All results are probabilistic estimates, not absolute truths. We've designed the report to be scientifically solid and read conservatively, but precision is limited by a few real factors:
Uneven ancient DNA sampling. Some regions and periods are densely sampled (Cretan Bronze Age, Bulgarian Chalcolithic, Croatian Late Antiquity, Serbian Iron Gates) while others are sparse (Cypriot Iron Age, Albanian Neolithic, Bosnian everything, Greek Epirus). Results in thinly sampled zones carry wider uncertainty.
DNA quality variation. Ancient DNA is degraded. Coverage in the dataset ranges from about 31,000 SNPs (our minimum threshold) to over 1.1 million. Lower-coverage individuals give less reliable distance scores, so we down-weight them.
Historical migration and admixture. The Greek and Balkan world has been a migration crossroads for tens of thousands of years. Population boundaries were gradients, not sharp lines, and the same signal can often be explained by more than one historical path.
Statistical noise in clustering. Even with high-quality data, fine-scale ancestry carries uncertainty — especially when telling apart closely related anchors like Serbian and Bulgarian, or North Macedonian and Northern Greek.