Ever since Bill James published his “age-27 peak” theory, the working assumption around baseball has been that the best players are usually those in the middle of their careers. This is largely a safe assumption when it comes to position players – Brian Roberts, the major league leader in batting average, is 27; home-run leader Alex Rodriguez is 29; and the leading MVP candidate is Miguel Tejada, also age 29. But the story is entirely different when it comes to pitchers, who do not follow the same predictable aging patterns that offensive players do.

Indeed, the majority of baseball’s best pitchers are either older than 36 or 26 and under. This year, while many stalwarts of the older generation have been hit hard (Randy Johnson, Greg Maddux, Curt Schilling, Kevin Brown, among others), several young starters have taken their games to a new level. According to Baseball Prospectus’s Value Over Replacement Player (VORP) metric, which measures the runs a player is worth as compared with a freely available replacement, three pitchers age 25 and younger rank among the top six in baseball. Five – Jake Peavy, Brett Myers, Dontrelle Willis, Jon Garland, and Josh Beckett – rank in the top 20. Here’s a look at each of these pitchers, with an emphasis on explaining how they’re managing to achieve their success, and what we can expect from them going forward.

*DONTRELLE WILLIS, MARLINS Willis has spent three seasons in the big leagues and has undergone just as many changes in his approach. In his 2003 debut, he was a strikeout pitcher who sometimes beat himself with walks, but nevertheless was dominant enough over stretches to win the NL Rookie of the Year award. Last season, Willis reverted to being a control pitcher, throwing fewer pitches per inning and walking fewer hitters, but also notching 20% fewer strikeouts and a 70-point jump in his ERA.

This season, we’ve seen still another transformation. Willis, who had a career groundball-to-flyball rate of 1.21 entering the season, has opposing hitters pounding balls into the dirt, using the sinking motion on his two-seamer to generate nearly twice as many groundballs as flyballs. As a result, he’s allowed just two home runs while inducing plenty of double plays with runners on base. His control has remained intact, while his strikeout rate is close to rebounding to its 2003 levels. This combination of power, finesse, and groundball induction is about as lethal as it gets, and Willis should be among the best pitchers in the league for a long time.

*JAKE PEAVY, PADRES Peavy led the majors in ERA last season, an accomplishment that was dismissed as a fluke in some circles because of his injury-shortened season and his favorable home park. But Peavy has been even better this season, having cut his walk rate nearly in half and finding as much success away from Petco Park (1.73 ERA away) as at home (2.27).

Like Maddux, Peavy mixes a run-of-the-mill fastball with hard-breaking pitches, and has come into his own thanks to a marked increase in his control. The single best indicator of success going forward is a pitcher’s strikeout-to-walk ratio, and Peavy’s stellar 6.1:1 figure gives him the highest upside of any pitcher on this list.

* BRETT MYERS, PHILLIES If you want a good working example of the importance of strikeout rate for a young pitcher, look no further. Last year, Myers struck out 5.9 hitters per nine innings and recorded an ERA of 5.52. This year, he’s striking out 9.2 hitters per nine innings and has an ERA of 2.06.

These changes are likely the result of two interrelated factors. First, Myers has a new pitching coach in Rich Dubee, who is instructing him to challenge hitters more frequently early in the count, resulting in more first-pitch strikes and fewer walks. Second, Myers is no longer pitching like he’s scared of his hitter-friendly home ballpark. Last season, his strikeout rate was nearly 25% higher on the road than at Citizens Bank Park. Though Citizens does give up its fair share of home runs, overcompensating by giving up more walks, more base hits, and fewer strikeouts was not a winning proposition.

* JON GARLAND, WHITE SOX If there’s one pitcher here who isn’t like the others, it’s Garland. He’s had his share of success, to be sure: He’s on pace to win 25 games, and is a good bet to be the AL’s starting pitcher in the All-Star Game. The problem is that he’s been doing it with smoke and mirrors.

Very few pitchers can sustain success for any length of time with a strikeout rate as low as Garland’s 4.2 K/9. Garland has been a better pitcher this year – like Myers, he’s benefiting from challenging hitters more, and like Willis, he’s generating more groundballs. But the 4.22 ERA he’s posted thus far in May is far more in line for a pitcher with his peripheral numbers than the 1.80 ERA he had in April.

* JOSH BECKETT, MARLINS It might be something in the salt water that surrounds Dolphins Stadium, but Beckett is following the same route to success as Willis, using an improved splitter to generate far more groundballs than he has at any point previously in his career. Compare the 0.82 GB/FB ratio that Beckett had in his 2002 rookie season to the 1.50 figure that induced 11 double plays this season, third most in the NL.

Can these pitchers maintain their early-career success into their late-20s? It’s worth remembering that it took Johnson until he was 29 to become a dominant pitcher and that Maddux, who later became known for his phenomenal control, was among the league leaders in walks allowed in three of his first four big-league seasons. Pitching is such an unusual craft that a small and seemingly trivial adjustment can produce a momentous difference in results. Each of the pitchers described here has made such an adjustment – and some of them will surely be mentioned alongside Johnson, Brown, and Maddux by future generations.

Mr. Silver writes for Baseball Prospectus. For more stateof-the-art commentary, visit baseballprospectus.com.