Guess the disruption

One puzzle a day. You see a public daily series with the place, dates and units removed, indexed so the pre-event level reads 100, and the day the event began marked as day 0. Pick which of 4 recorded disruptions it shows. The answer then reveals the event, its real dates and the source.53 puzzles are available in this build, drawn from 73 published records; the puzzle changes at midnight in your own time zone.

Today's puzzle

The daily puzzle needs JavaScript. A worked example with its answer is below, and every disruption has its own event page.

Loading today's puzzle...

Which disruption is this?

Worked example and answer

This is the answer view for one puzzle, shown here so the page reads without JavaScript. The series below is the one an unlabelled version of it would be drawn from, for Attacks on shipping in the Red Sea divert traffic around Africa.

Event: Attacks on shipping in the Red Sea divert traffic around Africa (Security incident)
Series: Transits, all vessel types, Bab el-Mandeb Strait
Started: 2023-11-19 (19 November 2023)
Ended: Ongoing at the time of this build

Transits, all vessel types, Bab el-Mandeb Strait, 2023-10-05 (5 October 2023) to 2024-02-01 (1 February 2024): values ranged from 27 transits per day on 2024-01-14 to 95 transits per day on 2023-10-27, with a mean of 64.4 transits per day. The latest value, for 2024-02-01, was 32 transits per day. The shaded band marks the recorded event, 2023-11-19 to 2024-02-01.

Source: IMF PortWatch (portwatch.imf.org), International Monetary Fund, using UN Global Platform AIS data. Statistics derived by World Freight Monitor. Terms.

Transits, all vessel types, Bab el-Mandeb Strait (transits per day).
Date	transits per day
2023-10-05	80
2023-10-06	90
2023-10-07	78
2023-10-08	71
2023-10-09	71
2023-10-10	73
2023-10-11	86
2023-10-12	80
2023-10-13	78
2023-10-14	88
2023-10-15	77
2023-10-16	86
2023-10-17	82
2023-10-18	77
2023-10-19	84
2023-10-20	87
2023-10-21	87
2023-10-22	75
2023-10-23	62
2023-10-24	71
2023-10-25	86
2023-10-26	62
2023-10-27	95
2023-10-28	83
2023-10-29	62
2023-10-30	84
2023-10-31	77
2023-11-01	61
2023-11-02	94
2023-11-03	87
2023-11-04	85
2023-11-05	69
2023-11-06	65
2023-11-07	67
2023-11-08	77
2023-11-09	84
2023-11-10	70
2023-11-11	86
2023-11-12	71
2023-11-13	84
2023-11-14	70
2023-11-15	76
2023-11-16	75
2023-11-17	87
2023-11-18	84
2023-11-19	59
2023-11-20	72
2023-11-21	87
2023-11-22	84
2023-11-23	75
2023-11-24	86
2023-11-25	74
2023-11-26	88
2023-11-27	65
2023-11-28	77
2023-11-29	89
2023-11-30	72
2023-12-01	64
2023-12-02	67
2023-12-03	90
2023-12-04	85
2023-12-05	72
2023-12-06	66
2023-12-07	75
2023-12-08	82
2023-12-09	77
2023-12-10	65
2023-12-11	62
2023-12-12	83
2023-12-13	81
2023-12-14	91
2023-12-15	76
2023-12-16	65
2023-12-17	57
2023-12-18	52
2023-12-19	56
2023-12-20	63
2023-12-21	53
2023-12-22	50
2023-12-23	50
2023-12-24	69
2023-12-25	60
2023-12-26	57
2023-12-27	45
2023-12-28	45
2023-12-29	70
2023-12-30	44
2023-12-31	49
2024-01-01	36
2024-01-02	52
2024-01-03	41
2024-01-04	54
2024-01-05	45
2024-01-06	45
2024-01-07	43
2024-01-08	51
2024-01-09	53
2024-01-10	43
2024-01-11	55
2024-01-12	35
2024-01-13	43
2024-01-14	27
2024-01-15	29
2024-01-16	38
2024-01-17	46
2024-01-18	32
2024-01-19	36
2024-01-20	39
2024-01-21	40
2024-01-22	45
2024-01-23	35
2024-01-24	31
2024-01-25	34
2024-01-26	29
2024-01-27	32
2024-01-28	31
2024-01-29	37
2024-01-30	28
2024-01-31	37
2024-02-01	32

This is a measured change in public data during the event window, compared with a baseline before it. It shows timing, not cause. We describe an event as causing a change only when an official source says so.

How the puzzle is built

Only events with a daily measured series qualify. The window runs 45 days before the event began through its end plus 14 days (an ongoing event is shown for 60 days), capped at 120 points and at the last day of data.
Values are shown as a percentage of the median of the pre-event days, so the scale does not give the place away. The reveal names the real series.
The three wrong choices are other recorded events of different types where possible.
Your streak (consecutive days answered correctly) is kept only in this browser. Nothing is sent anywhere.
Many disruptions leave no visible mark in one series, and some marks come from the calendar rather than the event, so some puzzles are hard on purpose. The reveal shows timing, not cause.

Series values come from the sources named in each reveal. Closest calls shows where severity and data disagree.