Calibrating Flavour Tagging Algorithms using $t\bar{t}$ events with the ATLAS Detector at $\sqrt{s}=13$ TeV

$b$-jets are identified in the ATLAS experiment using a complex multivariate algorithm. In many analyses, the performance of this algorithm in signal and background processes is estimated using Monte Carlo simulated events. As the event and detector simulation may not give a perfect account of real events, and given the large number of changes between Run-1 and Run-2, it is vital to calibrate the performance of this algorithm with data. The $t\bar{t}$ Probability Distribution Function method has been employed to measure the $b$-jet identification efficiency in data using a combinatorial likelihood approach. Results are presented incorporating the first $3.2~\text{fb}^{-1}$ of $pp$ collisions at $\sqrt{s} = 13~\text{TeV}$ collected by the ATLAS detector during Run-2.