The Colorado Rockies are winning but the offense is struggling

It’s a tad strange to write an article about what’s gone wrong during a season that’s going so well. As I type this, the Rockies are nine games over .500, first in the division, and fresh off a four-game series that they split with the Dodgers. Furthermore, the team unlocked the “Cheap Tacos” achievement on Sunday!

That said, the offense has had its fair share of struggles this season. After scoring 5.22 runs per game last season, the Rockies are scoring just 4.74 runs per game in 2017 (i.e. about 70 runs fewer over a full season). Take a look at the current rankings of team non-pitcher wRC+ (explained here):

Compare that to last year’s end-of-season rankings:

I’ll note first that there are good reasons to think that wRC+ underestimates the talent level of Rockies hitters. While park factors are useful (and essential) for comparing and adjusting for run environment (the overall rate of run scoring), they don’t take into consideration the difficulty of adjusting to how pitches move differently at different altitudes. That being, said, assuming the “Coors Hangover” hasn’t gotten worse in the past year, we can still fairly compare the Rockies to themselves using wRC+. As you can see, the Rockies have dropped 10 points of wRC+, which is worth roughly 60* runs scored over the course of a full season. Translate those to wins at the standard rate of 10 runs per win, and that amounts to six games in the standings come seasons’ end.

(*using a ballpark figure of 10 points of wRC+ being worth .01 runs per plate appearance).

The questions on my mind:

1. What specifically has been different about the Rockies’ offense?
2. What should our expectations be, going forward?

Finding the culprits

The elephant in the room here is: injuries. Currently the Rockies lack their ostensible starter in left field (David Dahl, assuming he wasn’t merely a figment of my imagination) and their top catching prospect (Tom Murphy); they just barely got Ian Desmond back and now have lost Trevor Story for a short while. As a result, the team is giving more plate appearances than intended to Gerardo Parra, Alexi Amarista, and Dustin Garneau. On the flip side, Mark Reynolds has been a revelation at first base, offsetting the drop in production to a degree. Furthermore, Story wasn’t exactly hitting much this year. All-in-all, it’s hard to blame the drop in hitting entirely on the injury bug. So, we go a-hunting for clues.

Here’s a rectangle full of numbers for us to look at:

The first thing that jumps out to me is that the Rockies have (so far) been worse in each of these categories compared to last season. It’s important to note that the overall league stats have changed slightly: league-wide, batters are striking out a half-percentage point more often, walking nearly a full percentage point more often, league BABIP has dropped by .08, and line drive rate has dropped a point. Even considering these changes, the Rockies have performed worse relative to the league than they did last season.

Perhaps most concerning is the drop in walk rate despite the league-wide increase in the free pass. Coupled with the increase in strikeout rate, the Rockies are needing to do more damage on balls-in-play, which they aren’t, as the drop in BABIP makes clear (partly, but not entirely, explained by a lower line drive rate and hard contact rate). What could be driving the problems at the plate?

The Rockies aren’t swinging at more pitches than last year, and they’re actually chasing fewer pitches out of the zone than last year. Normally, we would expect this to lead to an increase in walk rate. The real key here is the Zone%: the rate at which opposing pitchers are throwing pitches in the zone. This season, the Rockies have been pitched more aggressively. They rank second in zone rate.

I don’t have a complete explanation for this change. My suspicion is that it has something to do with the quality of pitching that the Rockies have faced:

• Nine games against the Dodgers, first in fWAR for pitchers
• Six games against the Diamondbacks, fifth in fWAR
• Seven games against the Giants, ninth in fWAR
• Three games against the Cubs, 10th in fWAR
• Four games against the Brewers, 12th in fWAR

Indeed, 29 of the team’s 39 games have been against teams with an above-average pitching staff (thus far).

Over the full season, that will normalize, as the Rockies get to face teams with weaker pitching staffs. It might even normalize this week: the Reds and Twins rank 24th and 28th, respectively, in pitching fWAR.

