Elections Vote Transfer Project

Interactive visualizations for analyzing Israeli Knesset election data, showing vote transfers between elections, party support patterns, and ballot-level analysis.

Live site: https://kolot-nodedim.netlify.app/

Project Structure

├── ballot*.csv                    # Raw election data per election (16-26), in project root
├── data/                          # Processed data files
│   ├── ballot_locations_*.json    # Ballot venue names (scraped)
│   ├── socio_eco21_T*.xlsx        # CBS 2021 socioeconomic data (T01, T07, T08, T12)
│   └── statisticalareas_demography2019.gdb/  # CBS 2011 statistical area boundaries (GDB)
├── party_config.py                # Election metadata, party names/symbols/seats per election (16-25)
├── generate_transfer_data.py      # Generates vote transfer matrices (cvxpy optimization)
├── generate_tsne_data.py          # Generates T-SNE clustering data
├── generate_map_data.py           # Generates geographic map data per election (with name normalization)
├── download_statistical_zones.py  # Loads CBS 2011 GDB, matches stations to zones via point-in-polygon
├── process_statistical_zones.py   # Joins CBS socioeconomic data to zones, creates station mapping
├── download_historical_ballots.py # Downloads K16-K20 ballot CSVs from CEC CKAN API
├── normalize_tsne_names.py        # One-shot: normalize settlement names in tsne_*.json
├── normalize_coordinates.py       # One-shot: normalize settlement names in station_coordinates.json
├── add_locations_to_tsne.py       # Adds venue names to tsne_*.json from ballot_locations
├── enrich_settlements_wikipedia.py # Fetches Wikipedia summaries for settlements
├── prepare_election_26.py         # Workflow script to generate all election 26 data
├── generate_og_image.py           # Generates OG social preview image (Pillow)
├── geocode_with_amenities.py      # Geocodes ballot stations via Nominatim
├── fix_venues_google.py           # Precise geocoding via Google Places API
├── requirements.txt               # Python dependencies
├── venv/                          # Python virtual environment
├── wikipages/                     # Wikipedia HTML files with party info per election
├── site/                          # Static website (served directly)
│   ├── index.html                 # Landing page dashboard
│   ├── sankey.html                # Sankey vote flow diagram (was index.html)
│   ├── sankey.js                  # Sankey visualization logic
│   ├── tsne.html                  # T-SNE ballot clustering visualization
│   ├── geomap.html                # Geographic map with ballot station markers
│   ├── scatter.html               # Party support scatter plot (X vs Y axis)
│   ├── dhondt.html                # D'Hondt/Bader-Ofer seat allocation calculator
│   ├── irregular.html             # Irregular ballot analysis
│   ├── fraud-sim.html             # Election fraud detection simulator (K24→K25, K25→K26)
│   ├── regional.html              # Regional elections simulator (Voronoi + D'Hondt)
│   ├── settlement.html            # Settlement profile page (?name=...)
│   ├── party.html                 # Party profile page (?name=...)
│   ├── discussions.html           # Giscus discussions page
│   ├── styles.css                 # Shared CSS (desktop)
│   ├── i18n.js                    # Internationalization (Hebrew/English), nav rendering
│   ├── i18n.css                   # RTL/LTR + language toggle styles
│   ├── og-image.png               # Social preview image (1200x630)
│   ├── favicon.svg                # Site favicon
│   ├── images/                    # Archive election photos (NLI Dan Hadani collection)
│   │   ├── archive_poster_1.jpg   # Election poster
│   │   ├── archive_poster_2.jpg   # Election poster
│   │   └── archive_photo_3-12.jpg # Historical election photos (voting, rallies, results)
│   ├── m/                         # Mobile-optimized pages
│   │   ├── index.html             # Mobile dashboard
│   │   ├── sankey.html            # Mobile Sankey
│   │   ├── tsne.html              # Mobile T-SNE
│   │   ├── geomap.html            # Mobile geographic map
│   │   ├── scatter.html           # Mobile scatter plot
│   │   ├── dhondt.html            # Mobile D'Hondt calculator
│   │   ├── regional.html          # Mobile regional simulator
│   │   ├── settlement.html        # Mobile settlement profile
│   │   ├── party.html             # Mobile party profile
│   │   ├── irregular.html         # Mobile irregular ballot analysis
│   │   ├── discussions.html       # Mobile discussions
│   │   └── styles.css             # Shared CSS (mobile)
│   └── data/                      # JSON data for frontend (source of truth)
│       ├── transfer_*.json        # Vote transfer matrices between elections
│       ├── tsne_*.json            # T-SNE clustering data per election
│       ├── map_*.json             # Geographic/settlement data per election
│       ├── all_transfers.json     # Combined transfer data for all election pairs
│       ├── settlement_wiki.json   # Wikipedia enrichment data per settlement
│       ├── wiki_official_results.json # Official per-party vote totals per election (from Wikipedia)
│       ├── station_coordinates.json # Geocoded ballot station coordinates
│       ├── socioeconomic_clusters.json # CBS socioeconomic cluster per settlement
│       ├── statistical_zones.json     # CBS 2011 statistical area boundaries (centroids + metadata)
│       ├── statistical_zones_socioeconomic.json # Per-zone socioeconomic cluster/index
│       ├── station_zone_mapping.json  # Ballot station → statistical zone (YISHUV_STAT)
│       └── station_socioeconomic.json # Per-station socioeconomic cluster + zone ID

