1-2-4-All / Round-robin / Lean Coffee / World Cafe / Open Space

Status: Documented, not shipped · Evidence: P · Family: Facilitation and group structures · Verdict: reject (2026-06-03)

What it is

This entry is not one method; it is a family of large-group facilitation formats that share a single underlying move: structure how a roomful of people take turns, split into sub-groups, and rotate, so that participation scales without the meeting collapsing into the usual failure modes (the loudest voice winning, the quietest going unheard, a few dominating the air time). They are the recognized members of this genre:

• 1-2-4-All (Liberating Structures): everyone reflects alone for a minute (1), then in pairs (2), then in foursomes (4), then the whole room shares (All) - a progressive funnel that lets every person contribute before the plenary.
• Round-robin: go around the circle and take one contribution from each person in turn, so no one is skipped and no one monopolizes.
• Lean Coffee: a structured but agenda-less meeting where participants post topics, dot-vote them into an order, and timebox each discussion with a thumbs-up/thumbs-down “keep talking or move on” vote.
• World Cafe: small groups talk at cafe tables for timed rounds, then rotate to new tables while one host stays to cross-pollinate ideas between sittings.
• Open Space Technology: participants self-organize the agenda on the spot, post sessions to a marketplace wall, and move freely between them under the “Law of Two Feet.”

The durable claim of every one of these is about social architecture, not idea content. The payoff is distributing attention and turn-taking across many independent humans: pairing-before-plenary to lower the cost of speaking up, rotation to mix who hears whom, self-selection to put people where their energy is, simultaneous parallel conversation so a hundred people can all engage at once instead of one at a time. Each format emits a group experience and a wall of shared notes, not an analytical artifact a solo reasoner produces. Strip the group away and there is nothing left to run: a single person cannot pair, rotate between tables, or vote with their two feet.

It is worth separating the real move from the popular packaging. The branded names (1-2-4-All, World Cafe, Open Space) and their mnemonics and “laws” are the packaging; the durable mechanism underneath is the much older facilitation craft of timeboxed turn-taking, sub-group breakout, and rotation to make participation scale and to defeat domination by hierarchy or volume.

When it helps / when it misleads

As meeting structures for real human groups, these formats genuinely help when there are many people in the room (World Cafe and Open Space are built for dozens to hundreds), when broad buy-in or surfacing of many voices matters more than a single rigorous deliverable, and when the default failure mode is a few people dominating while the rest disengage. The pair-then-share funnel, the table rotation, and the self-selected sessions are credible guards against that. They are convergence-and-engagement rituals, not analysis engines.

They mislead or simply do not apply when:

• There is no group. The entire mechanism is the management of many independent humans’ turn-taking, pairing, and movement. A single reasoner - or a single AI agent - has no other people to pair with, no tables to rotate between, no peers to self-organize around. Run “solo,” every one of these degenerates into “think about it, then write some notes,” which is not what the format is for and not why it is famous.
• A precise, accountable output is the goal. These formats are tuned for engagement and breadth of voice. They reliably produce a rich, messy harvest of conversation; they do not, by themselves, converge on a tested decision or a defensible analysis. Treating the wall of sticky notes as a finished answer is the classic misuse.
• The harvest is never synthesized. World Cafe and Open Space generate volume; the value depends entirely on the post-event synthesis, which the format does not perform. Without it, the energy in the room evaporates and nothing carries forward.
• The structure substitutes for diverse inputs or real authority. Rotating people who all share the same frame surfaces the same idea at every table. And none of these formats decides anything - Open Space explicitly has no decision step - so a group that needs a binding call still needs a separate decision mechanism after the format ends.

What the evidence says

The honest governing grade is P (practitioner), and the grading has to be careful, because these formats are taught with great confidence and their popular write-ups borrow conviction from research that measured engagement and participant experience rather than decision quality, and that studied human groups exclusively.

What the record supports. These are real, named, widely-practiced facilitation formats with clear origins and large practitioner followings. There is a genuine (if young) empirical literature, mostly qualitative, that they do what they claim on the participation dimension. For the World Cafe, Bumble and Carter (2021), a systematic review in the Journal of Disability Policy Studies, located 35 studies covering 152 World Cafe events with more than 4,985 participants - evidence of wide, documented use - while explicitly noting that the studies varied widely in reporting clarity and rigor. Lohr, Weinhardt and Sieber (2020), in the International Journal of Qualitative Methods, make the case that the World Cafe enhances participation and data richness as a qualitative-data-collection method, and state plainly that despite abundant anecdotal claims of effectiveness “there is a dearth of empirical research underpinning the claims.” For Open Space, Van Woezik, Reuzel and Koksma (2019), in Cogent Education, ran an evaluation in medical education and found the open, self-directed format raised student motivation and supported deep-learning strategies, though students also reported insecurity at having to structure their own learning. That is the shape of the well-supported claim: these formats reliably engage people and broaden participation, and that has been observed and written up.

