The second condition, how the season actually plays, is more relevant than the first condition, how the roster is perceived going in. Evaluating who has the best roster going in, of course, can be evaluated by a Fangraphs computer program. But better evaluated by the actual season.
-Veteran guys like Taillon, Imanaga, Steele, they can fall off the table pretty quickly. Fangraphs can have nice big-sample age-decline curves. But the Cubs will have individuals. Lackey, Hendricks, Lester, Q, we've got plenty of experience with proven vets who dropped off pretty sharply.
-The same will apply for whatever Free agent pitcher(s) Hoyer signs, Fried or Peralta or Eovaldi or whomever the one (or two) guys may be. How good Hoyer and Fangraphs thinks they might be may differ significantly from the reality. (Duh, Craig, everybody knows that.).
-Likewise wildly unpredictable what the young guys might do. PCA, Amaya, both might hit a lot better. But, also possible that they may be no better next year than last? Who knows? Shaw, Caissie, Triantos may be productive as opportunity allows, like 2016 Contreras, or 2015 Schwarber/Soler/Russell. Or not, who knows?
-Brown, Horton, Wicks, Assad, Little, Hodge, we might walk into the playoffs thinking we've never had a good, talented young bullpen like this before, ever... or not. Nobody will know in March.
Tangent on that: It would be fun if some unexpected relievers turned up successful. Pearson works on something, and actually ends up being excellent. Roberts comes back another year from surgery, and suddenly his fastball is up 3mph and his fastball command is back, and he's just a good little pitcher? Maybe something eventually clicks with Palencia, and he actually turns into a good reliever? Wesneski? It could be super fun if the young pitchers turn into a dynamite bullpen.