Why it matters

Closure feels like progress until the numbers flatten. A dashboard that shows the day-by-day violation count with the ratio of each day to the previous makes a stall visible, and a simple extrapolation tells you roughly where it will stop, which decides whether a different strategy is needed.

Worked example

1
2
3
4
5
6
7
8
9
10
11

v = [420, 260, 170, 130, 118, 112, 109]
ratios = [round(v[i + 1] / v[i], 3) for i in range(len(v) - 1)]
assert ratios == [0.619, 0.654, 0.765, 0.908, 0.949, 0.973]
stall_at = next(i for i in range(len(ratios) - 1) if ratios[i] >= 0.9 and ratios[i + 1] >= 0.9)
assert stall_at == 3                                  # ratios[3] and ratios[4] are both at or above 0.9
# extrapolate the floor from the last three differences (geometric decay of the differences)
d = [v[i + 1] - v[i] for i in range(len(v) - 3, len(v) - 1)]          # last two differences: -6 and -3
q = d[1] / d[0]
floor = v[-1] + d[1] * q / (1 - q)
assert d == [-6, -3] and q == 0.5 and floor == 106.0
print("PASS: ratios", ratios, "; stall detected from day", stall_at + 1, "; projected floor about", floor)

Violations fall from 420 to 109, but the day-to-day ratio rises from 0.62 to 0.97: each day removes a smaller share. Two consecutive days with a ratio of at least 0.9 (days 4 and 5) flag a stall. Extrapolating the last differences (-6, then -3, a ratio of 0.5) puts the floor near 106 violations, about a quarter of where closure started. Waiting will not reach zero; the remaining violations need a different kind of fix: structural changes, a change of the method, or an agreed waiver with evidence. The extrapolation is rough, but the decision it supports is not.

What belongs on the dashboard

What the review looks for

• The same definitions every day, so the series is comparable (union counts, same scenarios).

• A stall triggers a decision, not more of the same: structural fix, method change, or waiver.

• The dashboard distinguishes fixed from waived from open.

Related topics

• Avoiding Fix Breaks Fix

• Closure Strategy