A reliability score is a rating computed from a worker's history of attendance and hours worked, used to inform how future shifts are allocated among people who bid for the same work.

How it works

Reliability scores are usually built from a small set of inputs, recalculated at the end of each roster period:

• Attendance rate — how often a person showed up for shifts they were allocated.
• Hours fulfilment — whether they worked the hours expected, not just showed up briefly.
• No-show penalty — an extra deduction for not showing up without notice, weighted more heavily than a cancellation given with fair notice.
• Recency weighting — recent periods usually count for more than older ones, so a score reflects current behaviour rather than a distant track record.

The result is typically expressed as a percentage or a simple band (for example, green/amber/red), rather than raw attendance numbers, so it's quick to scan when allocating.

Why it matters

When more people bid for a shift than there's capacity for, a manager needs some basis for choosing beyond "first bid." A reliability score gives that basis without relying on personal familiarity or memory, and it rewards people who consistently show up and complete their shifts. It also gives workers a transparent reason for allocation outcomes, rather than an opaque manager's call.

Reliability scores are usually only visible to management, not to other staff, since they touch individual performance history.

In Chronomancer

Chronomancer computes a reliability score automatically when a roster period completes: 100 × attendance rate × hours fulfilment (capped) − no-show penalty, with recent periods weighted more heavily via exponential decay. The score is cached per contact, shown as a green/amber/red band with a percentage, and factored into allocation alongside priority tier — though by default it's hidden from managers running day-to-day rosters and reserved for admin-level views.