Speed testing according to the rules

TL;DR

When we carry out speed tests according to the guidelines of the Netherlands Authority for Consumers and Markets, it appears that our Dutch (!) internet provider (Ziggo) properly delivers the promised internet speed.

The differences between the speed tests to be used (a Dutch (!) Ookla Speedtest clone and M-Lab) are minimal. In both speed tests, at least the advertised speed was measured for each measurement.

 

Contents

  1. Introduction
  2. Method of measurement
  3. The measurements
  4. Conclusions
 

Introduction

According to the Dutch (!) Authority for Consumers and Markets you should use the Ookla Speedtest or M-Lab to measure the internet speed of your ISP.

Besides that should you test at least ten times during al least 3 days your internet speed, according to this Dutch (!) beleidsregel.

 

Method of measurement

For this test we use the Dutch (!) Ookla speedtest clone of Ziggo (our ISP) at https://www.ziggo.nl/speedtest and M-Lab.

We will test the internet speed twice a day during a week. We will use the following scheme for testing per day. First hour: Ookla, M-Lab. Second hour: M-Lab, Ookla, and so on.

 

The measurements

The following has been measured. Note that all speeds are measured in Mbps.

  1. Monday circa 12:30h. Ookla: 101.3 Mbps, M-Lab: 101.56 Mbps
  2. Monday circa 16:10h. M-Lab: 101.60 Mbps, Ookla: 101.6 Mbps
  3. Tuesday circa 10:30h. Ookla: 101.7 Mbps, M-Lab: 101.34 Mbps
  4. Tuesday circa 21:30h. M-Lab: 100.99 Mbps, Ookla: 101.6 Mbps
  5. Wednesday circa 06:50h. Ookla: 101.5 Mbps, M-Lab: 101.48 Mbps
  6. Wednesday circa 20:50h. M-Lab: 100.82 Mbps, Ookla: 101.0 Mbps
  7. Thursday circa 07:50h. Ookla: 101.6 Mbps, M-Lab: 101.13 Mbps
  8. Thursday circa 19:10h. M-Lab: 100.85 Mbps, Ookla: 101.7 Mbps
  9. Friday circa 17:00h. Ookla: 101.7 Mbps, M-Lab: 101.31 Mbps
  10. Friday circa 20.00h. M-Lab: 100.04 Mbps, Ookla: 101.1 Mbps
  11. Saturday circa 09:00h. Ookla: 101.6 Mbps, M-Lab: 100.86 Mbps
  12. Saturday circa 11:10h. M-Lab: 100.86 Mbps, Ookla: 101.7 Mbps
  13. Sunday circa 08:30h. Ookla: 101.7 Mbps, M-Lab: 101.41 Mbps
  14. Sunday circa 14:15h. M-Lab: 101.22 Mbps, Ookla: 101.8 Mbps
 

Conclusions

The following key figures apply:

Ziggo's Ookla Speedtest clone

  1. Median 101.6
  2. Average 101.5
  3. Standard deviation 0.2

M-Lab

  1. Median 101.2
  2. Average 101.1
  3. Standard deviation 0.4

The Ookla Speedtest clone and M-Lab give similar results. Although the median of Ookla is slightly higher than the median of M-Lab.

This is consistent with the findings of the research article A Comparative Analysis of Ookla Speedtest and Measurement Labs Network Diagnostic Test (NDT7). This article concludes that (...) Our results show that Ookla and NDT7 report similar speeds under most in-lab conditions, with the exception of networks that experience high latency, where Ookla consistently reports higher throughput (...).

The advertised speed is 100 Mbps. Hence we expect that 100.8 Mbps is the value that should be measured. You could say that M-Lab is slightly better than the Ookla clone.

Compliments to Ziggo

All measurements are above the 100 Mbps. So Ziggo (our ISP) delivers a consistent internet speed. Hence compliments to Ziggo.