The design of efficacious malaria vaccines is hindered by the antigen diversity contributing to immune escape. To identify specific antigens and their polymorphisms driving immune escape, we developed a novel population genomic model that quantifies allelic turnover within a host. The approach included massively parallel Nanopore long read targeted sequencing of Plasmodium falciparum antigen genes from clinical and asymptomatic infections of two longitudinal paediatric cohorts from Papua New Guinea. Genetic diversity was characterised in a total of 34 genes in 2-4 consecutive infections for each of 240 children resulting in sequence data from 464 P. falciparum isolates. To identify immune escape polymorphisms, we applied a stringent variant calling pipeline, and a novel method to compare significant differences in the turnover rate of variant alleles within-hosts compared to that in the general parasite population. Positive hits at known immune escape polymorphisms in apical membrane antigen 1 (ama1) and concordance with population genetic, biochemical, and structural predictions were used to validate our model. Overall, blood-stage antigens had a higher proportion of immune escape polymorphisms (an average of 30%) than antigens expressed in other lifecycle stages (an average of 1%). Filtering the PNG sequence dataset for these polymorphisms allowed a 30-fold reduction of the diversity for each antigen gene converting an average of 124 (range: 6 - 485) genotypes to 4 (range: 1-8) predicted serotypes. Layering this information onto global parasite genomic data to identify the most common serotypes in natural parasite populations will allow recommendations for the formulation of multivalent P. falciparum vaccines that may overcome the limited vaccine efficacy associated with high diversity. This provides a framework for improving malaria vaccine design and providing a deeper understanding of infection dynamics and immune escape in malaria.