I think having a specific deadline is a good idea, since it decreases the bias in whether I will report anything. I will report my results in the early February rationality diary, which will be ~4 weeks worth of data.

As for a definition of victory, I don't think victory is a boolean variable here; keeping the commitment 7 days a week is better than 6 is better than 5..., and I don't expect there to be any threshold or satiation effects worth caring about.

Also, I intend to report my results regardless of what they are. Failures provide useful information, and selective reporting makes reported result less trustworthy.