those are all mathematically correct models.
for example:
Target: xz1333
Distance: 0.8220% / 0.00821959 | R3P
52.4 Greek_Central_Macedonia
45.0 Italian_Trentino-Alto-Adige
2.6 Georgian_Ajar
Target |
Distance |
ALB_Çinamak_Anc |
Croatia_Imperial |
TUR_Marmara_Ilıpınar_Byz3(Slav) |
Italian_Trentino-Alto-Adige |
0.02279690 |
84.2 |
0.0 |
15.8 |
Greek_Central_Macedonia |
0.01605523 |
34.6 |
38.4 |
27.0 |
Average |
0.01942606 |
59.4 |
19.2 |
21.4 |
Greek_Central_Macedonia are similar to Albanians, but with increased Slavic and Imperial admixtures. Italian_Trentino-Alto-Adige are very similar to IA Balkans, with a bit of Slavic.
When you average those 2, you get results very similar to the average Albanian ones, as you can see in the table.
all modern and ancient Europeans are made of similar components. that's why you should use a specific selection of source samples, and not the full sheet.