🗳️ Voting Patterns Template

May 24, 2026 · View on GitHub

📌 Template Instructions: Copy to analysis/daily/{date}/{article-type}-run{N}/intelligence/voting-patterns.md. Replace every [REQUIRED] marker with analysis derived from get_voting_records, analyze_voting_patterns, analyze_coalition_dynamics, and compare_political_groups. See methodologies/per-artifact-methodologies.md §voting-patterns.

🎯 Purpose: Group-by-group coalition arithmetic for the period. Answers "how did each political group behave, where did blocs cohere, where did they fracture, and which cross-group coalitions carried majorities?" Distinct from coalition-dynamics.md (alliance-pair focused) — this file is bloc-behaviour focused.

📋 Document Metadata

Field	Value
Report ID	`[REQUIRED: VP-YYYY-MM-DD-runNN]`
Analysis Period	`[REQUIRED: e.g. Q1 2026 (2026-01-01 to 2026-03-31)]`
Roll-Call Votes Covered	`[REQUIRED: count]`
Plenary Sessions Covered	`[REQUIRED: list of part-sessions]`
Confidence	`[REQUIRED: 🟢 HIGH / 🟡 MEDIUM / 🔴 LOW — explain in body if not HIGH]`

1️⃣ Group Size & Theoretical Coalition Arithmetic

Group	Seats	% of 720	Role
EPP	`[#]`	`[%]`	`[REQUIRED: one-line strategic role]`
S&D	`[#]`	`[%]`	`[REQUIRED]`
PfE	`[#]`	`[%]`	`[REQUIRED]`
ECR	`[#]`	`[%]`	`[REQUIRED]`
Renew	`[#]`	`[%]`	`[REQUIRED]`
Greens/EFA	`[#]`	`[%]`	`[REQUIRED]`
The Left	`[#]`	`[%]`	`[REQUIRED]`
ESN	`[#]`	`[%]`	`[REQUIRED]`
NI	`[#]`	`[%]`	`[REQUIRED]`

Majority threshold: 361 votes (of 720). Simple majority of votes cast: varies by attendance.

2️⃣ Observed Coalition Patterns (Period)

%%{init: {"theme":"dark","themeVariables":{"primaryColor":"#1565C0","primaryTextColor":"#ffffff","primaryBorderColor":"#0A3F7F","lineColor":"#90CAF9","secondaryColor":"#2E7D32","secondaryTextColor":"#ffffff","tertiaryColor":"#FF9800","tertiaryTextColor":"#000000","mainBkg":"#1565C0","secondBkg":"#2E7D32","tertiaryBkg":"#FF9800","noteBkgColor":"#FFC107","noteTextColor":"#000000","errorBkgColor":"#D32F2F","fontFamily":"Inter, Helvetica, Arial, sans-serif"}}}%%
graph LR
    %% Palette note: this diagram uses political-group colours (EPP blue, S&D red, Renew orange, Greens green, ECR light-blue)
    %% which intentionally differ from the general Hack23 semantic palette defined in ai-driven-analysis-guide.md §Colour legend.
    %% Party colours take precedence over the semantic palette whenever a diagram depicts named political groups.
    EPP[EPP<br/>[#]] ---|[agreement %]| SD[S&D<br/>[#]]
    SD ---|[agreement %]| REN[Renew<br/>[#]]
    EPP ---|[agreement %]| REN
    EPP -.-|[agreement %]| ECR[ECR<br/>[#]]
    REN ---|[agreement %]| GRE[Greens/EFA<br/>[#]]

    style EPP fill:#1565C0,color:#ffffff
    style SD fill:#D32F2F,color:#ffffff
    style REN fill:#FF9800,color:#000000
    style GRE fill:#2E7D32,color:#ffffff
    style ECR fill:#0288D1,color:#ffffff

Coalition	Groups	Typical Majority	Use Cases	Cohesion (%)	Evidence
Grand Centre	EPP + S&D + Renew	`[# votes]`	`[REQUIRED: policy domains]`	`[%]`	`[REQUIRED: ≥1 RCV ID]`
Progressive-Centrist	S&D + Renew + Greens/EFA + Left	`[# votes]`	`[REQUIRED]`	`[%]`	`[REQUIRED]`
Conservative-Right	EPP + ECR + selective PfE	`[# votes]`	`[REQUIRED]`	`[%]`	`[REQUIRED]`
Opposition Blocs	PfE + ECR + ESN	`[# votes]`	`[REQUIRED]`	`[%]`	`[REQUIRED]`

3️⃣ Per-Group Behaviour

For each group: cohesion, defection highlights, sample votes.

