I said last time that my new paper on Better ways to evaluate research and researchers proposes a family of Less Wrong Metrics, or LWMs for short, which I think would at least be an improvement on the present ubiquitous use of impact factors and H-indexes. What is an LWM?