Team Builder is an automated way for your league to evenly distribute players onto teams without manually selecting each one. It's especially useful for blind drafts where player identities aren't shown, only their evaluation or data-based scores.
How Team Builder Works
When you run Team Builder, players are automatically assigned to teams based on their evaluations, player data, and any rules you've set up. The system uses optimization algorithms to balance teams as fairly as possible while respecting your constraints.
Bravara provides a step-by-step wizard that walks you through all required setup, so you don't have to guess what needs to be configured.
Common League Flows
Different leagues run drafts differently. Team Builder supports all of the following:
Leagues with evaluations: Teams are balanced primarily using evaluation scores.
Leagues without evaluations: Teams are balanced using player registration data only.
Mixed leagues: Some players have evaluations, others do not.
The wizard automatically adapts to the data you have.
The Team Builder Wizard
Step 1: Team Size
Configure how many players should be on each team.
Enter the desired number of players per team
The system shows how teams will be distributed
If numbers do not divide evenly, some teams may have one extra player
A warning appears if any players will not be placed
Step 2: Balance
Choose how players are ranked and distributed.
If most players have evaluations
Create draft rounds based on evaluation categories
Each round can emphasize a different skill (for example pitching, catching, or overall)
Players are ranked within each round by the selected category
If your league does not hold evaluations
Use Player Data Balance to create teams using registration data only.
Any registration field can be used (height, weight, experience, age, t-shirt size, etc.)
You may assign weights (1 to 5) to each field to control importance
Each field is normalized onto the league’s rating scale
How normalization works
For each selected field:
The lowest value in the league maps to the lowest rating (for example 1)
The highest value maps to the highest rating (for example 5)
All other players are distributed proportionally between those values
Example:
Shortest player receives a 1
Tallest player receives a 5
Everyone else is scaled evenly in between
This allows raw data like height, weight, or experience to behave like evaluation scores for balancing purposes.
Mixed data
Players with evaluations use evaluation scores
Unrated players use player data scores
If most players are unrated, the wizard automatically defaults to Player Data Balance
Step 3: Relationships
Control which players should be placed together or kept apart.
Siblings
Bravara automatically detects siblings using shared guardian email addresses
Siblings are placed on the same team by default
You may remove sibling connections if you do not want those players grouped together
Custom relationships
Add relationships manually for carpools, family friends, or other requests
Relationships can be added or removed directly in Team Builder
Each player is still ranked individually by skill or data. Relationships only affect team placement.
Step 4: Rules
Create advanced constraints for team formation. There are two types of rules:
Player Grouping Rules
Keep players together or apart
Use any imported registration field
Examples:
Keep players from the same school together
Separate players from the same neighborhood
Separate players based on conflict flags
Team Requirements
Enforce minimums, maximums, or ratios on each team
Examples:
Each team must have at least 1 goalkeeper
Each team may have at most 2 elite players
At least 30% of each team must have prior experience
For each rule, choose one of the following:
Must Satisfy: Team Builder fails if the rule cannot be met
Try to Satisfy: The system prioritizes the rule but may relax it for overall balance
Note: Any field imported during player import can be used in rules.
Step 5: Players
Review players who are already assigned to teams.
Frozen Players
Coaches’ children
Returning players
Special placement requests
Frozen players remain locked to their teams. Team Builder fills the remaining spots around them.
Step 6: Unrated Players (if applicable)
Appears when some players have no evaluations.
Unrated players are treated as approximately average when no player data is available
When player data exists, it is used instead of assuming average skill
Key Concepts
Imported Player Data
When importing players, you may include any columns you want.
Examples:
Height
Weight
Years of experience
School
Coach request flags
Friend request groups
These fields can later be used for:
Balancing teams
Grouping or separating players
Team minimums or maximums
Frozen Players
Players assigned to a team before Team Builder runs. These placements are respected and never changed.
Hard vs Soft Rules
Hard rules must be satisfied or Team Builder fails
Soft rules are best-effort and may be relaxed for balance
After Team Builder Runs
Making Changes
Trades and manual adjustments are allowed
All changes are tracked for transparency
Undoing Team Builder
Use Undo to reset team assignments
Running Team Builder again with identical settings produces the same result
Changing players, rules, or data will change results
Tips for Best Results
Evaluate as many players as possible when evaluations are used
Assign frozen players before running Team Builder
Verify guardian emails for accurate sibling detection
Start with soft rules
Review the wizard summary before committing
Friend Requests and Coach Requests
Friend requests, coach requests, and similar placement preferences are supported through imported data and Team Builder rules.
A detailed guide for handling friend and coach requests will be added here.
Premium Features
Some Team Builder features require a premium subscription:
Evaluation-based balancing
Player Data Balance
Advanced grouping and team requirement rules
Basic sibling handling and frozen players are always available.
Was this article helpful?
That’s Great!
Thank you for your feedback
Sorry! We couldn't be helpful
Thank you for your feedback
Feedback sent
We appreciate your effort and will try to fix the article