Sizing up Jordan Lyles' hot start for Rockies

Jordan Lyles was a 2008 supplemental first-round draft pick by the Houston Astros, who were in the midst of a string of franchise-stunting amateur drafts. But the scouting community liked this kid from South Carolina's Hartsville High School. He was "… loaded with projection" and had "… tons of upside based on frame and delivery" noted Baseball Prospectus. Both his fastball and his command were slightly above average, and he worked on his curve and sinker as he rose through the minors. The Astros fast-tracked him, and he quickly ascended the prospect rankings, going from Baseball America's No. 91 prospect prior to 2010 to No. 42 pre-2011 and appearing in the 2010 All-Star Futures Game. As he rose through the system, his K/9 dropped significantly, but his walk rate was steady at just over two per nine and he kept the ball in the park. He got the call to The Show in April 2011, at the tender age of 20.

He bounced between Houston and Triple-A Oklahoma City the next three years. In the majors, he gave up at least 10 hits per nine innings in each season, while his walks rose from 6.3 percent to 7.6 percent of all plate appearances and his strikeout percentage dropped from 16.1 percent to 14.5 percent. His ground ball percentage varied from year to year, but was generally above average (a three-year composite of 47.5 percent). However, Minute Maid Park was doing him no favors when he left the ball up, as he yielded 27 homers in 197.1 innings there from 2011 to 2013. Thus, his ERA ended up north of five each season.

Last December, the Astros parted ways with Lyles. He and teammate Brandon Barnes were sent to Colorado for Dexter Fowler. For a team that had stripped every last piece of discernible major league talent from its roster in a concerted effort to rebuild from the ground up, giving up on a still only 23-year-old former first-round pick was stunning. For Lyles, going to pitching purgatory Denver would likely take the bite off of his off-speed offerings, like it had for so many others. With his fastball only a tick or two above average and some homer-prone tendencies, many wondered what the Rockies saw in him (while also being confused as to why they would also give up the seemingly better, more accomplished center fielder in the same deal).

Flash forward five months, and it wouldn't be a reach to say that Lyles has been the most pleasant pitching surprise in the National League so far. After Monday's 8-2 triumph over the Texas Rangers, he sits at 4-0 with a 2.62 ERA, having yielded only two homers in 44.2 innings while giving up 7.5 hits per nine innings and cutting his walk rate back to two per nine. So what, if anything, has changed, besides the scenery? Well, when a player does get a "change of scenery," he also gets a change of defenders behind him, and in the case of Lyles, that might be making a palpable difference.

The Astros arguably had one of the worst team defenses of all 30 clubs for the three-year period ending in 2013. Here are their yearly defensive efficiency ratings (DERs) and ranks within baseball:

2011: .698 (27th)

2012: .693 (T-26th)

2013: .697 (T-24th)

Ironically, their average 25th ranking over those three years was better than only one other team … the Colorado Rockies. In fact, Colorado finished dead last in DER each of the past two seasons.

More importantly for Lyles, however, is that the Astros' infield defense was also well below average, as measured by BABIP for ground balls only:

2011: .250 (5.5 percent higher than MLB average of .240, ranked 25th)

2012: .252 (plus-5.9 percent, 25th)

2013: .252 (plus-5.0 percent, 23rd)

Here's where the trade to Colorado starts to get interesting. The following are the ground ball BABIPs and league-wide ranks for the Rockies during the same period:

2011: .230 (minus-3.0 percent, 10th)

2012: .254 (6.7 percent, 27th)

2013: .226 (minus-5.8 percent, 6th)

Going from the 23rd-best infield defense to the sixth could certainly help a guy with a career 47.5 percent ground ball rate. Furthermore, the Rockies had made significant defensive upgrades all over the diamond from 2012 to 2014. Here are the primary starters from those years for comparison:

Replacing Dexter Fowler (minus-16 defensive runs saved from 2011 to 2013) with Stubbs (plus-2 DRS so far in 2014) and Blackmon (plus-7), while also giving Barnes (plus-2) some starts in right seems to have made a huge impact on the Rockies' outfield defense. Here are their BABIPs and corresponding ranks on non-HR fly balls from 2011 to 2014:

2011: .144 (22nd)

2012: .148 (27th)

2013: .119 (28th)

2014: .086 (T-15th)

So let's see how Lyles' current and former teams stack up defensively so far in 2014 (through Monday):

Both squads have made significant strides, but the Rockies are better overall and more importantly for Lyles, significantly better than he ever pitched in front of in Houston. As a result, Lyles' fielding independent pitching (FIP) has flipped. In Houston, it always outpaced his ERA by at least 10 percent. Here in 2014, his stuff has played up thanks to his defense:

However, outperforming his FIP to such a degree, with such a low strikeout rate, usually signals that a hard regression is coming. There are other arguments that suggest we shouldn't totally buy in on this version of Lyles:

• Of the 116 pitchers to amass at least 30 innings this season, the .675 OPS of his opposing batters ranks 93rd highest. He's faced a soft schedule so far.

• His 8.3 percent HR/FB rate (based on homers/fly balls + popups + line drive homers) is at least three points lower than any prior season, and nearly two points lower than the major league average. Once the weather warms up in Colorado, will those fly balls start reaching the seats?

• His BABIP of .255 is nearly 50 points lower than his career-best, and over 40 points lower than the MLB average.

• His "swinging strike" percentage has dropped (again) to 10.4 percent (MLB average: 15.7 percent) while his in-play percentage has risen again to 33.8 percent (MLB average: 29.5 percent). The defense must continue its stellar play.

• Of the 107 ERA title qualifying starting pitchers through Wednesday, Lyles has the 22nd highest "well-hit average" at .171. Of the 21 pitchers with a higher well-hit average, only three have a lower BABIP. Batters are still making good contact against him. It could all be a run of good luck.

Best guess: Lyles will definitely regress from this hot start. However, with the improved defense behind him and a potent offense that can help take pressure off of him, he can certainly put up an ERA in the high threes and a WHIP around 1.30 this season. I'm sure the Rockies would take that.

Diane Firstman writes for Value Over Replacement Grit, a SweetSpot Network blog.