Running the Project

# Start local server
cd site && python3 -m http.server 8888
# Access at http://localhost:8888/

# Desktop pages served at root, mobile at /m/
# Each desktop page auto-redirects to /m/ on mobile devices

Architecture

Desktop vs Mobile

Every visualization has two HTML files: site/<page>.html (desktop) and site/m/<page>.html (mobile). Desktop pages detect mobile user agents and redirect to m/<page>.html. Mobile pages have:

Top bar: Fixed header with 🏠 home link, page title (flex:1 to push buttons left), ℹ info toggle, language toggle
Bottom tab bar: Horizontally scrollable pill-style nav (48px height, no home item since it's in top-bar)
BMC banner: Floating dismissible Buy Me a Coffee banner above tab bar (3s delay, sessionStorage persistence)
Touch-optimized bottom sheets instead of hover tooltips

Internationalization (i18n.js)

Central bilingual system (Hebrew/English). Key concepts:

Translation dict: All UI strings keyed by ID, with he and en values
data-i18n attributes: HTML elements get auto-translated via applyTranslations()
renderNav(activePageId): Generates the view-switcher nav bar on every page. The dashboard (index.html) uses injectLangToggle() instead (no nav bar)
navViews array: Defines page order and hrefs for navigation. Order: home → geomap → tsne → sankey → scatter → dhondt → regional → irregular
Helper functions: i18n.partyName(), i18n.settlementName(), i18n.settlementMatches(), i18n.fmtNum()
Language toggle: Persisted in localStorage, fires langchange event

Election 26 Feature Flag

Election 26 support is behind a URL feature flag: ?e26=1. When active:

i18n.SHOW_E26 is true (exposed in public API)
Each page conditionally adds election 26 buttons/options to its UI and makes K26 the default selected election/transition (sankey/dhondt/geomap/tsne/regional)
renderNav() propagates ?e26=1 to all nav links via addE26() helper
Data files (tsne_26.json, map_26.json, transfer_25_to_26.json) must be generated first via prepare_election_26.py

Simulated K26 scenario

K26 is a simulated scenario — ballot26.csv is generated by simulate_election_26.py from ballot25.csv:

TRANSFER_MATRIX: per K25 source party → K26 destination distribution (rows must sum to 1.0)
ROW_TURNOUT: per-source mobilization factor (<1 demobilized, >1 mobilized)
POP_GROWTH = 1.075: global multiplier modeling ~7.5% growth in eligible voters (Nov 2022 → Oct 2026)
Per-ballot Dirichlet noise with alpha=55 (concentration). alpha = α₀ = effective sample size; var = p(1−p)/(α+1)
Random seed 42 for reproducibility

The convex solver in generate_transfer_data.py recognizes the 25→26 transition (POP_GROWTH_PER_TRANSITION dict) and uses row_sum = 1.075 instead of 1.0 in the constraint — otherwise the matrix can't fit the population-grown K26 totals.

Hardcoded K26 totals appear in 3 places that must stay in sync when re-tuning:

site/dhondt.html parties26 array (+ site/m/dhondt.html)
site/data/wiki_official_results.json "26" entry (used by sankey side bar legend; has _26_note sibling key flagging simulation)
party_config.py ELECTIONS['26'] major_parties seats list

Set min_flow_threshold=15000 in generate_transfer_data.py to hide solver multicollinearity noise (~1% spurious cells) from sankey.

