**Coverage**: The more times each position in the genome is sequenced, the more confident the researcher can be in the base call at that position. At 7× coverage, some positions are not covered at all, some are covered once (meaning any error or contaminating read in that one coverage goes undetected), and the aggregate analysis has more uncertainty than at 15× coverage, where most positions have been read multiple times.
**Reference panels**: Population genetic analysis — assigning ancestry fractions, finding closest living relatives, modeling admixture — requires comparing the ancient individual to a dataset of reference genomes. In 2012, the number of ancient European genomes available for comparison was small. By 2023, tens of thousands of ancient European genomes had been sequenced. A larger and better-characterized reference panel means more precise ancestry estimates and better ability to distinguish signal from noise.
Each of these factors improved between 2012 and 2023. The result was not that the 2012 scientists made mistakes; it was that the 2023 scientists had better tools applied to better data, and the conclusions updated accordingly.
—
## Reading Ötzi’s Face: What Genomics Can and Cannot Do
The attention paid to what Ötzi “really looked like” reflects a particular kind of human interest in historical individuals: the desire to see them as people, not just data sources. The proliferation of facial reconstructions, television documentaries featuring actors representing prehistoric figures, and the sustained popular fascination with what any given ancient individual might have looked like reflects this impulse.
Genomics can contribute to answering this question, but its contributions are probabilistic and partial.