1/3/22, 11:28 AM Myanmar’s internet suppression KASPR Datahaus measures internet connectivity and speed in Myanmar every hour using a sample of active IP addresses. The sample includes all identifiable IP addresses throughout Myanmar except in the most densely-populated area of Yangon, where more than half of identifiable IP addresses were randomly sampled. It includes mobile data from cell towers but is primarily fixed-line and wireless internet connections. Researchers at Monash University’s IP Observatory computed an expected range of connectivity for each hour, based on the number of active connections during undisrupted service between November 2020 and March 2021. This expected range, which fluctuates depending on time of day and week, was the baseline against which the researchers compared connectivity since the service disruptions began. They also adjusted the data to account for monthly sampling differences. To illustrate the normal range, Reuters has shown the average high point and average low point of this expected range. The researchers used the same five-month period to establish a baseline for internet speed. They computed the expected hourly average ping response time, measured in milliseconds, of all IP addresses in the sample. They normalized the data to account for monthly sampling differences and variance in infrastructure quality. Hours with average speeds below the expected range were flagged as slow periods. The hours shown on charts are in local time. Additional reporting by Fanny Potkin and Nerijus Adomaitis Edited by Matthew Tostevin, Kay Johnson, Janet Roberts and Jon McClure Politicians, doctors and a fortune-teller: Myanmar’s new wave of detainees Speed and trust https://graphics.reuters.com/MYANMAR-POLITICS/INTERNET-RESTRICTION/rlgpdbreepo/ 30/31