Dot voting

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

What it is

Dot voting (also “dotmocracy” or “voting with dots”) is a group-facilitation move for converging a roomful of people from many options to a short list of favorites, fast and visibly. The options are laid out where everyone can see them - sticky notes on a wall, lines on a flip chart, cards on a board - and each participant is handed a fixed, limited allowance of marks (sticker dots, or strokes of a pen) and told to place them on the options they most prefer. The rule that does the work is the budget: because each person has only, say, three or five dots to spend across a long list, they cannot endorse everything, so a visible heat map of collective preference emerges as the dots accumulate. The artifact is that heat map - a ranked-by-density picture of where the group’s enthusiasm actually sits - which the facilitator then uses to cut the list down or pick a winner.

It helps to separate the durable mechanism from the popular packaging, because they are not the same size. The durable underlying move is scarce-budget preference aggregation: forcing each voter to allocate a limited stock of approval across competing options so that lukewarm-but-broadly-liked items can surface and nobody’s single passion dominates. That is the same logic the quality-engineering tradition calls multivoting and the social-choice tradition calls cumulative voting - dot voting is the low-ceremony, sticker-on-a-wall instance of it. The popular packaging adds the physical ritual: real dots, a shared wall, everyone voting in the same room at the same time. The packaging is where dot voting’s signature problem lives (see below), and it is also the part that is inseparable from having an actual human group, which is decisive for the verdict here.

When it helps / when it misleads

As a meeting structure for a real human group, dot voting helps when the options are already on the table (it is a converge move, not a generate-from-blank move), the list is too long to debate item by item, and the goal is a quick, legible read of where collective preference sits without a drawn-out argument. The limited-dot budget is a genuine improvement over a show of hands or “vote for your favorite,” because it lets an option that many people rate second-best - but nobody rates first - rise above a polarizing option that a few people love and many dislike. That is the property the multivoting literature praises, and it is real.

It misleads or simply does not apply when:

• There is no group. The entire payoff is the aggregation of many independent people’s private preferences into a visible collective picture. A single reasoner, or a single AI agent, has no other voters to pool, no spread of private preference to surface, and no wall of dots to read. Run “solo,” dot voting collapses into “list some options, then pick the ones I like,” which is not what the method is for and adds nothing a plain option review does not.
• The dots are open and sequential - the herding trap. This is dot voting’s defining weakness and the reason this entry is flagged. Because votes are placed in public, on a shared surface, in front of everyone, later voters can see where the dots are already piling up - and a well-documented bandwagon tendency pulls them toward the leader rather than toward their own honest preference. Early, senior, or simply faster voters anchor the result; the heat map can measure visibility and momentum as much as merit. The basic ritual does not fix this on its own.
• The vote is mistaken for the decision. A dot count is a fast preference snapshot, not a considered judgment. Treating the densest cluster as the answer skips the “so what / is the popular option actually the right one” step, and a group can dot its way confidently to a bad choice.
• The options were never made comparable. Dot voting presumes the items are roughly commensurable and already understood. Pointed at a list where people are voting on different interpretations of the same label, or on options of wildly different scope, the dots aggregate noise.

The mitigations that the facilitation literature recommends all amount to removing the herding channel: vote silently and simultaneously, conceal votes until everyone has placed them (closed or “blind” dotting), have junior people vote before senior people, or use a private tally rather than an open wall. A dot vote run without one of those guards should be read with the herding caveat attached.

What the evidence says

The honest grade for dot voting’s own move - “hand a group a limited dot budget and read the resulting heat map” - is P (practitioner). It is a real, named, widely taught facilitation technique, but the controlled evidence that exists sits next to it, not on it, and the strongest research finding about it is a warning, not an endorsement.

What the record supports. Dot voting is a recognized practitioner method with decades of documented use in workshops, retrospectives, and design sprints, and a clear popularizing lineage (Diceman’s Dotmocracy Handbook, the quality-engineering multivoting tradition, the design-sprint adoption of dot votes for convergence). As a limited-budget aggregation rule it is a legitimate, lightweight instance of cumulative voting / multivoting, whose claimed advantage - letting a broadly-liked option beat a narrowly-loved one - follows from the allocation rule itself. That is the extent of the directly supported claim: a respectable practitioner heuristic for fast group convergence.

