Percent change of each idealized search strategy for unique contacts compared to the empirical search strategy across 3 different target distributions.

Table entries are percent change in median search efficiency from observed ± 95% confidence interval. Two p-values are shown: the first indicates the significance of the change in median efficiency between the observed and idealized runs (N = 10 runs, each run consists of 4,100 samples, Fig 3B). The second p-value tests whether all raw efficiency values differ between observed and idealized runs (N = 41,000, Fig 3A). All p-values are calculated using the Mann-Whitney U test. The values in parentheses are the Hopkins aggregation statistic. All search strategies are statistically different from observations except LogMCRW in the most diffuse 40 μm DC clusters (in bold).