# Pour-over (/en/docs/foundation/phase-5/chapter-27)
## Phenomenon [#phenomenon]
Two people use the same beans, the same grinder at the same setting and the same recipe to brew a pour-over each.
The first pours slowly in small circles close to the coffee. The brew finishes in two minutes forty-five seconds and tastes clean and sweet.
The second holds the kettle high, pours a thick, fast stream and gives it a few stirs at the end. It takes four minutes to drain and tastes bitter and drying.
The numbers were identical. The only difference was **how the water was poured**.
## Variables [#variables]
Pour-over is **percolation**: water does not sit with the coffee; it keeps flowing down through a layer of grounds, carrying what it dissolves into the cup.
In that layer, several things pull on each other:
| Variable | What it is |
| -------------- | ----------------------------------------------------------------------- |
| Bed | The layer of wet coffee in the dripper: how deep, how flat, how compact |
| Pour structure | Stream thickness, height, how many pours, where you aim |
| Agitation | The churning from pour impact, stirring and swirling the dripper |
| Flow | How fast water moves through the bed |
| Drawdown | The total time from first pour until the water has drained |
Grind, ratio and temperature still matter, but this chapter looks only at pouring.
## Experiment [#experiment]
A thought experiment: same coffee, same total water, same total time, two ways to pour:
* A: a thin stream close to the bed, poured gently in several stages;
* B: a thick stream from high up, with a stir after each pour.
Which extracts more? Which drains more slowly? After brewing, how will the two beds look different?
## Observation [#observation]
B has far more agitation, so coffee and water mix more thoroughly and extraction is usually higher. But B also stirs up the fines, letting them travel down with the water and settle at the bottom of the bed and on the filter, which slows the drawdown. A slower drawdown means longer contact and more extraction on top. Stack the two and B easily overshoots.
Look at the beds afterwards: A's is usually fairly flat; B's may be uneven, with a ring of grounds stuck high on the walls and a dip in the middle. A flat bed usually suggests water passed through fairly evenly.
## Model [#model]
### Why percolation is efficient [#why-percolation-is-efficient]
In immersion, the water gets more and more concentrated, so dissolving slows down over time. Percolation is different: fresh water keeps arriving from above and carries away what has already dissolved, so the water around each particle stays relatively dilute and dissolving keeps going. That is why percolation often extracts more with the same amount of water.
### What sets the flow [#what-sets-the-flow]
How fast water passes through the bed depends on:
* **Grind**: finer particles mean smaller gaps and more resistance (Chapter 20);
* **Bed depth**: more coffee or a narrower dripper means a longer path;
* **Where the fines are**: fines washed to the bottom act like a thin layer of mud over the exit;
* **How much water sits above**: a higher water level pushes down harder.
Mathematical modeling of drip brewing identifies particle size, water flow and bed shape as the main factors in extraction, and espresso research found that grinding too fine can make the flow uneven. The same logic applies to pour-over: water follows the path of least resistance. Where resistance is unusually low, a channel forms — the coffee there is over-washed while the rest is under-extracted.
### Drawdown is a gauge, not a dial [#drawdown-is-a-gauge-not-a-dial]
You cannot "set" the time of a pour-over directly. Drawdown is the **result** of grind, bed and pouring working together. That is exactly why it is the most useful gauge in pour-over: when the same recipe suddenly drains at a different time, something changed in the bed.
### The bloom: everyone does it, the reason is unproven [#the-bloom-everyone-does-it-the-reason-is-unproven]
Almost every pour-over recipe begins with a bloom: wet the grounds with a little water, wait a few dozen seconds, then continue. It is near-universal practice. The usual explanation is that fresh coffee holds carbon dioxide, and letting it escape first lets water reach the grounds more evenly. That sounds reasonable, but no peer-reviewed study we found shows that blooming improves extraction. One small test found that stirring during the bloom reduced how much water the grounds absorbed, yet measured no difference in the final extraction.
So bloom if you like, but do not mythologize it. If you want to know whether it matters for your beans, it makes an excellent single-variable experiment.
## Transfer [#transfer]
Read the drawdown and the taste together:
| What you see and taste | One adjustment to make first |
| -------------------------------------------- | ----------------------------------------------------------------------------------- |
| Slow drawdown, bitter and drying | Grind a little coarser; or pour more gently and stir less |
| Fast drawdown, sour and thin | Grind a little finer |
| Normal time, but sour and drying, uneven bed | Pour more evenly, and give the dripper a gentle swirl at the end to flatten the bed |
| Same recipe, very different times every brew | Fix your pouring first: same height, same number of pours |
Notice the first row offers two options. They change different things: grind changes both dissolving speed and flow, while pouring changes agitation and where the fines go. **Pick only one at a time.**
## Misconceptions [#misconceptions]
**"Pouring technique is mystical, and more elaborate is better."** Pouring changes agitation, water level and where the fines end up — all explainable. An elaborate technique is not necessarily better; a repeatable one is what matters.
**"Blooming always makes coffee taste better."** It is common practice, but its effect has not been reliably shown.
**"You can set the drawdown time directly."** Drawdown is a result. To change it, change the grind, the dose or the pour.
## Real Lab [#real-lab]
If you only have a pour-over dripper, start with a simpler version: **one recipe, two ways to pour**. Brew one cup with gentle staged pours and no stirring, and another with a thick stream and a stir after every pour. Record both drawdown times, photograph each bed afterwards, then compare the taste.
## Check Your Model [#check-your-model]
**Predict.** You switch to a narrower, deeper dripper, keeping dose and grind the same. Before looking at any result, write down: what happens to the bed? To the drawdown? Which way might the flavor move?
The same coffee piles up deeper in a narrower dripper, so water travels a longer path through more resistance. The drawdown will probably lengthen, contact time increases, extraction may rise, and the flavor moves toward bitter. If it moves noticeably, grind a little coarser to compensate.
**Diagnose.** Lately your pour-over suddenly slows down near the end, and when it finishes there is a layer of fine mud at the bottom of the bed. What might be happening, and what do you change first?
Fines are probably being washed to the bottom and clogging the bed's exit. Change only the pouring first: a thinner stream, closer to the coffee, less stirring — and see whether the final stage recovers. If it is still slow, then consider grinding a little coarser, or check whether your grinder is producing more fines than before.
**Explain.** In your own words: why can two people using exactly the same numbers brew very different pour-overs?
A recipe's numbers fix the dose, water and grind, but not how water passes through the bed. Pour height, stream thickness and stirring change agitation, where the fines go and how evenly the water flows. Those in turn change the drawdown and the evenness of extraction, so the same numbers can produce completely different cups.