What the record does NOT support, and the laundering trap. There is no controlled or comparative study I can locate that measures dot voting as a protocol - “does placing limited dots on a wall produce better group decisions than a show of hands or open discussion?” - against any alternative. The genuinely experimental evidence about multivoting points one step away from dot voting’s signature ritual: Johnson, Awtrey and Ong (2022), in Academy of Management Discoveries, found across 93 student “pursuit teams” that groups taking an unofficial multivote (ten votes spread across three options) were about 50% more likely to identify the correct option than groups using plurality or ranked-choice voting - but the authors report the gain appeared before any discussion, attributing it to multi-allocation making individuals process the evidence more deeply on their own. That endorses multi-voting over single-choice voting (which is in dot voting’s favor as an allocation rule), yet it locates the benefit in a solo deep-processing effect and, crucially, studied a concealed allocation, not an open wall of stickers. It does not validate the public-dotting ritual, and transferring its “~50%” onto open dot voting would be laundering a cousin’s robustness onto the very feature that cousin avoided.

Where the evidence actually cuts against the ritual. The better-evidenced literature about dot voting’s public, sequential form is the literature on social influence in sequential voting, and it is a caution. Bikhchandani, Hirshleifer and Welch (1992) formalized information cascades: when people act in sequence and can see predecessors’ choices, it can become individually rational to ignore one’s own private information and copy the crowd, so a visible early lead snowballs regardless of merit. On the field side, Morton, Muller, Page and Torgler (2015), in the European Economic Review, used a French voting reform as a natural experiment and found that exposure to early-result (exit-poll) information significantly increased bandwagon voting - turnout-deciders shifted toward the expected winner - the first clean non-laboratory demonstration of the effect. Neither study is about dot voting, but both describe exactly the mechanism the open-wall ritual creates: later voters seeing where the dots already cluster and drifting toward the leader. This is why the registry flags the method rather than excluding it outright - the move is useful, but it ships with a known bias that must be named.

Transfer caveat (required). Every study above is on human groups and electorates in lab or field settings (student teams, voters). None studies dot voting, multivoting, or bandwagon effects performed by or with an AI agent, and the method’s entire value rests on social dynamics among many independent humans. The evidence is transferred from human group and election contexts and is not validated - and largely not even applicable - to a single AI agent.

Excluded figures (required). No traceable, single-authored effect size exists for the dot-voting ritual itself, so none is asserted here; the only sourced quantity is the Johnson et al. (2022) multivote result, which measures a concealed multivote and not open dotting, and is reported with that limit. Any popular “dot voting improves decisions by N%” framing has no nameable primary source and is excluded under the evidence rule.

Why it is / is not a skill here

Verdict: Flag (status flag, governing tier P). Dot voting is documented in the library but not shipped as a standalone skill, and the registry’s instruction is explicit: include it only with the herding caveat. The reasoning has two parts.

It is not a separable solo move. The Build burden for this library is to name one distinct, durable cognitive move that a solo agent performs and that no shipped skill already produces. Dot voting fails that test the same way the sibling group rituals do. Its capturable logic - take a set of existing options, weigh them against preference, and converge on the strongest - is already produced by the shipped converge skills (the option-review / weighted-comparison family the catalog ships for exactly this); the limited-allocation refinement is a parameter on that move, not a new move. Strip the human group away and “dot voting” is just “score these options against what matters and keep the top ones,” which the library already does. The note-and-vote entry makes the same call from the other direction: when it decomposed that design-sprint protocol, it pointed the converge half at the shipped option-review skill and named dot-voting here as the excluded, herding-flagged aggregation step. The two entries agree: the aggregation ritual is not a standalone agent skill.

Its distinctive part is a human-dynamics ritual an agent cannot reproduce - and it carries a known bias. What dot voting adds over a plain option review is precisely the physical, social act of many people placing visible dots on a shared wall. That is irreducibly a property of a human group, the same human-social-dynamics ground on which the family excludes the scaled-participation formats (“the value is human social dynamics an AI cannot reproduce”). And unlike a neutral ritual, this one comes with a documented failure mode - the open-wall herding / bandwagon effect above - which is why the disposition is flag rather than a silent fold: the library’s service to a reader is to record both that the move is real and that, where a human team does run it, it must be run with the de-herding guards (silent, simultaneous, or concealed voting; junior-before-senior order). A flag preserves that caveat; a quiet fold into the option-review skill would hide it.

Why flag rather than excl or fold. A bare excl (the disposition given to the broader scaled-participation formats) would under-serve the reader, because dot voting is a specific, famous, genuinely useful allocation rule worth locating and worth caveating, not just a class of workshop formats to wave off. A fold would over-claim that a shipped skill captures dot voting, when the captured part is only the commensurable converge logic and the uncaptured part - the social ritual plus its herding bias - is exactly what makes dot voting “dot voting.” Flag is the honest middle: documented, located against the shipped converge skills, and gated by the herding caveat. The learning value of the decision: a famous, genuinely useful group-voting ritual is not automatically an agent skill, and when its signature mechanism carries a documented bias, the library’s job is to name the bias, not to repackage the ritual under a tidy fold.

