๐ How Rare Is Tall + Blue Eyes? The Math Explained
๐งฎ The Math
๐บ๏ธ Regional Context
In Netherlands and Scandinavia, where both tall height AND blue eyes are common, this combo reaches 10%+. In East Asia or Sub-Saharan Africa, where both traits are very rare, it's essentially 0%. Global average: ~1.2-1.5% of all men.
๐ Why People Search This
This combination is iconic in media and culture as an "ideal" look, making people curious about its actual statistical rarity. In reality, while not ultra-rare globally, it IS concentrated in specific regions โ making it very rare outside of Northern Europe.
๐ฒ What "assuming independence" means for the combined figure
Multiplying a height share by an eye-colour share treats the two traits as independent. It assumes that knowing a man stands 185 cm or taller says nothing about the odds his eyes are blue. The joint frequency of two traits actually equals the product of their separate frequencies plus their covariance. Worldwide that covariance is positive, for a geographic reason: the populations where 185 cm is easy to clear are largely the same populations where blue eyes are common. When two traits cluster in the same places, the true overlap runs somewhat higher than the plain product, which makes any independence-based product a conservative baseline.
The flip side is that multiplication gets more reliable as the reference group narrows. Within a single ancestry group, most of the population structure behind that covariance drops away, and any remaining link between height and eye colour should be much weaker. The product is a rough guide for "every man alive" and a much better one for "men in Denmark". A ground-up global figure would instead need joint frequencies: the share of men in each population who are both tall and blue-eyed, measured in the same people.
๐ 185 cm meant something different to your grandfather
A cutoff in centimetres is not a fixed percentile: the share of men who clear it shifts with each birth cohort. The NCD Risk Factor Collaboration reanalysed 1,472 population-based studies, with height measurements on more than 18.6 million people born between 1896 and 1996 (eLife, 2016). The tallest group in that century of data was Dutch men born in the last quarter of the 20th century, whose average height surpassed 182.5 cm. Since even that average falls short of 185 cm, clearing the line means standing above the national average anywhere in the world. The same study found that height gains had levelled off for two to three decades in Finland and the UK.
Reported eye colour has shifted even more. Grant and Lauderdale analysed US national health surveys from 1971โ1975 and 1988โ1994 (Annals of Human Biology, 2002). They found blue eyes in 57.4% of non-Hispanic white Americans born between 1899 and 1905, compared with 33.8% of those born between 1936 and 1951. They put the change down to a declining trend in assortative mating by ancestry: what changed was who partnered with whom, not the genetics of eye colour. Both inputs to a tall-and-blue-eyed estimate are generational snapshots, not fixed constants.
๐๏ธ Why blue-eye figures come as a range
Blue-eye prevalence is usually quoted as a band rather than a single number, and one reason is that there is no standard scale for iris colour. Franssen, Coppens and van den Berg noted that most grading systems split eye colour into only two to five categories, which they judged too coarse, and proposed matching irises against a set of 24 graded eye photographs instead (Journal of Optometry, 2008). A scheme with an intermediate class will report a smaller blue share than a two-way blue-or-brown split of the same eyes, because grey and green-tinged eyes have somewhere else to go. Two countries' figures are only comparable if the studies used the same categories.
Genetics explains why that boundary is genuinely blurry. Duffy and colleagues had earlier tied a quantitative-trait locus at the OCA2 region of chromosome 15 to 74% of eye-colour variation, and in 2007 narrowed the signal to a three-variant haplotype in the first intron of OCA2 (American Journal of Human Genetics). A year later, Sturm and colleagues screened the neighbouring HERC2 gene and found that a single variant in its intron 86, rs12913832, predicted eye colour better than that haplotype, with an ordinal regression Rยฒ of 0.68; they concluded the conserved sequence around it is a regulatory region controlling OCA2 expression (AJHG, 2008). Strong prediction is not perfect prediction. The same genotype does not always produce the same reported colour, and the leftover variation lands in the in-between shades that grading schemes treat differently. Any blue-eye prevalence is therefore good only to within a few percentage points, and that uncertainty carries straight through the multiplication.