About

Fairness is rarely random. In classroom management, corporate workshops, or competitive gaming, true balance requires more than a shuffle algorithm. It requires weighted distribution. This tool solves the problem of uneven teams by allowing you to assign a numerical "skill value" to each entry. Whether balancing academic grades, employee seniority, or player MMR, the algorithm minimizes the variance between group totals.

Beyond mathematics, social dynamics matter. The strict exclusion logic allows you to flag individuals who must be separated or pairs that must remain together. Unlike static generators, this interface remains fluid: once the initial calculation is complete, you can manually drag and drop members between groups to finalize the arrangement before exporting to CSV or Markdown.

Formulas

When Skill Balancing is enabled, the tool attempts to minimize the difference between the total skill sum of each group. The target sum S_target for any group is defined as:

S_target ≈ N∑i=1 v_ik

Where N is the total number of participants, v_i is the skill value of the i-th participant, and k is the number of groups. The algorithm uses a Snake Draft sorting method (1, 2, 3... 3, 2, 1) to distribute high-value items evenly.

Reference Data

Feature	Standard Randomizer	This Tool (Snake Draft)
Algorithm	Pure Random (Fisher-Yates)	Sorted Snake Draft + Optimization
Input Format	Names Only	Name \| Skill Level (1-100)
Goal	Chaos / Chance	Mathematical Parity (∑ Group A ≈ ∑ Group B)
Constraint Handling	None	Keep Together / Force Separate
Variance	High (σ² is uncontrolled)	Minimized (Δ between group means is low)
Post-Processing	None (Static Text)	Drag & Drop Rebalancing
Persistence	Lost on Refresh	Saved to LocalStorage

Frequently Asked Questions

To balance teams, add a vertical bar and a number after the name. For example: "John Doe | 85". The generator will interpret "85" as the skill weight. You can use any scale (1-5, 1-100, Elo ratings), as long as it is consistent across the list.

Yes. While we don't have a complex syntax for this yet, the best workaround in the current version is to hyphenate their names as a single entry (e.g., 'Alice-and-Bob | 10') and then manually drag them if needed, or rely on the Drag & Drop feature to fix specific placements after generation.

Yes. We use your browser's LocalStorage to automatically save the input list and your settings. If you accidentally close the tab, your data will be there when you return.

We sort the list descending by value. Then we iterate through groups, adding the highest available member to the group with the current lowest sum. This greedy approach ensures that the standard deviation σ of group sums stays minimal.