After Friday's 5-1 victory over Boston, a reporter asked Cole Hamels to explain what has changed. The talented but troubled lefthander has pitched well ever since April 23, a start in Arizona that left Hamels sullen.

"You mean home run derby?" Hamels said, smiling.

Yes, home run derby. Hamels allowed four home runs in the span of seven batters and simply looked lost. After the game, he looked more frustrated with himself than ever before in 2010.

Well, how about this: In the month of May, Cole Hamels has a 2.36 ERA in 26 2/3 innings. He is 3-0 in four starts. He has 24 strikeouts, 10 walks and has allowed just three home runs.

And if you take away one really, really bad inning against Atlanta on May 9 (I know, you can always say this), Hamels looks even more impressive. The numbers go down to a 1.40 ERA in 25 2/3 innings.

So what changed?

"I try not to throw any pitches down the middle," Hamels said, coyly.

No, but really...

"From Arizona, I learned a lot," he said. "Most of the home runs I gave up were cutters down the middle. So I've learned how to adjust and throw that pitch."

The Pitch f/x data from Major League Baseball suggests Hamels has thrown more of his secondary pitches in May -- with a heavier reliance on curveballs than in the first month of the season. Granted, Pitch f/x data can be shaky, especially when it comes to identifying cutters because it has fastball movement and velocity.

In his first four starts, Hamels threw his cutter 8.0 percent of the time and his curveball 6.5 percent of the time.

In his last five starts, Hamels has thrown the curveball 10.2 percent of the time and his cutter 8.5 percent of the time. That means less of the fastball and change-up combo.

The cutter is a pitch Hamels began throwing competitively in spring training this year. It's an easier grip to master than most, pitching coach Rich Dubee said, and that's why Hamels adopted it as his fourth pitch.

But Dubee didn't want to stress too much reliance on the cutter. Hamels' curveball was still a viable pitch and he didn't want the pitcher to lose sight of that. Early in the season, it looked as if that's what was happening.

Now, it could be a combination of this: Hamels is using the curveball more while using the cutter in smarter spots. He's locating it far better and it truly has become a decent secondary option for him to use.

Already, Hamels looks closer to the 2008 version than the 2009 one.

If we use nine starts as the reference point, take a look at Hamels over the last three seasons:

2010: 9 GS, 57 1/3 IP, 60 H, 20 BB, 60 K, 3.92 ERA
2009: 9 GS, 48 1/3 IP, 58 H, 10 BB, 50 K, 5.21 ERA
2008: 9 GS, 65 1/3 IP, 49 H, 19 BB, 54 K, 2.89 ERA

His WHIP is almost exactly the same from 2009 (1.41) to 2010 (1.40). Here's one telling stat: Hamels' line drive percentage is down through nine starts. In 2010 it is 18 percent compared to 24 percent in 2009.

And with adding a new pitch, no one expected Hamels to figure out how to use all four pitches the right way immediately. It will continue to be something he tinkers with.

In May, it has worked.