The oracle is telling you which of the reference populations this calculator (MDLP World-22) uses are the closest match to you; the closest matches would have admixture results closest to what you have.
Single population is which reference population oracle thinks you’re closest to; in this case that’s Luiseño, though at a distance of 6.2 it’s only an okay (not great) match. Mixed Mode guesses a combination of two populations, and the percentage of each that would make a combo closest to what your results are.
Oracle 2, 3, and 4 are similar to Mixed Mode, but have to have perfect splits; Oracle 2 guesses two populations at 50% each, 3 guesses three at 33% each, and 4 guesses four at 25% each. Oracle-4 is giving you the ‘best’ match at a distance of 2.07, and that best match is 25% Athabaskan, 50% Karitiana, and 25% Pima. That means that, of all their reference samples, that combination best matches your particular admixture results.
It’s important to note that this doesn’t necessarily mean that any (or even all) of the results are actually part of your ancestry. They certainly could be, but they could just happen to mimic your actual ancestry the best of the all the options in World-22.
Read more here: Source link