#KneipenKultur Comeback
DE/EN

The Pub Quiz · Method

Where the knowledge comes from

Twenty questions on the decline of the pub, fourteen on pub lore — each with a source. This page says where every figure comes from, how scoring works and what the quiz stores about your answers. It is very little, but nobody should have to guess.

Where the questions come from

All questions live in a single file (static/data/quiz.json). The online version is built from it, so is the printable quiz night set, and so is this page. No question exists twice — otherwise the versions drift apart sooner or later.

The twenty main questions are arranged in four rounds: the state of play, the economics, the human side, the comeback. The short round takes ten of them, two or three from each round, so that the arc survives.

Alongside them sits the bar round: fourteen questions on language, beer and custom, arranged in two rounds. It can be played on its own and draws on different sources — the next section says which.

The bar round — different material, same rule

The fourteen questions of the bar round are not statistics. They are about the history of language, brewing and custom: where the German word “Kneipe” comes from, what the 1516 purity law says word for word, why village pubs in Brandenburg are called “Krug”. There is no time series to read a value off.

The project rule applies all the same: one source per claim, clickable, under the answer and in the list at the bottom. Here the evidence comes from dictionaries and institutions rather than statistics — among them the DWDS for German word history, the Online Etymology Dictionary for the English “pub”, the Historical Lexicon of Bavaria and the German Brewers' Association for the purity law, the German Legal Dictionary of the Heidelberg Academy for the Krugrecht, Destatis for the number of breweries, the LVR Institute for Regional Studies for the kiosks, the German Skat Federation for the Skat court.

  • Where something is legend, the answer says so. Much of custom is handed down and cannot be traced back to a source. Such passages are not told smoothly: the answer says that the derivation is in circulation and cannot be evidenced — and names the competing explanation where there is one.
  • What is only probable is called “probably”. In places the dictionaries are careful themselves. That caution stays in rather than being smoothed into fact.
  • Wikipedia appears three times in the source list. That is a deliberate exception: for a summary account of custom there is often no institutional source. In all three cases a firmer source carries the core — the DWDS or a hospitality trade service — and Wikipedia only the trimmings. Which question this affects and what comes from there is set out below with the sources.

So the same line applies to the bar round as to the figures: better an open uncertainty than a smooth claim. It is scored like the main questions — one point per correct answer, fourteen are possible.

How scoring works

Every correct answer gives one point — there is no further weighting. On estimation questions the order of magnitude counts: there is a tolerance band around the correct value, and a guess inside it takes the point.

  • Multiple-choice and true-or-false questions: one point for the correct answer, none otherwise.
  • Estimation questions: one point if the guess falls within the tolerance band. The bands are deliberately generous — the learning sits in the distance between guess and reality, not in failing.
  • A bullseye means the guess landed in the innermost fifth. That is a distinction, not an extra point.

The score gives the result type — passing trade, regular or part of the furniture. The boundaries sit at a third and two thirds of the points available — the same rule in the short round, the full round and the bar round, only calculated over ten, twenty and fourteen questions.

There is one deviation, and only in the printable set for a quiz night: there the main and bar questions run together in one evening, and the bar questions count half a point instead of a whole one. Otherwise word origins would decide the evening rather than the subject. Online each mode is played on its own — which is why every answer counts the same here.

What gets counted — and what does not

At the end of a round the site counts two things: for each question, which class of answer was chosen, and for the round as a whole, how many points came out. These are plain counters — numbers that go up by one.

  • No session, no cookie, no IP address. Nothing is stored that would allow two rounds by the same person to be linked.
  • No timestamps. Without them it is not possible to reconstruct after the fact who played when.
  • No individual round. What is stored is not “round 417 answered like this”, only “this answer came up n times”.

Why at all: the most interesting figure in the quiz is how far the guesses are off. As soon as enough answers have come in, that should appear in the answers — evidenced rather than asserted. Until then the figure appears nowhere.

Your progress during the game sits in your browser's local storage and does not leave the device. More on this in the privacy policy (German).

Where the figures wobble

A few answers are not a reading but a calculation from documented values — they are marked as such in the game and in the source list below. Where a source has two versions, or where two works carry one finding, that is noted there as well.

The figures that move most are the business counts and consumption per head: they are updated annually. The date of this version is at the foot of the page.

The sources

Every figure in the quiz hangs on a source; in the game it sits behind the footnote under the answer. Here they are all at once.

Where something was calculated rather than read off, where two works carry one finding, or where the source is more careful than the sentence, it is noted here and under the answer in question:

Question 1: One series, one source: VAT statistics from Destatis, the federal statistics office, category WZ 56.3 (German).

Question 2: Our own calculation from the Destatis time series (46,500 in 2002 down to around 21,000 in 2022).

Question 6: The documented figure is the margin per euro of turnover (minus 2 to plus 3 cents). The 48 cents of VAT and the conversion to a €3 beer are our own calculation.

Question 11: Listed first is the peer-reviewed 2017 version; the original study was a commissioned report for the pub organisation CAMRA.

Question 15: Two works, one source: the Chicago finding from “Heat Wave” (2002), the term from “Palaces for the People” (2018).

Question 101: The DWDS (German) calls the derivation from kneipen 'probable' — so it stands here as a conjecture, not as a fact.

Question 103: Two sources (German): the wording of the ordinance in the Historisches Lexikon Bayerns, and the explanation ('how it worked was not yet known') from the Deutscher Brauer-Bund.

Question 106: The core — a beer and a schnapps — is in the DWDS (German). The regional variants we summarise from Wikipedia; there is no institutional source for them.

Question 110: Two sources (German): the word in the DWDS dictionary, the Krugrecht in the Deutsches Rechtswörterbuch of the Heidelberg Academy. The DWDS explicitly does not cover the Krugrecht.

Question 112: A deliberate exception for Wikipedia (German article): there is no institutional source for the legend or its counter-explanation. What is documented here is mainly what is not documented.

Question 113: The core — beer felts as coasters and as covers, the name from covering — comes from HOGAPAGE (German); Wikipedia adds the manufacturing history from 1880 onwards.

Figures as of 2 August 2026.

← Back to the quiz