Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Version History

Version 1 Next »

This page introduces the calculation method of the Ecoscore, and its evolution introduced across versions.

Steps' score and ecoscore

Each step of a user journey is evaluation with regard to three criteria:

  • Energy: the rate of increase in power usage compared to the idle power usage

  • Time: The duration of the step in second

  • Mobile Data: the amount of data transmitted during the step, in ko

For each of these criteria, the step receives a grade between 1 and 5.

A user journey can contain steps of different types: Loading, action, and pauses. The “time” metric is not assessed for pause steps. Indeed, such duration is defined by the operator.

Each step is assigned a grade computed as the mean of its scores in the three metrics. The ecoscore of a user journey is calculated as the mean of all steps of this journey. For a given metric, the average scores of steps in this specific metric allows for calculating intermediary ecoscores for the journey in this metric. Ecoscores, which regard the entirety of the journey, are expressed over 100, while steps’ scores are expressed on a 1 to 5 scale.

Présentation sans titre(1).png

The ecoscore of a user journey is the mean of its steps’ scores, independently of the number of steps. Thus, it is possible to create a large number of small steps, with short duration and low data usage, to artificially increase the ecoscore.

Ecoscore V1

This first version of the ecoscore defines thresholds for the three metrics. Additional thresholds are defined for Action steps, which can be longer than Loading steps.

5

4

3

2

1

Energy

< x1.05

< x1.5

< x2

< x3

>= x3

Time

< 1

< 3

< 5

< 10

>= 10

Time (action)

< 5

< 10

< 15

< 20

>= 20

Network Data

< 10

< 30

< 50

< 100

>= 100

Ecoscore V2

This second version of the ecoscore intents to tackle limitations in ecoscore V1. First, it introduces new thresholds to better tackle specific situations, and second, some thresholds are redefined to improve the distribution of scores.

  • Specific thresholds are introduces for Pause steps in “Energy” and “Network data”. Such thresholds are structs as Pause steps are not supposed to sollicitate the hardware.

  • Energy steps (outside of Pause steps) are made more permissive

  • Highest Duration thresholds are made more strict

  • The distribution of Network Data is made exponential, with stricter thresholds for best grades and more permissive thresholds for low grades.

  • Specific thresholds are introduces for Network Data on thin clients (web), higher than on rich client (applications).

5

4

3

2

1

Energy

< x2

< x3

< x4

< x5

>= x5

Energy (Pause)

< x1.05

< x1.5

< x2

< x3

>= x3

Time

< 1

< 2

< 5

< 10

>= 10

Time (Action)

< 2

< 4

< 10

< 20

>= 20

Network data

< 5

< 25

< 50

< 300

>= 300

Network data (Web)

< 50

< 250

< 500

< 1000

>= 1000

Network data (Pause)

0

< 5

< 10

< 25

>= 25

Ecoscore V2.1

Ecoscore V2.1 was introduces to improve the distribution of Energy scores.

  • Energy thresholds, made more permissive in V2, turned out to be too permissive, making high ecoscores too reachable.

5

4

3

2

1

Energy

< x1.25

< x1.75

< x2.5

< x3

>= x3

Energy (Pause)

< x1.05

< x1.5

< x2

< x3

>= x3

Time

< 1

< 2

< 5

< 10

>= 10

Time (Action)

< 2

< 4

< 10

< 20

>= 20

Network data

< 5

< 25

< 50

< 300

>= 300

Network data (Web)

< 50

< 250

< 500

< 1000

>= 1000

Network data (Pause)

0

< 5

< 10

< 25

>= 25

  • No labels