroma as it shifts, moment to moment. And it’s the whole batch speaking, not the partial testimony of a few sampled beans.
In fact, a batch has been telling you, through smell, where it stands long before the first crack ever sounds. The crack itself is just the moment the batch finally gets loud enough for everyone to hear.
Beyond calculating thermal inertia, what a roaster should really rely on is the nose: a real-time flavor monitor, more accurate and more immediate than any single sample, because it carries none of the lag built into “pull it out, then look.” The trier freezes a single instant and reads it a step behind. Aroma is a whole river, flowing in front of you right now.
So I leave that gap closed — not merely out of caution, but because there’s no need to open it. The batch is speaking from start to finish. All there is to do is listen quietly, and then, at exactly the right moment, let it stop — no more, no less.