When all Rigid bodies are created and there are still unlabeled data segments, this is referred to as a gap.
To fill gaps with unlabeled segments, browse through the list of markers and analyze their movements to see gaps where the data should be. You can fill these gaps with segments by automatically searching for unlabeled segments.
The red crosshair shows the possible data segments that most likely fill the gap, and rebuild the data of the current marker by joining the two data segments. Keep dragging the Timeline indicator back and forth along the gap to visually analyze what unlabeled segment represents the best fit. Select the unlabeled segment that seems the most appropriate.