About
Great coffee should be knowable, not folklore.
Cup Engineering is an open lab for home coffee science: a cited wiki that reads like a story, a calculator that predicts your brew before you pour, and a community library of recipes, beans, gear and logged experiments — with an analytics dashboard where all of it adds up.
The problem it exists to fix
Coffee advice is folklore. It lives in forum threads that contradict each other, videos that can't be searched, and rules of thumb passed down without their reasons. Meanwhile the actual science — extraction kinetics, water chemistry, sensory studies — sits in journals most brewers never see. Between the two there was nowhere that connected the research to the cup in your hands: nowhere you could read why, predict what, brew it, measure it, and feed what you learned back in.
So this site is built as that missing loop. You learn from the wiki, predict with the calculator, brew from the library, log what actually happened, and the site analyzes what everyone's logs say — which sharpens the model and, chapter by chapter, the knowledge. Every page here is one stop on that loop.
The mission
Cup Engineering turns coffee science into instruments you can brew with — and every logged brew makes them sharper.
Cite every claim
Wiki articles carry numbered citations into a shared bibliography, tiered by trust — peer-reviewed and formal standards first, then institutional work, practitioner data, press, and makers last. When we last checked a reference is printed next to it.
Show uncertainty, don't bury it
The calculator gives ranges, not false precision. Articles distinguish well-replicated findings from single studies, practitioner observation, and our own editorial choices — and say which is which in the text.
Keep the methods open
The prediction model is deliberately simple, fully disclosed on the calculator page down to its coefficients, and honest about which of them are established and which are working assumptions.
Built by its brewers
Recipes, beans and brew logs come from the people who use the site, and every logged brew is a data point that will calibrate the model. Your data makes everyone's predictions better.
How the numbers work
The brew model starts from a per-method base extraction yield and nudges it for grind (surface area), temperature, contact time (with diminishing returns) and ratio; beverage mass comes from water minus what the grounds retain, and TDS follows from mass balance. The directions of those effects are well established in the literature; the coefficients are not yet fitted to data, which is why the calculator shows a “likely” range around every prediction. The full disclosure — every formula, every coefficient, and the evidence tier behind each — is on the calculator page, and the plan is to calibrate the model against the community's logged measurements as they accumulate. What has been logged so far is public on the analytics dashboard.
Who's behind it
Cup Engineering is an independent project, not a media company — built by a brewer who wanted numbers where there was folklore, and grown by everyone who logs a brew. Where an article is sponsored or a link earns a commission, it is labelled; the evidence standard doesn't bend for either.
The loop only turns if people log. Brew something tonight, write down what happened, and you've moved coffee science forward by one cup.
Log your first brew