EPP

Internal cohesion: [REQUIRED: % with evidence — e.g. 91% across 567 RCVs]
Dominant position: [REQUIRED: 1-2 sentences]
Notable defections: [REQUIRED: ≥1 named MEP + RCV ID, or note "none observed this period"]
Cross-group alliances: [REQUIRED: most frequent partners]

S&D

Internal cohesion: [REQUIRED]
Dominant position: [REQUIRED]
Notable defections: [REQUIRED]
Cross-group alliances: [REQUIRED]

Renew

Internal cohesion: [REQUIRED]
Dominant position: [REQUIRED]
Notable defections: [REQUIRED]
Cross-group alliances: [REQUIRED]

Greens/EFA

Internal cohesion: [REQUIRED]
Dominant position: [REQUIRED]
Notable defections: [REQUIRED]
Cross-group alliances: [REQUIRED]

ECR

Internal cohesion: [REQUIRED]
Dominant position: [REQUIRED]
Notable defections: [REQUIRED]
Cross-group alliances: [REQUIRED]

PfE

Internal cohesion: [REQUIRED]
Dominant position: [REQUIRED]
Notable defections: [REQUIRED]
Cross-group alliances: [REQUIRED]

The Left

Internal cohesion: [REQUIRED]
Dominant position: [REQUIRED]
Notable defections: [REQUIRED]
Cross-group alliances: [REQUIRED]

ESN / NI

Behaviour notes: [REQUIRED: aggregate position given fragmentation]

4️⃣ Bloc-Behaviour Index

Bloc	Members	Votes Won	Votes Lost	Win Rate	Trend vs. Prior Period
Grand Centre	`[EPP+S&D+Renew]`	`[#]`	`[#]`	`[%]`	`[↑ / → / ↓]`
Progressive-Centrist	`[S&D+Renew+Greens+Left]`	`[#]`	`[#]`	`[%]`	`[↑ / → / ↓]`
National-Right	`[PfE+ECR+ESN]`	`[#]`	`[#]`	`[%]`	`[↑ / → / ↓]`
EPP-led right	`[EPP+ECR±PfE]`	`[#]`	`[#]`	`[%]`	`[↑ / → / ↓]`

5️⃣ Stress Points & Outlier Votes

RCV ID	Topic	Outcome	Why outlier?	Coalition flipped?
`[REQUIRED]`	`[REQUIRED]`	`[For/Against/Abstain counts]`	`[REQUIRED: ≥30 words]`	`[yes/no]`
`[REQUIRED]`	`[REQUIRED]`	`[...]`	`[REQUIRED]`	`[...]`
`[REQUIRED]`	`[REQUIRED]`	`[...]`	`[REQUIRED]`	`[...]`

≥3 outlier votes required.

Upcoming Vote	Expected Coalition	Confidence	What would flip it
`[REQUIRED: plenary date + topic]`	`[REQUIRED]`	`[🟢/🟡/🔴]`	`[REQUIRED]`
`[REQUIRED]`	`[REQUIRED]`	`[🟢/🟡/🔴]`	`[REQUIRED]`

7️⃣ Voting Data Freshness

Required field — consumed by mcp-reliability-audit.md §"Data-source bridge" and by Stage-C completeness gate. Copy the value from VotingRecordsFallbackResult.freshnessLabel produced by getVotingRecordsWithFallback in src/mcp/ep-open-data-client.ts.

