One Algo to rule them All

There is a concerning trend at (least for me) where people enter the Finals not with just their best algo,
but with a SET of 6 … and if we go on like this the “the best algo” will be determined by luck, statistic and bluffs.

I feel personally responsible for this trend as this was made publicly available with the cross_check tool and its variations.
The inspiration for the tool was indeed to test the performance of different strategies,
but in the algo creation phase and the incorporate the statistically better variations in your final algo.

I personally feel we should limit the participation in the Finals to just one algo per player.
I even thing if switching (even for good reasons) is forbidden, it have better results for all:
Submitting a last minute quick change with out testing it publicly for at least a day … is not wise and not healthy!

What do you think ?

  • All 6 algoes should be eligible for the Finals
  • Limited number of algos should be eligible for the finals (2-3)
  • Just one algo, with option to switch if good reason is provided in short timeframe ( 24h)
  • One Algo and no compromises

0 voters

I’m conflicted because I enjoy the metagame of picking the best algo against particular opponents, but I can see how it is frustrating to test an algo all year then lose to the strategy of uploading 6 variations (or more with a team) on the last day and seeing what sticks, with no opportunity to be adapted against. Limiting it to one would make that strategy a lot riskier as you wouldn’t know which one would perform the best before picking it, and would encourage more testing throughout the year.

4 Likes

Please disregard my vote on the poll, it was just the easiest way for me to view the results. I will ensure we take the results into consideration for next season. There has been discussion about moving the transition period into its own competition, which would also naturally reset ratings without adjusting the main leaderboard and force a single algo to be used.

1 Like