🔁Virtuous Improvement Cycle

We appreciate the importance of curiosity and motivation for high prediction quality. Therefore, we are aiming to instigate and build upon a virtuous improvement cycle within Solex.ai to ensure the ever-improving price forecast inputs. In particular:

1. Firstly, we only reward a limited percentage of top-performing forecasters. Whilst constantly increasing our Awards Pool, we still ensure that the rate of our forecaster base growth remains higher than the rate of forecaster compensation increase. This makes it harder to reach the reward-paying bracket.

2. Secondly, depending on the forecaster's motivation, the growing competition leads them to either drop out or strive to do more research for better predictions. Although it is sad to lose our precious forecasters, the quality of our products is bound to increase regardless of their choice.

Solex.ai Virtuous Improvement Cycle

To evaluate and score our forecaster’s accuracy, we rely on a platform-tailored formula derived from the mean absolute scaled error (‘MASE’) metric. MASE is a statistical measure of the accuracy of forecasts that displays favorable properties, when compared to other methods, when calculating prediction errors and deviations. This measure is scale invariant (i.e. does not depend on the unit scale of the data), symmetric (positive and negative errors scored equally) and interpretable.

Ultimately, Solex.ai aims to assign awards for precision such that there exists a direct link between the relevant share of the Award Pool and the forecaster's results regardless of which asset they choose to focus on.