The Astros should add Michael Lorenzen.
Where there is smoke, there is a fire, and there seems to be plenty of smoke around Michael Lorenzen and the Astros.
For good reason. The Tiger's All-Star representative hasn't given up an earned run in his last three starts.
His 88th percentile fastball spin is fits the Astros M.O. perfectly. And he likely won't require the type of bidding war the Astros aren't equipped to win.
He's experienced as both a starter and reliever, and his swiss army knife versatility would be massive for Houston to keep the innings off of both starters and relievers alike down the stretch. Houston would be a great home for Lorenzen as the season comes to a close.
And speaking of experience as a starter and reliever...
The Astros should pursue Seth Lugo and Blake Snell and pass on Josh Hader.
The Padres are a trainwreck. Their roster is loaded, yet they sit 49-54 with a locker room that doesn't seem to like each other. Will they sell? Who knows. But they absolutely should. And if they do, the Astros should look for Seth Lugo.
Like Lorenzen, Lugo can start or come out of the pen. His 100th percentile curveball spin has had him linked to Houston forever. This may be the time for the Astros to make their move. He wouldn't be as expensive as Snell or Hader, and his versatility and ability to eat innings at a high level would go a long way towards keeping the Astros fresh down the line.
And Blake Snell needs no introduction. He's having his best year since winning the Cy Young in 2018. He's striking out 11.6 per nine, and has a sub-1.00 ERA since June. He'd immediately give the Astros another ace for October. But he's going to be costly. They should pursue him, but as a rental, he's not worth emptying the tank for.
And speaking of not worth emptying the tank, no matter how good Josh Hader may be, there's just not a world where the Astros should further empty their farm system for a reliever that would walk at the end of the season.
Brown should call San Diego to get Lugo, see if Snell can be had for an even feasible package, and not even ask about Hader.