Lineage and who to read

Dot voting’s exact origins are not cleanly attributable - voting with marks on a shared list is old workshop folklore - but the modern named lineage runs through three streams. The quality-engineering tradition formalized the limited-allocation version as multivoting (also “nominal prioritization”), taught as part of the Nominal Group Technique toolkit by the American Society for Quality (ASQ) and the Six Sigma community; this is where the “give everyone a fixed number of votes so a broadly-liked option can win” rule is most rigorously described. The participatory-democracy stream produced the “dotmocracy” name and its most-cited handbook: Jason Diceman introduced Dotmocracy sheets at the Karma Food Co-op in Toronto around 2004 and published the free Dotmocracy Handbook (Version 2.2, 2010), which formalized dot stickers on an agreement scale for large-group decisions. The design / facilitation stream (Google Ventures design sprints, agile retrospectives, design-thinking workshops) adopted dot voting as a standard convergence step and is where most practitioners meet it today.

For the evidence about its public, sequential form - the part this entry flags - read the social-influence literature, which is where the herding caveat is actually grounded: Bikhchandani, Hirshleifer and Welch (1992) on information cascades, and Morton, Muller, Page and Torgler (2015) on bandwagon voting under early-result information. For the allocation rule’s one supportive controlled result, read Johnson, Awtrey and Ong (2022) on multivoting, noting that it studied a concealed multivote and located the gain in solo processing. “Dot voting,” “dotmocracy,” and “multivoting” are descriptive, generic terms in common use - no registered trademark to clear - so this entry is documented descriptively and flagged for its herding caveat, not for branding.

Named sources

• Jason Diceman, Dotmocracy Handbook: A Simple Tool to Help Large Groups Find Agreement, Version 2.2 (self-published, 2010; freely available). The canonical articulation of dotmocracy: participants place dot stickers on an agreement scale to surface large-group collective preference. Practitioner / foundational. (P)
• American Society for Quality (ASQ), “Multivoting” and “Nominal Group Technique” (quality-resources). The quality-engineering formalization of limited-allocation voting: each member gets a fixed number of votes so an option favored broadly but first by none can rise to the top. Practitioner / methodological reference. (P)
• “Dot-voting,” Wikipedia. Documents dot voting as a form of cumulative voting and notes that its origins are unclear, with the Karma Food Co-op / Diceman dotmocracy line as the best-attested modern thread. Tertiary / encyclopedic. (P)
• Sushil Bikhchandani, David Hirshleifer and Ivo Welch, “A Theory of Fads, Fashion, Custom, and Cultural Change as Informational Cascades,” Journal of Political Economy 100(5) (1992): 992-1026. Foundational model of information cascades: when people act in sequence and observe predecessors, it can be rational to ignore private information and copy the crowd. Explains why open, sequential dot voting herds; does not test dot voting. (M, for the cascade mechanism - not for dot voting)
• Rebecca B. Morton, Daniel Muller, Lionel Page and Benno Torgler, “Exit Polls, Turnout, and Bandwagon Voting: Evidence from a Natural Experiment,” European Economic Review 77 (2015): 65-81. Natural experiment on a French voting reform: early-result information increased bandwagon voting toward the expected winner (and reduced turnout). The first clean field demonstration of the bandwagon effect; grounds the herding caveat, on electorates not workshops. (M, for bandwagon voting in elections - not for dot voting)
• Michael D. Johnson, Eli Awtrey and Wei Jee Ong, “Verdicts, Elections, and Counterterrorism: When Groups Take Unofficial Votes,” Academy of Management Discoveries (2022). Experimental: across 93 student teams, multivoting groups were about 50% more likely to pick the correct option than plurality or ranked-choice groups, with the benefit appearing before discussion. Supports limited-allocation multivoting over single-choice voting, but studied a concealed multivote and located the gain in solo processing - it does not validate the open-wall dotting ritual. (M, for the multivote allocation rule on human groups - not for open dot voting)

Excluded under the evidence rule: no single-authored, traceable effect size exists for the open dot-voting ritual itself; any popular “dot voting improves decisions by N%” framing has no nameable primary source and is not counted toward this entry’s grade. The one sourced quantity (Johnson et al. 2022) measures a concealed multivote, not open dotting, and is reported with that limit.