When Jeff Luhnow used the first pick of the Rule 5 draft on Josh Fields, I had high hopes for the reliever.
Jose Veras was signed to be the closer for this season, but everyone knew that was not a permanent solution. Veras did have success to start the season saving 19 games for the Astros prior to being traded to the Tigers for outfield prospect Danry Vasquez. It then took the Astros over a month to find his permanent replacement in Fields.
Now for starters, Fields only has five saves on the season and his ERA still sits at 5.56 so it still might be a little premature to anoint him closer for 2014 and perhaps beyond. But what we do know that is clear, is that Fields has earned the right to at least be the closer for the rest of 2013 and for the start of 2014.
This is also assuming that Luhnow does not import a veteran closer in the offseason. That is something I don’t think is likely, although it appears that the bullpen will be much improved for next season. Although, you do reserve the right to be skeptical until it happens.
After landing on the disabled list after pitching only .2 innings on the season, Fields returned in June and had his share of struggles. He did save two games in August, but that was really attributable to the fact that the Astros bullpen was just a revolving door. Fields’ June ERA was 4.66 (in 9.2 innings), followed by a 9.o0 July ERA (in 7 innings), and then a 6.10 ERA in August (10.1 innings).
September though, has been a different story for Fields. Heading into Friday’s game, the closer is three of three in save opportunities with a 2.84 ERA allowing two runs on one hit and five walks while striking out five batters.
Walks have continued to be a problem for Fields, but he is still a young pitcher who is learning, so at least he deserves a shot here. The fact that he has struck out 9.26 batters per every nine innings so far this season certainly does not hurt either.
I have a hard time blindly giving Fields the closer job based just on three saves, but so far he is the most deserving candidate.