Dear organizers and moderators,

I am writing this message in a constructive spirit, before the final disclosure of the winners, to share my perspective on the competition and on the use of public leaderboard feedback.

First of all, I would like to thank you for organizing the competition. Despite the difficulties that occurred during the process, including cyberattacks, leaderboard instability, and large-scale probing attempts, I still consider this a valuable learning experience.

During the competition, I made several attempts to reach a competitive score through standard approaches: model development, validation strategies, feature engineering, and regular experimentation. However, at some point it became increasingly difficult to reconcile the highest leaderboard scores with approaches that did not make use of the information provided by the public leaderboard.

The use of leaderboard feedback was also not an isolated or hidden issue. It appeared to be a visible part of the competition dynamics, and it was discussed during the competition. Since this behavior was not explicitly prohibited in the rules, and since no clarification or warning was issued while the competition was still ongoing, I interpreted the available public feedback as something that could be used within the ordinary submission constraints.

I fully understand that this is a delicate matter. Leaderboard feedback can create problematic incentives, and it can affect the fairness and reliability of a competition. At the same time, I believe that these concerns should be evaluated in light of the rules and guidance that were actually available to participants during the competition, rather than through criteria introduced only after the deadline.

My account finished in third place, with a score significantly higher than many other submissions. I believe that, in a competition primarily based on leaderboard performance, this result should be given substantial weight. Additional criteria such as code quality, methodology, reproducibility, and documentation are certainly relevant, especially when comparing submissions with very similar scores. However, when there is a meaningful score difference, I think the leaderboard result should remain a central part of the final evaluation.

My intention is not to dispute the organizers’ role, but to ask that the final decision take into account the full context: the wording of the rules, the lack of an explicit prohibition during the competition, the visible use of leaderboard feedback by participants, and the actual ranking achieved.

Finally, could you please let us know whether there is an expected date for the winner disclosure?

Thank you for your work and for considering this perspective.