Settlement Profile Pages

URL pattern: settlement.html?name=<settlement_name> (URL-encoded Hebrew, normalized name)

Loads all 5 map files (map_21.json through map_25.json) to show voting trends
Loads settlement_wiki.json for Wikipedia data (thumbnail, description, extract)
Loads socioeconomic_clusters.json and station_coordinates.json
Desktop: search box, hero section, voting trends chart (Chart.js), party table + Leaflet mini-map (dots colored by winning party), sortable ballot table
Mobile: has top-bar, trends chart, party table, and sortable ballot table (no map)
Party colors use merged lookup from ALL elections (not just latest)
Cross-page links: geomap popups, tsne tooltips, scatter tooltips all link to settlement profiles

Party Profile Pages

URL pattern: party.html?name=<canonical_hebrew_name> (URL-encoded)

15 party families with per-election incarnations, merges, notes, and gaps
PARTY_FAMILIES config duplicated in desktop party.html and mobile m/party.html
Loads wiki_official_results.json for official per-party vote totals
Leader photos from map_*.json party info leader_image field (e.g. images/leaders/netanyahu.jpg)
Desktop: hero with leader photo + color bar, election history table (seats, votes, %, leader), trends chart, strongholds + mini-map, vote migration (from all_transfers.json)
Mobile: hero with leader photo, election history table, trends chart, top 10 strongholds

Social Preview (OG Tags)

All HTML files have og:type, og:title, og:description, og:image, and twitter:card meta tags. The OG image is at site/og-image.png (1200x630, generated by generate_og_image.py). The og:image URL must be absolute (https://kolot-nodedim.netlify.app/og-image.png).

Visitor Counter

The dashboard pages (index.html, m/index.html) use counterapi.dev for visitor counting. Uses sessionStorage to avoid double-counting within a session. URLs must have trailing slashes or the API returns 301 which fails with CORS.

Settlement Name Normalization

The CEC changed settlement name formatting between elections 22→23 (stripped hyphens, geresh, gershayim, parentheses). generate_map_data.py has normalize_name() that ensures consistent names across all elections. Key rules:

Remove: -, –, ', ׳, ", ״, (, )
Collapse multiple spaces, normalize יי → י
NAME_OVERRIDES dict for special cases (e.g. גולס → ג'וליס)
After regenerating tsne data, run normalize_tsne_names.py to apply same normalization
After changing coordinates, run normalize_coordinates.py
settlement_wiki.json keys must also be normalized

Key Technical Details

Ballot Matching Logic

When comparing elections, ballots are matched by settlement_name|ballot_number. Subdivided ballots (e.g., 14.1, 14.2) follow this logic:

Exact match first: 14.1 matches 14.1
Fallback for .1 only: 14.1 can match 14 if 14.1 doesn't exist in the other election
.2, .3, etc. do NOT fallback: 14.2 only matches 14.2, never 14

This is implemented in:

site/scatter.html: getBaseKey() function
site/geomap.html & site/m/geomap.html: getComparePct() function (party comparison mode)
generate_transfer_data.py: get_base_ballot_id() function

Data Formats

tsne_*.json (per election):

{
  "parties": [{"name": "...", "symbol": "...", "color": "#..."}],
  "stations": [
    {"n": "settlement", "b": "ballot", "l": "location", "v": total_voters, "e": eligible, "t": turnout, "p": {"party": proportion}}
  ]
}

map_*.json (per election):

{
  "election": {...},
  "parties": [...],
  "settlements": [
    {
      "name": "...", "lat": N, "lng": N, "voters": N, "eligible": N,
      "turnout": N, "ballotCount": N, "proportions": {...},
      "winningParty": "...", "cluster": N,
      "ballots": [
        {"b": "ballot_num", "v": voters, "e": eligible, "t": turnout, "l": "location", "p": {"party": pct}}
      ]
    }
  ],
  "stats": {"totalSettlements": N, "totalBallots": N, "totalVoters": N, "totalEligible": N, "totalLists": N, ...}
}

transfer_*.json (between elections):

