Before the season began, I prepared a WAR projection for the Astros---and the accuracy of those results followed the Astros' season into shambles. We will try again. Reset. Project. I attempt to project the Astros' Wins Above Replacement (WAR) for the remaining 50 games of the season. I use the same WAR spreadsheet, inputting projected playing time, wOBP, fielding, and baserunning for each position player and innings, ERA, and leverage for each pitcher. But it's not the same this time because the roster has changed a lot since the start of the season.
Projecting performance for the newest rookies is a puzzle. ZIPS had a projection for J.D. Martinez, and I used it. For Altuve, Paredes, and Shuck, I used a major league equivalent (MLE) calculator. Sure, the MLE calculation isn't perfect, but what else do we have? I made educated guesses for the rookies' base running and fielding.
For the other players, I looked at their current stat line so far this year, the pre-season ZIPS forecast, and the updated ZIPS forecast. In most cases, I either assumed a continuation of current performance or I used the ZIPS update. Generally, it doesn't make a lot of difference whether the current season performance or ZIPS is used. That may indicate that ZIPS has been reasonably accurate in its projections of Astros players. The current fielding and base running performance is used as the basis for projecting those performances for the remainder of the year. I also assumed that Aneury Rodriguez will return to the rotation when Lyles is shut down, and guessed that Harrell and Sosa will be called up in September. In addition, I think that Johnson and Wallace will be called back up in September, and I included playing time for them.
The WAR results are summed in order to give us a W/L projection. Based on this projection, the Astros will go 21-29 for the remainder of the season. This would give the Astros a final record of 58-104. Details of the projections follow.
Below are the player WAR projections for the remaining 50 games. The position players will provide a scant 1.36 wins above replacement. The pitching staff fares somewhat better with 5.27 wins above replacement over the next 50 games. Wandy and Bud Norris provide the best WAR among pitchers, and Altuve and Bourgeois provide the best WAR among position players. Of course, when I say this or that will happen, take it with a grain of salt. This projection is built on pile of uncertainty, and projecting a small sample size like 50 games is particularly less reliable.