Our Solution

Using algorithms to generate thousands of fair redistricting options

The Traditional Problem

Politicians Choose One Map

Traditionally, redistricting works like this: politicians draw one map behind closed doors, present it to the public, and say "take it or leave it." This process is:

  • Opaque - No one knows how the lines were drawn
  • Unfair - Designed to benefit the party in power
  • Unaccountable - No alternatives to compare against
πŸ‘₯

Politicians meet behind closed doors

β†’
πŸ—ΊοΈ

Draw one map

β†’
πŸ“’

Present to public: "Take it or leave it"

Our Algorithmic Approach

Instead of one map chosen by politicians, we generate thousands of fair options using computers.

1

Generate 10,000+ Maps

Our algorithms create thousands of possible district configurations, each following legal requirements like equal population and contiguity.

2

Score Each Map for Fairness

We evaluate each plan using multiple criteria: compactness, demographic balance, and population equality.

3

Present Top 100 Plans

The fairest 100 plans are presented to the community with full transparency about how they were created and scored.

How Our Algorithm Works

Markov Chain Monte Carlo

We use a sophisticated mathematical technique called "Markov-Chain Monte Carlo" (MCMC). The algorithm starts with one map that is legally compliant and gradually improves it by making thousands or millions of small changes to reach a final map.

Fairness Metrics

Each map is scored using established fairness measures: Compactness (how round the districts are), and Demographic balance (fair representation of all communities), Each map is evaluated using a combination of these metrics.

Ensemble Analysis

By generating thousands of maps, we create an "ensemble" that shows what fair districts naturally look like. This gives us a baseline to compare against and ensures our results aren't just one random outcome which may not be fair.

Why This Approach Works

Transparency

Every step of our process is documented and publicly available. You can see exactly how each map was created and scored.

Objectivity

Computers don't have political preferences. They follow the rules we give them and create maps based on fairness, not politics.

Community Choice

Instead of one option, communities get tens of different fair choices. They can pick the plan that best represents them.

Legal Compliance

All our maps follow legal requirements like the Voting Rights Act and equal population standards.

Real Results: Santa Clara County

We've Already Done This

In Santa Clara County, our algorithm generated over 50,000 redistricting plans and identified 100 of the fairest options. Each plan was evaluated using:

  • Polsby-Popper scores - A measure of how compact each district is
  • Reock scores - Another measure of how compact each district is
  • Demographic segregation indices - Ensuring fair representation of all communities
  • Population balance validation - Making sure each district has an equal number of people
  • Contiguity checks - Ensuring districts are connected (no islands)
View Santa Clara Results
50,000+ Plans Generated
100 Top Plans Selected

Ready to Bring This to Your Community?

Our algorithmic approach can work for any city, county, or region in California. Learn more about our methodology or get involved.