Give the players a continuous update of the progress of the ensemble algorithm, so that they know how well it is doing compared to their own strategies (and the other bots). This should continue even after the paper about it is submitted and published.
Freywa
oh I thought this was going to be a suggestion of a feedback mechanism for the ensemble algorithm.
The reporting to us (as players) sounds good (if perhaps time consuming unless automated) but I think I like the idea I thought you were suggesting - so here’s that idea.
If all the strategies that contribute to the ensemble algorithm are seperate “strategy modules”.
The various strategies have been ‘marked’ (here http://eterna.cmu.edu/htmls/strategy_…) for how successful they were after each actual lab result.
I thought you were then suggesting weighting the ‘contribution’ of each “strategy module” based on those ‘ordering’ marks - including a continuous update of those ratings based on any fresh lab results.