{
  "from_election": {...},
  "to_election": {...},
  "nodes_from": [...],
  "nodes_to": [...],
  "transfer_matrix": [[...]],
  "stats": {"common_precincts": N, "r_squared": 0.xx}
}

settlement_wiki.json (settlement → Wikipedia data):

{
  "תל אביב יפו": {
    "title": "...", "description": "...", "extract": "...(max 500 chars)",
    "wiki_url": "https://he.wikipedia.org/wiki/...", "thumbnail": "https://..."
  }
}

wiki_official_results.json (per-party official vote totals):

{
  "21": [
    {"name": "הליכוד", "seats": 35, "votes": 1140370},
    {"name": "כחול לבן", "seats": 35, "votes": 1125881}
  ]
}

Generating Data

source venv/bin/activate

# Vote transfer matrices
python generate_transfer_data.py
cp data/transfer_*.json site/data/

# T-SNE clustering
python generate_tsne_data.py
python add_locations_to_tsne.py   # Adds venue names from ballot_locations_*.json
python normalize_tsne_names.py    # Normalize settlement names

# Geographic map data (reads tsne, applies normalization, counts lists from CSV)
python generate_map_data.py

# Wikipedia enrichment (rate-limited, ~3 min for 1100 settlements, resumes from cache)
python enrich_settlements_wikipedia.py

# Election 26 (simulated scenario workflow)
python simulate_election_26.py                          # generate ballot26.csv (alpha=55, seed=42)
python prepare_election_26.py --real-csv ballot26.csv   # run tsne/transfer/map for K26
# After re-tuning the matrix, also refresh hardcoded K26 totals in:
#   site/dhondt.html, site/m/dhondt.html, site/data/wiki_official_results.json, party_config.py

IMPORTANT: Scripts write to data/, but the web server reads from site/data/. Always copy generated files:

cp data/transfer_*.json site/data/
cp data/tsne_*.json site/data/

(Some scripts like generate_map_data.py and enrich_settlements_wikipedia.py write directly to site/data/)

CRITICAL: The l (location/venue name) field in tsne_*.json is NOT generated by generate_tsne_data.py. It must be added separately by add_locations_to_tsne.py, which reads from data/ballot_locations_*.json. If you regenerate tsne data without re-running add_locations_to_tsne.py, station search by venue name will break (the l field will be missing).

Statistical Zones & Socioeconomic Data

CBS 2011 Census statistical areas provide intra-city socioeconomic resolution. The pipeline:

source venv/bin/activate
python download_statistical_zones.py   # Load CBS 2011 GDB → match stations to zones
python process_statistical_zones.py    # Join CBS 2021 socioeconomic data → per-station mapping

Key concepts:

YISHUV_STAT: Zone ID format = settlement_code * 10000 + stat_area (e.g. 30001335 = Jerusalem area 1335)
CBS 2011 boundaries: From statisticalareas_demography2019.gdb (EPSG:2039 → WGS84 conversion)
CBS 2021 socioeconomic index: Publication 1955, tables T12 (stat areas), T01 (local authorities), T07/T08 (regional localities)
Matching priority: T12 exact stat area → T01 settlement-level fallback → T07/T08 regional fallback
Point-in-polygon: Station coordinates matched to zone polygons via Shapely
Coverage: ~31,550/31,590 stations get a cluster (99.87%)

CRITICAL: Must use CBS 2011 boundaries (not 2022 ArcGIS). The 2021 socioeconomic data is keyed to 2011 Census stat areas. Using 2022 boundaries only matches 67.6% of T12 zones; 2011 boundaries match 100%.

Output files (all write directly to site/data/):

statistical_zones.json — compact zone data with centroids
statistical_zones_socioeconomic.json — per-zone cluster/index/rank
station_zone_mapping.json — station key → YISHUV_STAT
station_socioeconomic.json — station key → {zone, cluster}

Historical Elections (K16-K20)

Elections 16-20 (2003-2015) are downloaded via download_historical_ballots.py from the CEC CKAN API. Key differences from K21-K25:

Encoding: UTF-8-BOM (not ISO-8859-8)
Ballot field: מספר קלפי (not קלפי)
K16-K17 ballot numbering: Uses x10 numbering (10, 20, 30...) — ballot_number_divisor: 10 in party_config normalizes to 1, 2, 3...
K17 eligible voters: בזב column is empty — abstention estimated from national totals
Abstention toggle disabled: For transitions 16→17 and 17→18 (unreliable per-ballot eligible data)

