Tampa Bay Rays: Solving the Mystery of James Shields

Eli MargerCorrespondent IMarch 24, 2011

ST PETERSBURG, FL - OCTOBER 23:  James Shields #33 of the Tampa Bay Rays gets set to throw a pitch against the Philadelphia Phillies during game two of the 2008 MLB World Series on October 23, 2008 at Tropicana Field in St. Petersburg, Florida. The Rays won 4-2. (Photo by Jamie Squire/Getty Images)
Jamie Squire/Getty Images

Over the course of four full seasons, Tampa Bay Rays pitcher James Shields has gone from a top-flight starter to a homer-prone liability.

Since entering the league in the middle of the 2006 season, Shields has exhibited an elite changeup and a good curveball and slider to complement a serviceable fastball.

But "Big Game James," as he is known to the Rays faithful, has been on an alarming downward spiral, culminating in a 2010 season in which he allowed 34 home runs and ended up with a dismal 5.18 ERA.

So, what happened? Let's start at the peak.

In 2008, Shields was the No. 2 starter on the Rays AL championship team, serving as the tail end of a great one-two punch with Scott Kazmir. Shields went 14-8 with a 3.56 ERA in a stellar season that included two shutouts. After a shaky debut season in 2006, Shields had pitched well in 2007. 2008 was simply the combination of Shields' potential reaching a peak and a chance to play on a team with great defense and run support.

That season, Shields had a nice mix of pitches—a fastball (44.9 percent of his pitches), a cutter (19.2 percent), the changeup (26.3 percent) and a curveball (9.6 percent). All of his pitches had positive values except for his fastball, which was just less than zero at a minus-1.5.

This is not a bad number by any means. It simply shows that Shields' fastball cost the Rays about 1.5 runs over the course of the season.

Though hitters did make contact with his pitches, Shields was able to keep himself out of trouble. He induced ground balls 46.3 percent of the time, still a career high. His line-drive rate, often an indicator of how well hitters make contact, was a career low. His BABIP (batting average on balls in play) was a very good .287.

His control was very impressive in 2008, especially when it came to painting the edge of the strike zone. In this heat chart, courtesy of Fangraphs, you can see that a high volume (the red, pink, and white) of fastballs were thrown down the middle and on the outer part of the strike zone.

We can safely assume that most of the fastballs down the middle were either first-pitch fastballs or 3-0 pitches, so the fact that a high volume of pitches still hit the corners is very impressive. Compare this to his 2010 chart, where you see high volumes of fastballs clustered in or around the middle of the strike zone.

When examining the changeup, a similar trend appears. Since we examined his fastball success against righties, let's look at changeups against lefties. In 2008, the chart is incredible. Shields threw his great changeup in exactly the spots you'd like to see—down and away from the left-handed hitter. A select few pitches missed up and towards the middle of the strike zone.

Now, look at 2010. He still did a decent job of locating his changeup, but many more pitches missed in the middle of the zone and up.

The bottom line is this: In 2008, Shields was a control pitcher, combining excellent command with great movement on his pitches. In 2010, for whatever reason, he was missing spots and missing them badly.

The result? Ten more home runs allowed, an ERA 1.5 runs higher and a decrease in WAR from 4.1 to 2.2.

But we're not done yet.

When examining the PitchFX data for Shields, there was a very interesting development between 2008 and 2010. Remember that 44.9 percent fastball rate in 2008? Those were all four-seam fastballs. But in 2010, PitchFX recorded only 28.4 percent four-seamers. This is because a new pitch came into Shields' arsenal—the two-seam fastball, thrown 18.7 percent of the time.

Here is the chart of this pitch against lefties. You will notice that a large volume of two-seam fastballs were thrown basically right down the pipe.

It is very hard to grasp the effects of this new pitch, but the most convincing argument against Shields' two-seamer I could find was his pitch value. As mentioned before, in 2008 Shields had a fastball value of 1.5 runs against the Rays. In 2010, this value plummeted to 24.7 runs, an astronomical figure.

In 2009, a year in which he threw 8.2 percent two-seamers, his fastball value was a horrible 13.1 runs against. And it wasn't just his fastball that killed him in 2010. His cutter registered a value of 9.2 runs against.

In short, this is what it seems like—Shields' four-seam fastball is fine. But his moving fastballs (cutter and two-seamer) kill him. They either don't move enough or cannot be located well enough for them to be effective pitches. Mariano Rivera's cutter is deadly because not only does it break a ridiculous amount, but it can be located wherever Mo wants it to be. Shields' cutter and two-seamer seem to be the polar opposite.

If Shields is to return to form in 2011, he must reverse this trend of throwing more two-seam fastballs. If he has to throw them, he must improve their effectiveness, whether it be through better movement or location.

Shields is not a strikeout pitcher—he learned that the hard way in 2010.

The highest K/9 ratio of his career coincided with his worst season. He is a control pitcher that relies on deception and location, much like a Pedro Martinez or Greg Maddux.

Parents always tell their kids to not be someone they're not. As much as James Shields may want to be David Price, he will always be James Shields. He will always have a killer changeup, a good fastball and a good curveball. He will never be able to blow pitches past hitters without sacrificing command.

If James Shields can figure that out, his 2011 campaign will be much like 2008.

And if that happens, the Rays season in 2011 may turn out much like that magical year.