July 30, 2025
In rock climbing, grades (the difficulty of a climb) are a big part of the status of a climber. In theory, grades are determined by consensus. The first person to climb a route proposes a difficulty grade, and those who repeat the climb can either accept it or suggest an upgrade or downgrade. The more people climbing a route, the more established the grade becomes.
This system works well at a local level. Grades are negotiated within the local community of climbers, which means that different communities reach different consensuses. This is why a V5 in Fontainebleau (France's most historical bouldering destination) will be similar in difficulty to other V5 climbs in Fontainebleau but harder than one in Albarracín (a bouldering spot in Spain). Thus, a seemingly universal classification system (some countries use different scales, but there are conversion tables) is in practice highly context-dependent. Understanding climbing difficulty requires some familiarity with the local climbing culture.
For amateurs, the different standards result in jokes, whining, and, at worst, snarky comments about climbers from the areas with softer grades. However, for the global climbing elite, context-dependent grades are a real issue.
In professional climbing, the stakes are higher. Athletes compete for recognition and sponsors. Competition does not only happen in terms of difficulty. There are multiple logics of success in climbing. Alex Honnold, perhaps the world's most famous climber, does not climb much harder than some of the local climbers here in Durham, but none of them are anywhere close to making a living out of climbing. This happens because Honnold made his fame by risking his life in free solo ascents (climbing with no rope) and speed records. Nevertheless, climbing hard grades is still the most direct path to global recognition.
So we have a grading scale that is only understandable at the local level, but which determines how rewards are distributed among a global climbing elite. The system is set up for conflict. Great climbers such as Fred Rouhling or James Pearson have fallen in disgrace for proposing grades for their climbs that other climbers thought were too high.
Although the community eventually reaches some consensus, this system is distasteful. To work, it requires that climbers call each other out in public. The elite climbing community is small, and nobody wants to hurt the reputation and career of a friend.
What is the alternative? An algorithm. The latest trend in climbing grades is Darth Grader, an app that uses an formula to determine the difficulty of a climb. It works by dividing up the climb into sections, assigning a grade to each, and then using a proprietary formula to determine the overall grade of the climb. But this still requires the climber to evaluate each section, meaning that the amount of subjectivity, if such a thing exists, is either the same or higher than in the previous system. What the algorithm does is hide the subjectivity of the score behind a formula. The result is an air of objectivity, which will allow lower-status climbers to justify their grade proposal and lower the risk of being shunned.
In terms of how this case fits with other examples of the spread of quantification, this one is a lot like Law school rankings. Both systems are relatively decentralized, so the adoption of this new device happens via horizontal diffusion. This occurs despite most inside actors agreeing that the device is silly, but it is hard to resit because it outsources some of the status allocation process to a set of rules.