Evaluators of education initiatives commonly report results as standardized effect sizes. While this eases comparability across studies, educators and policymakers struggle to interpret the practical importance of results on this scale. This stimulates demand for translating standardized effects to more readily interpretable metrics, such as the number of years of learning necessary to induce the effect. However, the years of learning translation has serious drawbacks and may not be the best choice. We examine four options for translating effect sizes: converting to years of learning, benchmarking against other effect sizes, converting to percentile growth, and estimating the probability of scoring above a proficiency threshold. We describe how each of these perform on a set of desirable properties and use a small survey of researchers to rank them. Years of learning ranks lowest while percentile growth ranks highest and is our recommended translation option.