Field	Value
Data source	`[REQUIRED: "mcp" / "ep-open-data-portal" / "unavailable"]`
Freshness label	`[REQUIRED: 🟢 MCP (...→...) / 🟡 EP Open Data Portal fallback (...→...) / 🔴 voting data unavailable for window...]`
EP MCP tool	`get_voting_records` (`dateFrom` / `dateTo`)
Fallback used?	`[yes — queried EP Open Data Portal /api/v2/decision / no — MCP data sufficient / n/a — both sources empty]`
Attribution	`[REQUIRED when fallback active: "European Parliament Open Data Portal — https://data.europarl.europa.eu — CC BY 4.0"]`
Confidence adjustment	`[REQUIRED: explain whether 🟡 or 🔴 widens WEP bands or degrades Admiralty grade per osint-tradecraft-standards.md §3]`

When source = "unavailable": Do NOT substitute structural-proxy cohesion scores without an explicit 🔴 LOW-confidence flag on every claim in §§2–4. Widen all WEP bands by ~10 pp. See analysis/methodologies/osint-tradecraft-standards.md §3.1 and .github/prompts/07-mcp-reference.md §11 item #6.

8️⃣ Confidence Ledger

✅ Roll-call IDs present: [count cited in this file]
⚠️ Aggregate vs. per-MEP: [REQUIRED: note where EP roll-call delay limits claims to LOW]
🔬 Tools used: analyze_voting_patterns, analyze_coalition_dynamics, compare_political_groups, get_voting_records (per session), ep-get-voting-records (fallback, if active)

9️⃣ EP MCP Tool Inputs

EP MCP tool	Used for which section	Notes
`get_voting_records`	§1 RCV-evidence column (every claim cites RCV ID)	Aggregate margins; flag LOW if <4 weeks.
`ep-get-voting-records` (fallback)	§1 RCV-evidence column when MCP returns empty	EP Open Data Portal `/api/v2/decision`; activate via `getVotingRecordsWithFallback`; CC BY 4.0 attribution required.
`analyze_voting_patterns`	§2 Per-MEP behaviour rubric	When per-MEP feed available; aggregate otherwise.
`analyze_coalition_dynamics`	§3 Group-cohesion proxy	Two-window deltas.
`compare_political_groups`	§4 Seat-share normalisation	Confirms majority arithmetic.
`track_legislation`	§5 Procedure context	COD/CNS/APP per RCV.
`get_meeting_decisions`	§6 Adopted-decision ↔ RCV cross-check	Verifies passage.
`get_adopted_texts`	§8 Outcome confirmation	Adopted-text reflects RCV result.
`correlate_intelligence`	§9 Anomaly alerts	Defection / abstention spikes.
`detect_voting_anomalies`	§9 Outlier RCVs	Configurable sensitivity threshold.

🔟 Worked Pass-1 → Pass-2 Example (Strasbourg-I 2025 voting digest)

❌ Pass-1 (thin, 22 words):

"Several important votes happened. AI Act passed. CRA implementation was approved. Coalitions held. Some defections noted but minor."

✅ Pass-2 (compliant, 110 words, sourced):

Top 5 RCVs (Strasbourg-I 2025-10-21/24): (1) 2024/0123(COD) Critical Raw Materials Act final 412/189/68 — passage margin 178; (2) 2024/0145(COD) AI-Office implementing-act amendment 392/214/41 — Renew lost 11 of 84 (Strack-Zimmermann + 10) per analyze_voting_patterns; (3) 2025/2014(RSP) Ukraine-aid 50bn extension 487/126/52 — broadest cross-party majority of session; (4) 2024/0099(COD) CRA implementation amendment 367/241/45 — closest passage of session; (5) 2025/2901(RSP) Iran-MEK delisting 287/198/153 — high abstention signal (153 = 22 %). PfE+ESN+ECR cohesion on RCVs (1)+(4): 84 % per analyze_coalition_dynamics — right-flank consolidation continues.

1️⃣1️⃣ Worked Top-5 RCV Table

#	Procedure code + short title	Date	For	Against	Abstain	Margin	Anchor group split
1	2024/0123(COD) Critical Raw Materials Act — final	2025-10-22	412	189	68	+178	EPP 184/188 (98 %); S&D 134/136 (99 %); Renew 79/84 (94 %); ECR 41/78 (53 %); PfE 0/84; Greens/EFA 47/53 (89 %)
2	2024/0145(COD) AI-Office IA amendment	2025-10-22	392	214	41	+178	EPP 184/188 (98 %); S&D 132/136 (97 %); Renew 73/84 (87 %, 11 defectors); ECR 22/78 (28 %); PfE 0/84
3	2025/2014(RSP) Ukraine-aid 50 bn extension	2025-10-23	487	126	52	+361	EPP 188/188; S&D 136/136; Renew 84/84; ECR 56/78; PfE 18/84; ESN 0/30
4	2024/0099(COD) CRA implementation amendment	2025-10-23	367	241	45	+126	EPP 181/188; S&D 121/136 (15 dissenters); Renew 81/84; Greens/EFA 12/53 (24 % only); ECR 39/78
5	2025/2901(RSP) Iran-MEK delisting	2025-10-24	287	198	153	+89	EPP 117/188; S&D 91/136; Renew 51/84; high abstention reflects free-vote on terrorism listing

