I haven't written roster selection strategies for the past 2 years, but not because I'm keeping any insights or strategies to myself.
Last year I was just too busy with work.
This year I didn't have anything to add to the conversation, other than I still believe that deadball pitchers are the better option in the 90M theme. The modern pitching teams will make less errors but they'll give up more HR's. A bit of a wash, really. It really boils down to my opinion that, due to cost efficiencies, I'd rather have Cicotte, Bernhard and Ford starting against, say, Halladay, Keuchel, Carpenter and Kluber. I mean, if I could afford a premium Kershaw and Pedro at this cap, that would be a different story. However that's not really feasible, unless you want a crap offence.
The other thing I've noticed this year is quite a bit of parity. A lot the team stats for Batting Ave, OBP, WHIP, etc are pretty damn similar. So, LUCK will be a major factor. Things like number of extra inning games & resulting fatigue, strength of division/schedule, teams underperforming vs their pythag. win expectancy and just plain flip of the coin luck between evenly matched teams. I'm not a big believer that poor records in one run games are necessarily an indicator of bad luck; I'm not sure of the cause/effect relationship, but I've noticed that mediocre and bad teams generally have crappier records in one run games. Last year, I constantly cursed my bad luck in one run games, but at the end of the year I had to admit that my teams just weren't very good.