Geocoding Stations

Station coordinates are in site/data/station_coordinates.json. Sources by priority:

manual: Hand-corrected coordinates (when automated sources were wrong)
google_venue: Precise venue coords from Google Places API (fix_venues_google.py)
venue: Venue coords from Nominatim (geocode_with_amenities.py)
settlement: Fallback to settlement center coordinates
Google API key: set GOOGLE_MAPS_API_KEY env var (never commit the key)
After fixing coordinates: Re-run download_statistical_zones.py + process_statistical_zones.py to update zone assignments

Party Configuration

party_config.py contains the ELECTIONS dict with metadata for each election (21-26):

Election name (Hebrew + English), date, file path, encoding
Eligible voters, votes cast, turnout percentage
Major parties: symbols, names, seats
Reference: wikipages/ folder has Wikipedia HTML with correct party info per election

Visualization Pages

index.html (Dashboard)

Landing page with animated hero stats and card grid linking to all visualizations
Stats: elections (25), lists, settlements, ballots, eligible voters, voted, visitors — all with rolling number animation
Stats loaded from all 5 map_*.json files (max ballots across elections, latest for rest)
Visitor counter integrated as regular stat (counterapi.dev)
Archive photos: 12 historical NLI images in site/images/, 2 randomly chosen per page load, flanking the 9 view cards
Credit: Dan Hadani Archive, Pritzker Family National Photography Collection, National Library of Israel
Photos hidden on screens < 1200px
Discussions link next to language toggle
No nav bar (uses injectLangToggle() only)

sankey.html (Vote Transfer Flow)

Shows vote flow between consecutive elections as Sankey diagram (K16→K17 through K24→K25)
Uses sankey.js for rendering
Data: transfer_*.json and all_transfers.json
Abstention toggle: Shows "did not vote" pseudo-party (לא הצביעו) using *_abstention.json files
Abstention disabled for K16→K17 and K17→K18 (unreliable per-ballot eligible voter data)

tsne.html (Ballot Clustering)

T-SNE clustering of ballot boxes by voting patterns
Color modes: turnout, party support, socioeconomic cluster
Socioeconomic data from CBS 2021 index (1-10 scale), per statistical area within cities
getStationCluster(station) returns {cluster, statArea} — per-station lookup first, settlement-level fallback
Tooltips show cluster + full YISHUV_STAT zone ID (e.g. "30001335")

geomap.html (Geographic Map)

Leaflet map with clustered ballot station markers
Color by party support, turnout, or socioeconomic cluster
Per-station socioeconomic coloring (different ballots in same city show different clusters)
getStationCluster(station) same pattern as tsne — per-station first, settlement fallback
Settlement search with autocomplete
Uses CartoDB Positron tiles with CSS brightness(0.7) for dark theme
Party support comparison mode: When a party is selected, a row of election pills appears below the legend. Clicking one loads the comparison election's tsne data and shows a diverging bichromatic scale (orange = lost support, teal = gained support, clamped ±20%). Party matching across elections uses the party symbol (not name). Ballot matching uses settlement|ballot key with .1 suffix fallback. State: compareElection, compareData, compareLookup, compareCache (cached fetches). Tooltips/popups show diff with both election percentages. Clusters show weighted average diff.

scatter.html (Party Comparison)

Compare party X support vs party Y support across ballots
Can compare same or different elections
Settlement filter with autocomplete search
Units toggle: percentages vs absolute votes

dhondt.html (Seat Calculator)

Interactive Bader-Ofer (modified D'Hondt) seat calculator
Adjust vote counts to see seat changes

irregular.html (Anomaly Detection)

Identifies statistically unusual ballot boxes

fraud-sim.html (Election Fraud Detection Simulator)

