# There was no average winning recipe: what 23 WAC 2025 recipes reveal

Canonical: https://brew.supply/wac-2025-recipe-analysis-report

We analyzed 23 recipes from the 2025 World AeroPress Championship final. The rules kept dose and served beverage mass relatively tight, but competitors made very different choices around brew water, bypass, temperature, timing, agitation, pressing, and service.
The sample covers 23 of 66 finalists. Only first, second, and third place are known; the other 20 records are rank unknown. All competitors used the same coffee and could use no more than 18 g.

## Five findings

- Water architecture showed the largest split. Brew-water ratios ranged from 5.56:1 to 13.33:1.
- Bypass was common, but it was not one method. Exact amounts ranged from 0 g to 88 g, and timing, cooling, and service role also differed.
- The podium used three different systems. First and second used 100 g of brew water with large bypasses; third used 208 g of brew water and only 12 g of bypass.
- Temperature and timing did not converge. Temperatures ranged from 82–95°C, press starts from 50–160 seconds, and exact press durations from 15–60 seconds.
- Written recipes described pressing poorly. Most records included some timing, but few defined press endpoint, output, or flow rate.

## Where the recipes landed

| Variable | Documented range or pattern |
| --- | --- |
| Dose | 16–18 g |
| Final beverage | Mostly 150 g because of the judging minimum |
| Brew-water ratio | 5.56:1–13.33:1 |
| Bypass | 0–88 g |
| Temperature | 82–95°C |
| Press start | 50–160 s |
| Press duration | 15–60 s |
| Stir count | 0–25 |
| Swirl count | 0–28 |

Served mass tells us little about recipe strategy because the rules pulled competitors toward the same minimum. Brew water, bypass, temperature, timing, agitation, pressing, and service show much more of the actual design choice.

## Variable atlas

The main report keeps the full distributions because they are a central research outcome. Each graph states its own eligible n, keeps sensitivity values visually separate, and marks podium recipes. Stirring and swirling remain separate, and grinder clicks remain inside compatible systems.
- [Coffee dose](https://brew.supply/wac-2025-recipe-analysis-report#coffee-dose)
- [Brew water](https://brew.supply/wac-2025-recipe-analysis-report#brew-water)
- [Brew-water ratio](https://brew.supply/wac-2025-recipe-analysis-report#brew-water-ratio)
- [Bypass amount](https://brew.supply/wac-2025-recipe-analysis-report#bypass) and [bypass share](https://brew.supply/wac-2025-recipe-analysis-report#bypass-share)
- [Final beverage mass](https://brew.supply/wac-2025-recipe-analysis-report#final-beverage) and [final-beverage ratio](https://brew.supply/wac-2025-recipe-analysis-report#final-beverage-ratio)
- [Brew temperature](https://brew.supply/wac-2025-recipe-analysis-report#temperature) and [water TDS](https://brew.supply/wac-2025-recipe-analysis-report#water-tds)
- [Press start](https://brew.supply/wac-2025-recipe-analysis-report#press-start) and [press duration](https://brew.supply/wac-2025-recipe-analysis-report#press-duration)
- [Stirring](https://brew.supply/wac-2025-recipe-analysis-report#stirring), [swirling](https://brew.supply/wac-2025-recipe-analysis-report#swirling), and [cooling](https://brew.supply/wac-2025-recipe-analysis-report#cooling)
- [Compatible grinder groups](https://brew.supply/wac-2025-recipe-analysis-report#grind)

## Water architecture

Brew water is water documented as contacting the coffee. Bypass is water added outside that phase. Competitors used strong concentrate-plus-bypass systems, moderate concentrates, and fuller-volume or low-bypass systems. Bypass amount and timing both mattered: water could be present before brewing or added after pressing for dilution, cooling, or service.
No single broad water structure dominated the available sample. Brew-water ratio describes recipe inputs, not extraction yield.

## Three podium systems

| Variable | Némo Pop | Jan Ahrend | Dharun Vyas |
| --- | ---: | ---: | ---: |
| Dose | 18 g | 18 g | 16 g |
| Brew water | 100 g | 100 g | 208 g |
| Brew-water ratio | 5.56:1 | 5.56:1 | 13:1 |
| Bypass | 70 g | 86 g | 12 g |
| Temperature | 84°C | 88°C | 88°C |
| Press duration | 20 s | ≈44 s, derived | 60 s |

First and second shared a concentrate structure but differed in bypass timing, orientation, filters, preparation, cooling, and pressing. Third used a fundamentally different water structure. Némo Pop separated from the other 22 in the strongest statistical split, but unusualness does not explain the win.

## Three additional patterns

Temperature and timing remained broad. Cooling was documented in all seven recipes in one post-brew neighborhood and one of eight in the other large neighborhood. Written recipes documented press timing more often than press endpoint, output, or flow rate.
The strongest statistical split separated Némo Pop’s recipe from the other 22. The report uses clustering to find nearby recipes, not to define fixed recipe types.

## What competitors can test

Choose a broad water structure, hold final beverage mass constant, and change one main variable at a time. Record stirring and swirling separately. Define press start, duration, rate, output, endpoint, and whether the press reaches the hiss. Include cooling and service in the recipe.
Explore the data contains the complete recipe records, names behind each point, detailed filters, evidence status, sample-edge distinctions, and downloads. Methods and provenance contains the full eligibility audit, normalization rules, grinder conversions, clustering details, and workbook coordinates.

## Downloads

The public appendix includes approved normalized recipe fields, recipe architecture, podium profiles, numeric badges, and the data dictionary. It excludes handles, raw source wording, recovery notes, legacy scores, and private attribution material.
- [Normalized public recipes](/data/wac-2025-recipe-analysis-report/normalized-recipes-public.csv)
- [Recipe architecture](/data/wac-2025-recipe-analysis-report/recipe-architecture.csv)
- [Podium profiles](/data/wac-2025-recipe-analysis-report/podium-profiles.csv)
- [Numeric badges](/data/wac-2025-recipe-analysis-report/numeric-badges.csv)
- [Data dictionary](/data/wac-2025-recipe-analysis-report/data-dictionary.csv)
- [Variable distributions](/data/wac-2025-recipe-analysis-report/variable-distributions.csv)
- [Variable summaries](/data/wac-2025-recipe-analysis-report/variable-summaries.csv)

## Limitations

This is a convenience sample, not a census. The missing 43 recipes may contain patterns that are not visible here. Usable counts vary by variable, and written gaps may reflect incomplete records rather than what happened on stage.
Most final-beverage values sit at 150 g because that was the minimum beverage mass required for judging. This provides service context rather than evidence of a preferred method. Water TDS is not beverage TDS, so the report does not calculate extraction yield. No chart identifies a cause of podium placement.

## Sources

- [Canonical WAC 2025 research workbook](https://docs.google.com/spreadsheets/d/1Mp_TLg-i2lH9T7MwB-2z1JbJGQwXflVXWpAx341H2EM/edit)
- [Official first-place recipe: Némo Pop](https://worldaeropresschampionship.com/pages/1st-nemo-pop-australia-2025)
- [Official second-place recipe: Jan Ahrend](https://worldaeropresschampionship.com/pages/2nd-jan-ahrend-switzerland-2025)
- [Official third-place recipe: Dharun Vyas](https://worldaeropresschampionship.com/pages/3rd-dharun-vyas-india-2025)