The most common metrics used to evaluate the calibration quality are precision and accuracy. Precision is defined as the ability of the eye tracker to reliably reproduce the same gaze point measurement. Accuracy is defined as the average difference between the real stimuli position and the measured gaze position. The required level of accuracy and precision depends on the nature of the eye tracking study.
Tobii Pro Lab calculates precision and accuracy using the same samples the tracker used for calculating the calibration. The way it works is that the tracker collects raw samples during the calibration and determines an eye model that fits the participant. When this is done, the eye model correction is applied to the raw samples and Tobii Pro Lab uses the corrected samples to determine accuracy and precision. This means, the eye tracker first computes the average position of the samples, and then measures the angle/distance from this point to the stimulus point.
Read more in "Tobii Test Specifications Accuracy and Precision Test Method"