What the record does NOT support, and the laundering trap. None of that evidence shows these formats produce better decisions or better analysis than an ordinary meeting, and the authors of the strongest sources say so directly: Lohr and colleagues name the “dearth of empirical research,” and Bumble and Carter flag the uneven rigor of the very literature they reviewed. I can locate no controlled or comparative study that measures any of these formats as a protocol against an alternative meeting on a decision-quality or problem-solving outcome. The participation-and-engagement findings must not be transferred onto a claim of analytic superiority. Equally, no traceable primary source supports the popular “X times more productive” or “best ideas happen at the coffee break, so design the whole event like one” intuitions as measured effects; the coffee-break story is Harrison Owen’s founding anecdote for Open Space, not an experiment, and should be read as origin lore, not data. There are no nameable effect sizes to quote here, and inventing one would be laundering; the honest statement is that the evidence is qualitative, supportive on participation, silent on decision quality, and thin overall - which is exactly a P.

Transfer caveat (required). Every study above is on human groups in workshops, classrooms, and community settings (medical students, disability-community members, qualitative-research participants). None studies any of these formats performed by or with an AI agent, and the whole value of the genre rests on social dynamics among many independent humans - turn-taking, pairing, rotation, self-selection. The evidence is transferred from human group contexts and is not validated for, and largely not even applicable to, a single AI agent.

Why it is / is not a skill here

Verdict: Reject (status excl). These are excluded on the merits as out of scope for a library of solo-agent thinking artifacts. The registry states the governing principle outright: this is a workshop reference, and “the value is human social dynamics an AI cannot reproduce.”

The Build burden for any entry is to name one distinct, durable cognitive move that a solo agent performs and that no shipped skill already produces, plus the separable artifact it emits. Every format in this family fails that burden on both the artifact test and the agent-applicability test, and the failure is the same in each case:

• The artifact is a group experience, not a deliverable. 1-2-4-All produces a funneled plenary, World Cafe a cross-pollinated set of table harvests, Open Space a self-organized session marketplace. The “output” is what happened to and among the people; there is no analytical artifact a lone reasoner emits. A library whose product is a solo thinking artifact has nothing to capture.
• The mechanism is irreducibly social. Pairing-before-plenary, rotating between tables, voting with your two feet, taking one turn each around a circle - none of these has any meaning for a single agent. They are devices for governing many independent humans, which is precisely the part an AI cannot reproduce. This is the same reason the family’s sibling note-and-vote is excluded and dot-voting is flagged: a famous, genuinely effective group-facilitation protocol is not automatically an agent skill.
• Whatever solo residue exists is already shipped. The one fragment that does survive solo is the silent-generation-before-discussion moment inside 1-2-4-All and round-robin - “write your own candidates before anyone talks.” That is the Nominal Group Technique / brainwriting mechanism, and the library already ships it as brainwriting (“Brainwriting 6-3-5 / NGT”), into which silent-writing-before-discussion is a recorded fold. So even the capturable sliver adds no new move; it is brainwriting, and the rest is the human-dynamics part the library does not ship.

Why reject (excl) rather than fold: folding into brainwriting would falsely assert that brainwriting captures these formats, when the social architecture (rotation, self-selection, two-feet movement, scaled simultaneous parallel conversation) is the whole point and is exactly the non-reproducible part - a fold would hide the exclusion instead of recording it. And there is no single clean fold target: the silent-generation sliver lands on brainwriting, the dot-vote inside Lean Coffee lands on the flagged dot-voting, and the irreducible remainder is the excluded social-dynamics part. Naming one foldInto would misrepresent the entry. Reject is the more honest disposition, and it mirrors note-and-vote: these are real, useful meeting structures that are out of scope for a solo-agent thinking library. The learning value of the NO is the durable principle for the whole facilitation-and-group-structures family - the library documents these famous formats, points their capturable sliver at brainwriting, and excludes them rather than ship a hollowed-out ritual under a famous name.

