From this blog comparing IOI 2026's performances to CF ratings, we find that the correlation has dropped significantly, e.g., when comparing to this blog comparing IOI 2025's. Also, if we see the IOI 2026 rankings, we see that Team China performed not "as well as" before (still Rank 1, orz).

These phenomena all link to one mostly recognized opinion, which is that this year's problems are very ad-hoc, and non-traditional. In my opinion, this trend will gradually continue over the years, the reason being the claim that AI will gradually replace people who only know "algorithms" but have less thinking ability.

Not just IOI, but other contests, such as APIO and (China) NOI, are following the same trend (I feel this, being a participant in both competitions).

And that is why I think Codeforces contests will, or should, also follow this trend. This also benefits people aiming for various global competitions. Currently on Codeforces, the biggest difference with IOI is the non-existence of "Subtasks" or "Partial Scores". Some may argue that it is hard to set ad-hoc problems without subtasks guiding, which I think is true to some extent, but I think problem setters can definitely give it a try.

Do you agree? Perhaps write your thoughts in your comments.