To clarify what this question is likely talking about:
We know that there was a continuous Greek presence in Thrace up to Constantinople, the Pontus (Black Sea), and Cappadocia, after the arrival of the Ottomans.
We know that there was a substantial Greek population in Western Asia Minor in the 19th century, which is linguistically distinct from Thrace, the Pontus, and Cappadocia: the dialect spoken in places like Smyrna was closer to Crete and the Cyclades.
We know that there was movement of Greek populations within the Ottoman Empire; we know for example that Bithynia was resettled from Epirus in the 17th century, and we know that the Tsakonian colony near Erdek/Artaki cannot have been indigenous, and likely dates from the 18th century.
We know that to the Ottomans, Christian and Muslim, ethnicity was not particularly important; religion was. “Greek population” would have been understood by everyone at the time to mean “Christian Orthodox population”.
So at issue is: whether the Christian population in Western Asia Minor disappeared, or remained continuous: whether that population entirely represents resettlement from Greece, or whether a Christian population remained in place in Western Asia Minor in the 15th century.
This is a topic Dimitra Triantafyllidou and I have disagreed on, and unlike me, she has read the mainstream tract outlining the theory of discontinuity, The Decline of Medieval Hellenism in Asia Minor, and found it unconvincing.
If the population was discontinuous, and that’s how I’m interpreting the question, the answer is assimilation. The Greeks of Western Asia Minor weren’t massacred, they were converted to Islam. That there was conversion is known to be true; the debate is really whether there were any Christians left in Western Asia Minor in 1500 or not.
Presumably Thrace and the Pontus held out because they were conquered later, once the Millet system was in place, so there was less social pressure for conversion; and Cappadocia presumably was relatively inaccessible.