I think there are two elements to this answer. To cut to the short answer skip to the bold summary at the bottom...
Firstly, genetic variation exists because of mutation. Genes get mutated every generation, the . Larger populations will have more mutants within them because: more individuals = more nucleotide base pairs (C's G's A's and T's) = more potential sites of mutation. However mutation is not likely to explain the persistence of variation because mutation rates are very low (1 in 100,000 to 1,000,000 gametes have a newly mutated loci at any individual locus) and singleton alleles have only a 50% chance of reproduction (assuming no selection) so are likely to be lost by drift (Falconer & Mackay, Intro to Quantitative Genetics 1996).
You also talk about acne, which is likely to have a large component of environmental variance. Therefore you should remember that not all phenotypic variance is genetic its source and it is highly likely that an individual trait has some degree of environmental variance component. Simplistically:
Phenotypic variance = genotypic variance + environmental variance
So the bigger question is why does variation persist? There are many potential causes of this which continue to be widely debated. Essentially it seems paradoxical because selection should reduce variation as it drives the fixation of all loci to the fittest allele. However, selection is transient, both spatially and temporally, and is not efficient against rare alleles (especially recessive alleles because they are hidden by dominant traits - e.g. disease "carriers"). Another important point is that some mutations will be neutral, therefore remain unaffected by selection.
In the spatial context, this means that selection is not always favouring the same allele in all places a species inhabits. Selection might be different based on the where it is occurring (within a species, traits like fur would be beneficial to populations in cold climates but not to those in warmer climates - here I am assuming that the sole effect of fur is to improve the ability of retaining heat).
Temporally there are also key elements. Principally, over time selection changes. Again sticking with my fur example, climates change. Ice ages come and go bringing with them different selection coefficients for fur growth.
Another variance in selection can be sexually antagonistic selection, where different alleles are favoured in either sex. In this case selection does not deplete variation but instead maintains it. It has recently been shown that sexual antagonism is prevalent throughout the genome.
the divergent reproductive strategies of the sexes could promote the
maintenance of sexually-antagonistic variation (Sharp & Agrawal 2012... yesterday!)
Other hypotheses suggest mechanisms by with selection can maintain variation such as assortative mating.
Long story short, you stated that you expect variation to reduce as a consequence of selection. However genetic variation persists for many reasons, and can even be maintained by selection in several ways. Furthermore, phenotypic variation which is what you actually describe with your acne example (and I with my fur example) can be caused by non-genetic components of variation.
Cox & Calsbeek 2009, Sexually Antagonistic Selection, Sexual Dimorphism, and the
Resolution of Intralocus Sexual Conflict.
Falconer & Mackay 1996, Introduction to Quantitative Genetics.
Singh & Krimbas 2000, Evolutionary Genetics: from molecules to morphology.
Sharp & Agrawal 2012 (in press, accepted on-line version released yesterday, print may be 2013) Male-biased fitness effects of spontaneous mutations in Drosophila melanogaster, Evolution.
Innocenti & Morrow 2011, The Sexually Antagonistic Genes of Drosophila melanogaster, PLoS Biology.
Arnqvist 2011 Assortative mating by fitness and sexually antagonistic genetic variation, Evolution. (also see his book sexual conflict).