In First Time Up: An Insider’s Guide for New Composition Teachers Brock Dethier advocates an holistic model of grading. Dethier poses the question of whether the “errorless bullshitting argument truly deserves more credit than the heartfelt but barely literate personal essay” (68). I would take the latter over the former any day. Lack of sentence level errors is a lot less important than a faulty or poorly supported argument. Dethier also relates that while early in his teaching career he graded grammar mistakes fairly harshly, he now tries to grade on the basis of the “whole interacting mishmash of factors” (68). Although mechanical errors need to be corrected, I want to give students as much credit as possible for effective prose. Finally, Dethier incorporates the concept of “risk” or “degree of difficulty” (69) into his rubrics. I think this is a great idea and plan to employ it in my own teaching. Students should know up front that writing a piece which has few errors, but “says nothing, goes nowhere” (69) will not be graded as highly as more ambitious work, even if it has more sentence level errors. Class discussion of what constitutes a risky paper vs a safe paper can help students as they choose writing topics.
Given your interest in rewarding risk, I think you might want to take a look at Lisa Blansett’s Anti-Rubric Rubric. The way she conceptualizes writing projects seems to provide helpful nuance to the idea of risk.