I think the issues people tend to run into most with evaluating new cards is that they're evaluating them overwhelmingly against existing cards, trying to slot them into existing decks, and/or assuming that the meta of the new set is largely the same as it is now.
I feel it makes way more sense to do something like what Kibler does and try to assess each card as:
1.) What decks (existing or hypothetical) does this fit into?
2.) Assuming that deck works, would this be a good card in that deck?
Then, after the full set is revealed, it makes sense to go back and construct hypothetical decks based on your earlier assessments, and then try to guess whether or not the decks would actually be good. More specifically, I think it makes sense to just start listing out a bunch of strengths and shortcomings with the list you created, since it will both reveal what types of metas the deck would actually work in, and also guide you to making improvements.
This also helps tell what decks might be good in the future. A lot of the current core cards from Mid-Range Shaman were introduced during TGT, but Shaman was not an amazing class at that time by any means. However, shoving the whole thing together, you could see the potential with a few more cards. The same thing happened with Dragon decks when people identified that what they most needed was some kind of strong Dragon related heal or taunt around turn 3 or 4, and Twilight Guardian ended up providing that.
There are a lot of cards that are going to seem bad in isolation, or with only one or two synergy cards revealed, but I don't think it makes sense to do day-by-day reveal evaluations assuming that linchpin cards will never exist.
I feel it makes way more sense to do something like what Kibler does and try to assess each card as:
1.) What decks (existing or hypothetical) does this fit into?
2.) Assuming that deck works, would this be a good card in that deck?
Then, after the full set is revealed, it makes sense to go back and construct hypothetical decks based on your earlier assessments, and then try to guess whether or not the decks would actually be good. More specifically, I think it makes sense to just start listing out a bunch of strengths and shortcomings with the list you created, since it will both reveal what types of metas the deck would actually work in, and also guide you to making improvements.
This also helps tell what decks might be good in the future. A lot of the current core cards from Mid-Range Shaman were introduced during TGT, but Shaman was not an amazing class at that time by any means. However, shoving the whole thing together, you could see the potential with a few more cards. The same thing happened with Dragon decks when people identified that what they most needed was some kind of strong Dragon related heal or taunt around turn 3 or 4, and Twilight Guardian ended up providing that.
There are a lot of cards that are going to seem bad in isolation, or with only one or two synergy cards revealed, but I don't think it makes sense to do day-by-day reveal evaluations assuming that linchpin cards will never exist.