Even shift distribution

What it means

Over the scheduling period, morning, afternoon, and night assignments should be distributed as evenly as possible across all employees. The engine calculates a fair share for each shift type and penalizes deviations from that target.

Why it exists

Even if the hard core shift balance constraint ensures no single employee is overloaded on one shift type, it does not guarantee fairness across the team. One employee could have 7 mornings while another has 5, both within the hard limit. This soft constraint pushes the optimizer toward equal distribution across the whole team.

How the penalty is calculated

For each core shift type, the engine computes the total number of slots across the period and divides by the number of employees to get a target per person. For each employee, it measures the absolute deviation from that target and multiplies by the weight.

Example: if the morning target is 7 per person and one employee has 9, their deviation is 2. That contributes 2 × 10 = 20 to the total penalty.

Example

Suppose a 30-day period has 22 morning slots to fill across 5 employees. The target is roughly 4 per person. A roster where counts are 4-4-4-5-5 scores better than one where counts are 3-3-4-6-6, even though both satisfy the hard balance constraint.

Parameters

Interaction with other constraints

This constraint works alongside prior-month history bias. History bias nudges assignments away from shifts an employee already worked heavily last month, while shift balance nudges toward equal counts this month. Both contribute to the same penalty total.