the last several posts in this thread highlight the knowledge gaps we have about predicting DCS. Many different algorithms and models for how it works, but in the end we don't truly KNOW. Until we do we have to stay conservative or accept the risk that comes with not being conservative.