The Gravity Table

Substrate series

The main table assumes a malted-barley wort. Ferments built on another sugar source sit outside the sequence, the way the lanthanides and actinides are pulled out below the real table. These entries have no number because gravity, colour and bitterness don't mean the same thing for honey or apples.

How positions are derived

Mendeleev ordered elements by atomic weight and wrapped the list whenever properties began to repeat. Beer has a repeating pattern of its own: the same character recurs at rising strength. Mild becomes bitter, then ESB, then old ale, then barleywine. Enkel becomes dubbel, tripel, quadrupel. Session IPA becomes IPA, double, triple. Strength sets the rows, and four measurable choices set everything else.

What the numbers merge

An orbital holds two styles. When a third lands on the same shell, colour and spin, it becomes an allotrope: the same element by every measured value, different in form, like diamond and graphite. The most widely brewed style names the element. That choice is cosmetic and moves nothing. These merges come from the numbers, not from my opinion.

Off the table

Checked against BJCP's tags

BJCP tags each of its styles by judgment: strength, colour, fermentation and dominant flavour. This table places styles by computation from their numbers. Where the two agree, that is independent support; where they part, each case below says why. BJCP calls its tags informal, so a disagreement is a question, not an error on either side.

Where the analogy breaks

• Nothing in beer is quantized. Style ranges overlap, and real beers wander outside them. Positions come from range midpoints, not from any particular beer.
• The shell edges and colour bins are conventions. I chose them so that well-known ladders, such as bitter to ESB or IPA to double IPA, land one shell apart. Mendeleev's periods were also found by experiment; Bohr's electron shells and Pauli's exclusion principle (1913–1925) later explained their lengths.
• Spin uses BU:GU, which ignores hop aroma, roast and acidity. For the wild block it sorts mostly by trace bitterness.
• BJCP adjusted its 2021 colour ranges for consistency with its colour names, and says this "was not the result of any scientific analysis". The colour columns inherit that.
• Rows follow gravity, not alcohol. The ABV guide beside each row assumes 75% apparent attenuation and the simple formula, which reads low for strong beers; very dry beers sit lower than their strength and sweet ones higher (see the BJCP tag check).
• sit exactly on an edge, so the tie-break convention places them (an edge value goes to the heavier shell, the darker column, or ↑); their details say so. More broadly, of would change shell, column or spin if one midpoint moved by . That is normal for any boundaries: shifted at random they leave as many or more, because sorting a continuous spread into bins always leaves some styles near a line.
• Some styles have no published ranges. Those are my estimates and are marked †. One historical style, Berliner Braunbier, has too little data and stays unplaced.

Sources

• BJCP 2021 Beer Style Guidelines (rev. 1.25): ranges for most styles, the style tags the placements are tested against, commercial examples, and the short histories, summarized in our own words.
• Brewers Association 2026 Beer Style Guidelines: ranges for styles BJCP does not define, among them Adambier, Kuit and Schoeps.
• BJCP 2025 Cider and 2026 Mead Style Guidelines: strengths and examples for the substrate series.
• Untappd: the 300 style names placed here, the top-rated lists behind some estimates, and the links on each example beer. Never a source of figures.
• Ron Pattinson's analyses of historical brewing records, Shut up about Barclay Perkins: the gravities of Mumme and Broyhan, Adambier's sourness, and notes on Dampfbier and Burton ale.
• Brewery specifications: the commercial figures behind the estimates marked †, named in each style's details.
• Jeff Alworth, The Beer Bible, 2nd ed. (Workman, 2021): the Quadrupel examples, and cross-checks on Steinbier, Sticke, Aventinus and several notes.
• The SRM colour chart computed with AJ deLange's spreadsheet (franklinbrew.org, 2007): the swatches, a display approximation, since SRM does not map exactly to a colour.

An audit script checks every guideline range, and BJCP's example lists, against the PDFs; each style's details name the source it uses. Ranges marked † are estimates: no published guideline covers the style.