After allowing just one hit and recording thirteen strikeouts against the Giants, Clayton Kershaw allowed six runs in only 4.1 innings against the Astros. What was the difference in pitching a gem and getting pounded? Did he throw softer, not get his pitches to break, have bad location, or was he just a victim of bad luck? Let's take a look.
Here's a chart showing the break on each of Kershaw's pitches in inches. The chart shows pitches from his last start in Houston and the start before against the Giants. 0 represents a ball with no spin, so fastballs are above 0 on the y-axis because they have backspin, causing them to rise. Curveballs are below 0 on the y-axis because their topspin causes them to drop. The viewpoint is from behind homeplate.
Clayton's fastballs and changeups in Houston didn't have as much vertical rise, meaning they didn't have as much backspin on them as in San Francisco. I don't think this would hurt his pitching though. His curveballs actually had more break against the Astros and he threw a higher precentage of them for strikes (53% compared to 35%). So movement wasn't the problem. One other note, you may notice that Kershaw used his changeup a lot more against the Astros. He used it 11 times compared to just 5 against the Giants. It doesn't seem to be a big factor, although the Astros did have a single and a double off of his change up.
Here's a graph showing the velocity of Kershaw's fastballs. The pitches are in order of when they are are thrown so the points on the left are at the start of his outing and those on the right are at the end of his outing. You can see a dropoff in velocity at the end of his start in San Francisco, so he was probably fatigued then. Otherwse there seems to be a gradual drop in velocity, which is normal. 
His average fastball velocity against the Giants was 95.3 MPH and against the Astros it was 93.9 MPH. He threw slightly harder against the Astros but still didn't generate a similiar amount of swings and misses. Chalk that up to the Astros offense. He also threw a higher percentage of fastballs for strikes in Houston, 69% compared to 64.5%.
Here's a chart showing his curveball velocities in the two starts. He threw at basically the exact same velocity on curveballs in each start. Not much to take from this, it seems from the data available the Kershaw's curveball was at least as good if not better in Houston, but he just didn't strike guys out with it. His only strikeout only a curveball was when he got Lance Berkman to swing through a curveball in the dirt.
This is a graph of Kershaw's release point in the two starts:
It seems he was releasing the ball just a bit lower against the GIants. I doubt this had any serious effect but it is interesting that there was a difference between the two starts. These last two charts show the final pitch of each at-bat and what the result was. The first chart shows his start against the Giants and the second chart shows his Houston start.
It looks like the main causes of Kershaw's poor start are just facing a better offense and having some bad luck. Pitches the Giants' hitters were unable to handle the Astros' hitters laced for line drives all over the place. The Astros were 7-12 on pitches in the strike zone while the Giants were just 1-14. You can also see that Carlos Lee's home run was actually out of the strike zone.
I wouldn't be too worried about Kershaw's poor performance on Tuesday. He was throwing harder and got more break on his pitches, but Astros' hitters were on their game. He couldn't get swings and misses in two-strike counts and the Astros got lucky on balls in play with a .410 BABIP. Against the Giants his 13 strikeouts kept them from even putting the ball in play much and their only hit was on a home run. As with most things, he's not as bad as he looked against the Astros and he's not as good as he looked against the Giants.