Interactive fraud detector sandbox for K24→K25 (real data) and K25→K26 (simulated scenario)
Fraud injection: Parametric controls inject synthetic ballot-level fraud — select target settlement, number of ballots, intervention strength (% of ballot votes to reassign)
Attack scenarios: Predefined templates for party-to-party fraud (e.g., Likud→Shas redirections, bloc-to-bloc shifts)
Multiple detectors: Geo (geographic isolation), manifold (t-SNE neighbor distance), matched-filter (1D variance detection)
Portfolio fusion: Combines detector scores via weighted combination; optimizable threshold shown on dynamic ROC curve with AUC
Output: Ranked ballot list with before/after vote distributions, seat-equivalent impact, per-detector anomaly scores
Druze exclusion: 14 specified Druze settlements excluded by default (settlement-level checkbox to re-include)
Visualization modes: ROC plot (dual curves for current vs population level), side-by-side tampered ballot comparison
First-time intro modal: Explains concepts (detector types, fraud scenarios, ROC interpretation)
Default state: K24→K25 transition, top-100 ballots, 400-ballot injection, 17% intervention strength

regional.html (Regional Elections Simulator)

Voronoi-based regional election simulation with D'Hondt allocation

settlement.html (Settlement Profile)

Per-settlement deep dive with Wikipedia info, voting trends, party breakdown
URL parameter: ?name=<settlement_name>

Paper

The paper/ directory is a separate git repo (harelc/elections-paper) embedded in this repo. It has its own remote, branch (main), and commit history. Commit and push it independently:

# Compile the paper (uses tectonic, NOT pdflatex/latexmk)
cd paper && tectonic paper.tex

# Commit paper changes (separate repo!)
cd paper && git add -A && git commit -m "..." && git push
# Then update the submodule ref in the parent:
cd .. && git add paper && git commit -m "Update paper submodule"

# Key files:
# paper/paper.tex         — Main paper source
# paper/references.bib    — Bibliography
# paper/figures/           — Generated figures (PDF)
# paper/generate_figures.py — Script to regenerate figures
# paper/bootstrap_analysis.py — Bootstrap confidence interval analysis

Common Issues

Browser cache: After regenerating data, hard refresh (Cmd+Shift+R) or use incognito
Python environment: Always use source venv/bin/activate for generate scripts
Data sync: Scripts write to data/, web serves from site/data/ — copy after generating
Map tiles: Using CartoDB Positron (no auth required, unlike Stadia Maps)

Election IDs

16: January 2003
17: March 2006
18: February 2009
19: January 2013
20: March 2015
21: April 2019
22: September 2019
23: March 2020
24: March 2021
25: November 2022
26: 27 October 2026 (estimated) — simulated scenario, feature-flagged behind ?e26=1

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Elections Vote Transfer Project

Project Structure

Running the Project

Architecture

Desktop vs Mobile

Internationalization (i18n.js)

Election 26 Feature Flag

Simulated K26 scenario

Settlement Profile Pages

Party Profile Pages

Social Preview (OG Tags)

Visitor Counter

Settlement Name Normalization

Key Technical Details

Ballot Matching Logic

Data Formats

Generating Data

Statistical Zones & Socioeconomic Data

Historical Elections (K16-K20)

Geocoding Stations

Party Configuration

Visualization Pages

index.html (Dashboard)

sankey.html (Vote Transfer Flow)

tsne.html (Ballot Clustering)

geomap.html (Geographic Map)

scatter.html (Party Comparison)

dhondt.html (Seat Calculator)

irregular.html (Anomaly Detection)

fraud-sim.html (Election Fraud Detection Simulator)

regional.html (Regional Elections Simulator)

settlement.html (Settlement Profile)

Paper

Common Issues

Election IDs

FilesExpand file tree

CLAUDE.md

Latest commit

History

CLAUDE.md

File metadata and controls

Elections Vote Transfer Project

Project Structure

Running the Project

Architecture

Desktop vs Mobile

Internationalization (i18n.js)

Election 26 Feature Flag

Simulated K26 scenario

Settlement Profile Pages

Party Profile Pages

Social Preview (OG Tags)

Visitor Counter

Settlement Name Normalization

Key Technical Details

Ballot Matching Logic

Data Formats

Generating Data

Statistical Zones & Socioeconomic Data

Historical Elections (K16-K20)

Geocoding Stations

Party Configuration

Visualization Pages

index.html (Dashboard)

sankey.html (Vote Transfer Flow)

tsne.html (Ballot Clustering)

geomap.html (Geographic Map)

scatter.html (Party Comparison)

dhondt.html (Seat Calculator)

irregular.html (Anomaly Detection)

fraud-sim.html (Election Fraud Detection Simulator)

regional.html (Regional Elections Simulator)

settlement.html (Settlement Profile)

Paper

Common Issues

Election IDs