Discrete updates, cycles and boundaries

VectorPaint majority cycle: nine complete rankings produce a directed triangle of pairwise victories

Check pairwise vote tallies 6:3,5:4,7:2 from groups 2/3/4, distinguishing a5/3/1 control and first-choice counts.

English translation of the Chinese original, translated on . Original publication dates, attribution, code and version notes are retained.

Make a three-option pairwise-comparison diagram. Every individual ranking is complete and transitive, yet aggregated strict majorities can cycle. All nine voters and rankings are fictional mathematical inputs, not polling data or evidence for a real voting strategy. Counts are primary; triangle arrows display the comparison relation.

Declare equally weighted complete strict rankings

Options are A,B,C. Each fictional voter has weight 1 and a complete strict ranking without ties or abstention. Groups are 2 voters A>B>C,3 voters B>C>A,4 voters C>A>B, totaling 9. Display group sizes 2/3/4 rather than count three groups as three equal ballots. Create ranking-table, pair-count, relation-arrow and control layers; save native work. Individual rankings are acyclic; the object tested later is the aggregated majority relation, not an individual preference.

Compare A versus B using every group size

For A versus B, group 1 contributes 2 to A, group 2 contributes 3 to B, and group 3 contributes 4 to A. Thus A wins 6:3, so the strict-majority arrow is A→B. C’s position gives no extra ballot in this pair, and C-first voters still rank A relative to B. Every pair must total 9. Keep group contributions and verify the total before drawing. First-choice counts 2,3,4 describe another rule, not this pairwise comparison.

Check B versus C and C versus A separately

For B versus C, groups 1+2 give B 2+3=5 and group 3 gives C 4: B wins 5:4. For C versus A, group 1 gives A 2 and groups 2+3 give C 3+4=7: C wins 7:2. List each winner-first pair with option names, preventing a5:4 label from migrating to C→A. A ballot contributes once per pair, without multiplying group size by rank. Recover every result from rankings instead of choosing a cycle first and fitting counts afterward.

Use the cycle to test for a Condorcet winner

A beats B, B beats C, C beats A: a directed three-cycle. Under this lesson’s strict definition, a Condorcet winner must beat every other option by a strict pairwise majority. A loses to C, B to A, C to B, so none qualifies. This does not mean no rule can select an outcome or that individual rankings contradict themselves. First-choice counts are A 2, B 3, C 4, so plurality would choose C, which still loses pairwise to B. Keep rule-specific outputs in separate columns. Do not multiply majority margins or treat arrows as transition probabilities: each edge summarizes a different pairwise question on the same rankings, with no random state-update rule.

Draw relation arrows from winner to loser

Use a suggested 2200×1600 canvas, ranking rows at Y=200–550 and nodes A=(650,850),B=(1550,850),C=(1100,1350). Direct all arrows winner→loser, labelled 6:3,5:4,7:2 with pair names. These are relations, not time, causality or geography; spacing does not encode margins. Draw thin arrows before labels and check one incoming/one outgoing edge per node. A loser→winner convention would require a new global declaration, not a local reversal under the old legend.

Change only group sizes to obtain a winner control

Keep the same three ranking types but save group sizes 5,3,1, still 9 equal voters. A beats B(5+1):3=6:3; A beats C5:(3+1)=5:4; B beats C(5+3):1=8:1. A is now a Condorcet winner, so three options do not inevitably produce a cycle. Change only group sizes, not rankings or arrow conventions, and retain total 9 per pair. This is a mathematical control, not advice for mobilizing, manipulating or improving real election outcomes.

fictional equal-weight voters, complete strict rankings
2: A>B>C
3: B>C>A
4: C>A>B
A vs B: 6:3; B vs C: 5:4; C vs A: 7:2
majority cycle: A->B->C->A; no Condorcet winner
first choices A,B,C=(2,3,4)
control group sizes=(5,3,1)
A vs B: 6:3; A vs C: 5:4; B vs C: 8:1
control Condorcet winner: A

Diagnose counts with totals and expanded ballots

If a pair total differs from 9, check omitted/duplicated groups. If totals agree but directions differ, expand 2/3/4 groups into nine ballots and inspect relative order. Arrows all favoring plurality leader C may indicate confusion between rules; calling the cycle individually intransitive confuses aggregation with ballots. Change one step and retain versions. Ties, abstention or unequal weights require new definitions/counts. One fictional profile proves possibility, not real-world cycle frequency or a group’s enduring preference.

Reopen and reconstruct from rankings rather than memorize a triangle

Save editable groups, counts and arrows; inspect SVG/PDF and a separately opened PNG. At reduced size distinguish group sizes from margins; enlarge the 5:4 edge label. Reopen, hide the triangle, recalculate three pairs and restore it; also recalculate A’s two wins in the control. Deliver 2/3/4 profile,5/3/1 control, rule definitions and diagnosis notes, marking unperformed output steps pending. This manual diagram implies no App voting analyzer, survey dataset or automatic social-choice verdict; definitions and original inputs suffice for verification.

VectorPaint dependency diagrams: match arrows to feasible orders; VectorPaint Simpson reversal: higher in each group but lower after aggregation.

Method sources: Stanford Encyclopedia of Philosophy, Voting Methods, Condorcet’s method.

Version and operation scope: checked on 18 November 2026 against the official VectorPaint listing and Apple public lookup: version 1.0.8, requiring iOS / iPadOS 17.0 or later. Draw manually with published shapes, pen, Bézier nodes, grids/guides, alignment, layers/groups, text and file output through visible installed controls. Prepare definitions and calculations externally; infer no built-in formula, statistics or graph solver. Keep editable native work and inspect SVG, PDF or PNG as needed. SVG/PDF containing imported bitmaps do not automatically become entirely vector. Canvases and samples are original teaching settings, without physical-device execution or real measurement; mark unperformed actions pending.