1️⃣2️⃣ Per-Group Behaviour Rubric (4 EP voting behaviours)

Behaviour	Definition	Worked observation (Strasbourg-I 2025)	Score 1-5
Discipline	% of group voting with whip on top-5 RCVs	EPP 98 %, S&D 97 %, Renew 89 %, Greens/EFA 87 %, ECR 65 %, PfE 92 %, ESN 88 %	4 (Grand Coalition); 3 (right-flank)
Defection	named MEPs voting against own group ≥1 of top-5	Strack-Zimmermann + 10 Renew on RCV 2; 15 S&D dissenters on RCV 4	2 (low — Grand Coalition); 4 (S&D fissure on CRA)
Abstention	abstention rate on contested RCVs	RCV 5 (Iran-MEK) 22 % abstention — high signal	5 on RCV 5; 1-2 elsewhere
No-show	% of MEPs absent (not voting / not authorised substitute)	Top-5 average 8 % (515-pp turnout / 720 seats); rises to 14 % on Thursday RCV 5	2 (Tue/Wed); 4 (Thu)

1️⃣3️⃣ Anti-patterns — REJECT on Pass-2 Review

#	Banned pattern	Why it fails
1	RCV cited without procedure code (e.g. "the AI Act vote")	Unindexed; reviewer cannot trace.
2	Per-MEP defection claim while EP roll-call publication delay >4 weeks	Tradecraft violation.
3	Top-5 RCV table without abstain column	Abstain is a signal, not noise.
4	Group behaviour score without RCV citation	Score is judgment-only.
5	"Coalition held" verdict on a vote where one anchor group ≤80 % cohesion	Definition mismatch.
6	Defection narrative without naming defectors (where public-record permits)	Hides the political signal.
7	Margin column missing or computed from only For-Against (ignores Abstain)	Margin = For - Against; absent column = unverifiable.

1️⃣4️⃣ Cross-References — Controlling Methodology

../methodologies/per-artifact-methodologies.md#voting-patterns — construction rules.
../methodologies/political-classification-guide.md — group taxonomy + procedure-code conventions.
../methodologies/osint-tradecraft-standards.md — Admiralty per RCV; LOW-confidence flag for <4-week-old votes.
./coalition-dynamics.md — group-pair aggregate view consumes per-RCV detail here.
./session-baseline.md — RCV count baseline.
../methodologies/electoral-domain-methodology.md — legislature voting tradecraft.

1️⃣5️⃣ Stage-C Completeness Signals

scripts/validate-analysis-completeness.js checks for this artifact:

Check	Threshold	Source
Line floor	per article-type (breaking 150, week-in-review 150, month-in-review 180, motions 200)	`reference-quality-thresholds.json`
Required H2 substrings	"Top RCVs" / "Roll-Call" / "Behaviour" / "Anomalies"	structural contract
Mermaid block	≥1 (xychart of For/Against/Abstain across top-5 preferred)	visual contract
Tradecraft markers	Admiralty per RCV; LOW flag for <4-week-old votes; abstain column populated	`osint-tradecraft-standards.md`
Source diversity	≥3 EP MCP tools (must include `get_voting_records` or `ep-get-voting-records` fallback)	per-artifact rule
Top-5 coverage	5 distinct RCVs cited with procedure codes	template logic
Voting data freshness	§7 "Voting Data Freshness" table present and `source` field populated	D-02 fallback compliance

Document Control: /analysis/daily/{date}/{type}-run{N}/intelligence/voting-patterns.md · Template v1.2 · Depth floor: per article-type minimum defined in ../methodologies/reference-quality-thresholds.json (authoritative — e.g. breaking 150, week-in-review 150, month-in-review 180, motions 200).