Attendance Bias
Calibration adjustment for camera occlusion, hidden rows, partial views.
Attendance bias is a calibration factor applied to raw sensor counts to account for systematic undercounting or overcounting. Common sources of bias include camera occlusion (pillars, balcony overhangs), partial field-of-view coverage, double-counting at multi-entrance venues, and hidden seating areas outside the sensor's line of sight.
Kingdom Metrics allows church administrators to set a bias adjustment per camera — for example, +8% if the sensor consistently misses a balcony section that holds about 8% of the congregation. The adjustment is applied after the AI count and before the number is stored as the official attendance record.
Why it matters for your church: No sensor placement is perfect. A well-calibrated bias adjustment bridges the gap between what the camera can see and the full picture. The result is an attendance number your team can trust — close enough to ground truth that it is more accurate than a volunteer headcount, and consistent enough week-over-week to produce reliable trends.