How Press Timing Changed in World AeroPress Championship Recipes

Across the 46 available recipes, press start is documented in 30, duration in 33, and total time in 28. That leaves 16, 13, and 18 missing values respectively. A page that reports one n for “timing” would hide which comparison is actually supported.

The overall medians are 85 seconds for press start, 30 seconds for duration, and 107.5 seconds for total time. They come from different subsets. They are not three measurements of one shared cohort, and adding or subtracting the medians would create a synthetic recipe that may never have existed.

Brew Supply's operating rule is to name the field before reading the number. “Press at 1:30” is a start. “Press for 30 seconds” is a duration. “Finish at 2:00” can support a total time, but only if the source makes the timer origin and finish clear. A word such as slowly is useful texture, not a numeric replacement.

The available starts span 15–180 seconds around a 62.5-second median. Durations span 20–75 seconds around a 30-second median, while total times span 70–210 seconds around a 102.5-second median. The extreme values are real normalized records, but the missingness makes the range more informative than a smooth trend line.

Only 6 early recipes support all three numeric fields, and 3 available recipes support none of them. The archive can still describe a press qualitatively when no normalized number appears. That prose may be sufficient to understand the method yet insufficient for a period statistic.

Early recipes also differ in orientation, water volume, recirculation, steep structure, and what the written method treats as completion. The broad timing envelope therefore describes heterogeneous podium methods and historical documentation. It does not show that competitors deliberately optimized across a 15-to-180-second start range.

The median press start is 65 seconds, duration is 30 seconds, and total time is 95 seconds. However, total time is available in only 7 of 12 recipes, so even this apparently compact period does not support a complete twelve-recipe timing distribution.

The timing sits inside the period's concentrate-and-bypass architecture. Many methods combined large coffee doses, small brew-water stages, vigorous agitation, a short clock, and substantial dilution. A 30-second press in that system is not operationally interchangeable with a 30-second press in a fuller-volume recipe using another grind, filter, and endpoint.

Five records support a complete start-duration-total triple. All available recipes contain at least one normalized timing value, but that improvement over the early archive is not proof that every press was fully specified. Numeric timing can coexist with unstated pressure, flow changes, hiss handling, pressed output, or plunger position.

The modern start median is 90 seconds with a 50–115-second range. Duration has a 30-second median and 15–60-second range. Total time has a 120-second median and 70–135-second range. Eleven recipes support all three fields.

That concentration is a useful chronological finding, but it is not one repeated recipe. The period includes different pours, temperature stages, agitation, filters, bypass amounts, press outputs, cooling routines, and service decisions. The 90-and-30 pattern locates a common clock structure; it does not make the surrounding systems equivalent.

One available modern recipe has no normalized numeric timing triple, and four lack a total-time value. The right conclusion is that modern documentation is denser and the usable values are more concentrated—not that every podium recipe pressed at 1:30 for exactly 30 seconds or that doing so improved placement.

This identity is a useful quality check. It confirms that the three normalized fields are internally coherent for the complete records and makes the year-by-year matrix easier to audit. It does not turn any one field into redundant data: start and duration still describe different actions, while total time marks their combined endpoint on the recipe clock.

The identity is not used to fill blanks. If a source supports start and duration but the normalized total is absent, this page keeps total time absent. Likewise, a duration is not inferred by subtracting start from an ambiguous finish time. Derivation can look mathematically harmless while silently erasing uncertainty about timer origin, pauses, or what the source counted as the end.

This is the distinction between validating a complete record and reconstructing an incomplete one. Brew Supply uses the equation to detect contradictions among values that are already supported. It does not use it to convert partial documentation into a denser dataset than the archive warrants.

Historical recipe wording varies from precise clock instructions to phrases such as press slowly, press gently, press all the way, or stop at a physical or audible cue. The normalized fields preserve numeric timing when the source supports it, but this page does not translate a qualitative rate into seconds or score one endpoint as more advanced than another.

Endpoint differences can change output and workflow even when start and duration match. One brewer may stop before the final air is expressed; another may continue until the plunger reaches a position; another may target grams in the server. Time alone therefore cannot establish equal pressure, equal flow, equal retention, or equal final beverage mass.

The complete historical report and official recipe record are the right places to inspect exact wording. The timeline here answers the chronological timing question, while the source method restores the sequence, endpoint, output, bypass, cooling, and service context needed for a faithful rebrew.

In the separate 25-of-66 finalist sample, press starts span 50–160 seconds around a 92.5-second median. Exact durations span 15–60 seconds around a 30-second median. Only first, second, and third placement are known.

The historical dataset contains podium recipes across many editions, coffees, rules, and documentation conventions. The finalist dataset examines one same-coffee season with a much wider rank-unknown sample. Pooling them would duplicate the 2025 podium, overweight one year, and blur two eligibility rules.

The similar modern medians are a comparison, not a merger and not an optimum. Neither dataset is a controlled timing experiment. Each recipe changes grind, water, temperature, agitation, filtration, pressing, bypass, cooling, and service together, so prevalence cannot be translated into a causal placement advantage.

If you want to rebrew a podium recipe, open its complete archive record. Copy the timer origin, pour and agitation sequence, press start, duration, endpoint wording, output, bypass, cooling, and service steps before changing anything. Record substitutions instead of treating the three timing numbers as the full method.

If you want to design a home schedule, use the evergreen press-time guide. Hold coffee, dose, grind, water, temperature, agitation, filter, orientation, bypass, and final target stable. Brew the same timing schedule twice before changing start or duration, and record resistance and grams out alongside the clock.

Brew Supply's judgment is that the historical medians are best used as navigation, not prescriptions. A 90-second start and 30-second duration point toward a group of modern methods worth reading. They do not tell you which schedule suits your coffee, your grind, your filter resistance, or your desired cup.