Stability metrics assess the reliability of video playback by measuring interruptions during streaming. A higher Stability Score indicates fewer disruptions, leading to a smoother viewing experience, while a lower score reflects increased interruptions.
This metric is calculated using both the frequency and duration of buffering events, providing a nuanced evaluation of playback stability. Some of the key components are defined below.

The Stability Score evaluates the level of interruptions experienced during video playback. A higher score indicates a more stable viewing experience, while a lower score signifies increased disruptions.
This calculation considers both the frequency (buffer count) and duration (buffer percentage) of interruptions, providing a nuanced evaluation. Early Exits Before Video Start (EBVS) views are excluded from scoring to ensure accuracy.
Importance
The Stability Score is crucial for understanding the reliability of video playback. A high score indicates fewer interruptions, leading to a better viewer experience. Monitoring this metric helps identify issues that may disrupt content delivery, allowing platforms to optimize performance and enhance viewer satisfaction.
How to use Stability Score
API usage
Use the List overall values API to retrieve overall values for Stability Score. You need to pass stability_score as the metricId along with other query parameters to receive the response.
The Buffer Ratio quantifies the extent of buffering experienced across the platform.
Buffering Duration represents the cumulative time viewers spend buffering, while Total Viewing Time encompasses all time spent engaged with video content.
Importance
Understanding the Buffer Ratio is vital for assessing how much of the viewing experience is affected by buffering. A high ratio indicates that a significant portion of viewing time is spent waiting for content to load, which can lead to viewer frustration.
How to use Buffer Ratio:
API usage
Use the List overall values API to retrieve overall values for Buffer Ratio. You need to pass buffer_ratio as the metricId along with other query parameters to receive the response.
Buffer Frequency measures how often rebuffering events occur during video playback. This metric provides insights into instances of video stuttering and can indicate potential issues with playback stability.
Importance
Monitoring Buffer Frequency is crucial because frequent interruptions, even if brief, can be as frustrating for viewers as longer buffering events. Understanding this metric helps identify patterns that may affect user experience.
How to use Buffer Frequency
API usage
Use the List overall values API to retrieve overall values for Buffer Frequency. You need to pass buffer_frequency as the metricId along with other query parameters to receive the response.
Buffer Fill measures the duration, in seconds, that viewers experience rebuffering per video view. This metric provides a detailed perspective on how long viewers are waiting during rebuffering events.
Importance
While longer videos may have more opportunities for rebuffering events, Buffer Fill offers valuable insights into the viewer experience by quantifying delays in seconds rather than percentages. This nuanced perspective aids in understanding the impact of rebuffering on overall satisfaction.
How to use Buffer Fill metric
API usage
Use the List overall values API to retrieve overall values for Buffer Fill. You need to pass buffer_fill as the metricId along with other query parameters to receive the response.
Buffer Count quantifies the number of rebuffering events that occur during video playback. This metric provides insights into how frequently rebuffering interrupts the viewing experience.
Importance
Understanding Buffer Count is essential because it highlights how often viewers encounter interruptions due to buffering. Frequent occurrences can lead to viewer frustration and abandonment of content.
How to use Buffer Count
API usage
Use the List overall values API to retrieve overall values for Buffer Count. You need to pass buffer_count as the metricId along with other query parameters to receive the response.