Tradeoffs - negative reciprocal causal relationships in net benefits between trait magnitudes - have not always been studied in depth appropriate to their central role in life-history analysis. Here we focus on allocation tradeoffs, in which acquisition of a limiting resource requires allocation of resource to alternative traits. We identify the components of this allocation process and emphasize the importance of quantifying them. We then propose categorizing allocation tradeoffs into linear, concave and convex relationships based on the way that resource allocation yields trait magnitudes under the tradeoff. Linear relationships are over-represented in the literature because of typically small data sets over restricted ranges of trait magnitudes, an emphasis on simple correlation analysis, and a failure to remove variation associated with acquisition of the limiting resource in characterizing the tradeoff. (We provide methods for controlling these acquisition effects.) Non-linear relationships have been documented and are expected under plausible conditions that we summarize. We note ways that shifting environments and biological features yield plasticity of tradeoff graphs. Finally, we illustrate these points using case studies and close with priorities for future work.