Is what true exactly? My comment above was strictly about whether Numerai cares [from a technical standpoint] if you submit the same model in the same slot each week. (They don’t.) There is sometimes a misconception that Numerai is tracking your model over weeks or months to tune the metamodel better or something, but like I said they recreate the metamodel each week from scratch with the predictions submitted only for that round, and anything submitted for previous rounds or previous performance of your model (or your ranking on the leaderboard) is not factored in at all. Only your stake amount for that round is a factor in how your model is weighted in the metamodel. [However, the implementation of this actually wasn’t 100% right for the last so many months because for a while they had the stake-change deadline in the middle of the week, but now it has been moved back to be the same as the submission deadline where it should be so they can be sure to have accurate staking amounts when they do their real life trades.]
But anyway, it sounds like you are saying there is benefit TO YOU in sticking with the same model for a while (not worrying about every down round, etc) and not changing up all the time. And that is probably true, at least as long as it is a decent model. And yes, what’s good for you is good for Numerai if you make wise decisions about which models to submit and how much to stake on them. But that wasn’t what I was addressing at all.