Arkham creates an Elo ranking for cryptocurrency prediction traders

Arkham, a company focused on blockchain analysis, introduced a new classification system inspired by the Elo model, focused on evaluating the performance of traders operating in cryptocurrency-related prediction markets.

The proposal seeks to measure the accuracy of predictions

made by participants, creating a ranking based on how correct they are in relation to the implied probabilities of each event. In practice, correct predictions raise the trader's score, while mistakes negatively impact their position.

One of the model's distinguishing features lies in the weight assigned to each correct prediction. Predictions considered less likely by the market generate larger gains in score, rewarding those who identify opportunities outside the prevailing consensus.

On the other hand, expected results or those already widely priced in offer more limited advancement in the ranking. This mechanism seeks to highlight traders who demonstrate a sharper reading of less obvious scenarios within the prediction market.

Another relevant point is that the system does not consider the size of the bets placed. This allows smaller participants, but with a high hit rate, to appear above traders with greater capital, even if the latter have higher absolute profits.

According to Arkham, the objective is to separate performance based on accuracy from other factors, such as trading volume or total financial return. The initiative seeks to offer an alternative metric for evaluating consistency over time.

Despite this, the model also has limitations. The classification does not take into account aspects such as risk management, capital allocation, or accumulated losses in unsuccessful trades.

In practice, this means that a trader may have a high score even without necessarily presenting the best long-term financial strategy. Even so, the system emerges as an additional tool for tracking the behavior of participants in prediction markets within the cryptocurrency segment.