Lineage and who to read

The five formats have distinct, well-documented origins.

• 1-2-4-All is one of the 33+ Liberating Structures developed by Henri Lipmanowicz and Keith McCandless, partners since 2002, drawing on complexity science from their time on the Plexus Institute advisory board; the canonical reference is their book The Surprising Power of Liberating Structures (2014) and the open liberatingstructures.com menu. The structures are published under a Creative Commons license; “Liberating Structures” is the project’s name, so the entry credits Lipmanowicz and McCandless descriptively.
• World Cafe was discovered by Juanita Brown and David Isaacs in 1995, when rain forced a planned outdoor strategic dialogue at their Mill Valley home indoors onto kitchen tables with paper tablecloths; the rotating-table structure that emerged was formalized over the following decade in The World Cafe: Shaping Our Futures Through Conversations That Matter (2005). “The World Cafe” is associated with The World Cafe Community Foundation and is credited to Brown and Isaacs.
• Open Space Technology was created by Harrison Owen in 1985, after he noticed the best conversations at a conference he had organized happened during the coffee breaks and asked why the whole event could not be run that way; its doctrine is the Four Principles and the one Law of Two Feet, set out in Owen’s Open Space Technology: A User’s Guide. Credited to Harrison Owen.
• Lean Coffee was started in Seattle in 2009 by Jim Benson and Jeremy Lightsmith as the least structure necessary for a coherent agenda-less meeting (post topics, dot-vote the order, timebox with a keep-going vote); documented at leancoffee.org and in Benson’s lean/Kanban work.
• Round-robin is a generic, un-branded turn-taking convention with no single author; it predates and underlies the others.

For the honest evidentiary picture, read Lohr, Weinhardt and Sieber (2020) on the World Cafe as a qualitative method and their candid note on the thin empirical base; Bumble and Carter (2021) for the scale of documented use and the uneven rigor across it; and Van Woezik, Reuzel and Koksma (2019) for a rare structured evaluation of Open Space, supportive on engagement, silent on decision quality.

Named sources

• Henri Lipmanowicz and Keith McCandless, The Surprising Power of Liberating Structures: Simple Rules to Unleash A Culture of Innovation (Liberating Structures Press, 2014); liberatingstructures.com, “1-2-4-All.” The canonical articulation of the pair-to-plenary funnel and the broader Liberating Structures repertoire. Practitioner / foundational. (P)
• Juanita Brown and David Isaacs, The World Cafe: Shaping Our Futures Through Conversations That Matter (Berrett-Koehler, 2005). The origin and formalization of the rotating-table dialogue format. Practitioner / foundational. (P)
• Harrison Owen, Open Space Technology: A User’s Guide (Berrett-Koehler; orig. 1992). The Four Principles, the Law of Two Feet, and the self-organizing-agenda format; includes the coffee-break origin story. Practitioner / foundational. (P)
• Jim Benson and Jeremy Lightsmith, Lean Coffee (Seattle, 2009); leancoffee.org. The agenda-less, dot-voted, timeboxed meeting format. Practitioner / foundational. (P)
• Katharina Lohr, Michael Weinhardt and Stefan Sieber, “The ‘World Cafe’ as a Participatory Method for Collecting Qualitative Data,” International Journal of Qualitative Methods 19 (2020). Argues the format enhances participation and data richness, and states plainly there is a dearth of empirical research underpinning effectiveness claims. The honest evidentiary anchor. (P, qualitative)
• Jennifer L. Bumble and Erik W. Carter, “Application of the World Cafe to Disability Issues: A Systematic Review,” Journal of Disability Policy Studies 32(3) (2021): 193-203. Located 35 studies of 152 World Cafe events (4,985+ participants); documents wide use and flags wide variation in reporting clarity and rigor. (P, review)
• Tamara van Woezik, Rob Reuzel and Jur Koksma, “Exploring Open Space: A self-directed learning approach for higher education,” Cogent Education 6 (2019), article 1615766. Structured evaluation in medical education: Open Space raised motivation and supported deep-learning strategies; students reported insecurity at self-structuring. Supportive on engagement, silent on decision quality, single institution. (P, study)

Excluded under the evidence rule: the popular “X times more productive” claims and the “best ideas happen at the coffee break” intuition (Open Space origin lore, not an experiment) have no traceable primary source measuring decision quality and are not counted toward this entry’s grade.