I run my model before every game is played and it’s based off of the same information that goes into the preseason projections. In doing so, I take the starting line up of each team, including that night’s starting pitcher, and estimate how many games that particular configuration of talent would win if it played all 162 games. (For instance, behind Yu Darvish in his last start, the Rangers sported, by my calculations, a 104-win line up.) I averaged the sum of those wins for the Texas’ first 40 games and calculate the Rangers, on average, sent a 96-win team on the field each game. Yet, I only estimated those same players would win 88 games over the course of the season.