Assessing avalanche risk in the backcountry is a challenging task. First and foremost since avalanche hazard cannot be accurately determined: We do not know snow stability in space and time with sufficient accuracy. Therefore, it is not surprising that many approaches have been suggested for assessing avalanche hazard when travelling in the backcountry – ever since the first educational avalanche book has been published. Given the uncertainty due to the insufficiently known stability of the snowpack, most approaches rely on experience, comparing contributory factors of avalanche formation to avalanche occurrence. They were developed from rules of thumb to quantitative statistical analyses, to the present machine learning models. For backcountry travelers, the era of the probabilistic approaches started with the advent of Munter’s reduction method – and variants thereof derived subsequently. These methods were promoted in contrast to what was termed the analytical approach which essentially involves assessing and weighing the contributory factors to avalanche formation. Obviously, these terms – probabilistic and analytical – are ill-posed, since any approach can only describe the hazard with a relative probability. We suggest leaving the dichotomy behind, describing the approaches as knowledge-based and rule-based, and moving forward toward a comprehensive risk assessment. The latter implies combining estimates of release probability and consequences, taking into account adequate risk reduction measures, and finally assessing the risk. Such a primarily knowledge-based approach considers simple and relevant observations for both estimating the release probability and the consequences, and builds on our present understanding of avalanched release.