As far as results on balls-in-play, the story seems a bit different. We’re going to use a bit of Statcast data. Over at Baseball Savant, they publish a metric called “xBA”, or “expected batting average.” It uses the launch angle and exit velocity to estimate the probability of a given batted ball falling for a hit, and then adds up those hit probabilities (combined with actual strikeout rates) to estimate a what a player’s batting average “should” have been, given average luck. This is not park adjusted, meaning a Rockies player will usually have a better real-life batting average than expected batting average. However, as long as we compare apples-to-apples, meaning Rockies players to themselves, we avoid those thorny issues for the most part. Here are the numbers to compare this season to last season, for Rockies hitters:

While Tony Wolters and Mark Reynolds have seen upticks in their expected production (as has been mirrored on the field), four of the major cogs of last season’s offense have experienced fairly steep declines. Carlos Gonzalez and Trevor Story in particular have cratered relative to last season. Over at Fangraphs, Travis Sawchik hypothesized that Story’s struggles stem in part from going from above-average to extreme in fly-ball rate. Hopefully upon return from the DL, Story will be able to cut down on the popups and regain some of the line-drives that helped him post a strong rookie season.

Gonzalez, on the other hand, is a bit of a mystery. He’s striking out a tad more than last season, but not by much. His walk rate has improved, up to just over league average. In the past series with the Dodgers, he seemed to break out a bit, collecting six hits and reaching base eight times.

That being said, his relative lack of power output (just two home runs on the season) is concerning. Looking at Statcast data, his exit velocity has declined by 1 mph relative to league average, and he’s lost nearly as much speed on his swing. His average launch angle, which has typically been quite low for a power hitter already, has dropped from 8.5 degrees to just 5.7 degrees. As a result, his ground-ball rate is up to a career-worst 53.1 percent. Since he tends to pull his grounders, he’s become an easy victim for defensive shifts.

Here’s chart of his rate of hard contact and rate of ground-balls:

Carlos experienced a similar dip in hard contact and increase in ground-ball rate early last season. As he made more hard contact, he began to elevate the ball more often. This season, while he’s been hitting the ball much harder recently, his ground-ball rate has spiked even higher. Due to the frequency of shifting, those hard-hit grounders aren’t doing as much damage as their exit velocity would indicate.

Gonzalez is by no means an old player: he’s only 31 and a half. While we shouldn’t be expecting a steep decline in production, it is difficult to predict which hitters have strong mid-30s performances and which players take steps back. Going forward, he’ll likely need to do the opposite of what Trevor Story needs: elevate the ball.

So far, all of the stats referenced here have been exclusively for position players. However, as a National League team typically gives around five percent of its plate appearances to pitchers, it’s worth seeing how the Rockies have stacked up in that department. Last season, the Rockies had the second-worst group of hitting pitchers in the NL. Their -33 (!) wRC+ mark beat only the Reds, and they compiled -1.0 WAR as hitters (remembering that replacement-level hitting for a pitcher is a much lower standard already). This season, they are once again the second-worst unit in the NL, with a -60 wRC+ (head of the Braves’ -67 mark). They’ve already piled up -0.6 WAR in their 90 total plate appearances. Here’s a chart comparing the difference between a team’s wRC+ with pitchers included versus without pitchers:

Without pitchers included in the data, the Rockies have an 89 wRC+. Include their pitchers, and that drops down to 80 wRC+, third-to-last in the league.

I don’t have good ideas for getting pitchers to hit better, especially with the three rookies on the staff (although Marquez has shown off a sweet stroke). This is something the team will probably just have to live with.

Rest-of-season outlook

With about 25 percent of the season complete, the Rockies’ corps of hitters has some work to do. Despite a much improved pitching staff, the drop in offense has given a thinner margin for error than is normally desirable (especially with three rookies in the starting rotation). That said, there are reasons to think that the remaining games will be a better display of the lineup, as the Rockies are able to shift plate appearances to better hitters (*cough* bench Parra for Tapia please *cough*), and experience the needed positive regression from Carlos Gonzalez and a healthy Trevor Story. Plus, the Rockies still get to play the Padres a bunch more this season!

With the Rockies having their best-ever start through this many games, and more talent in the rotation and bullpen then they’ve ever had, there’s real hope for the Rockies to contend for their first-ever division title. It will be critical that the bats to step up (as they did in the series finale against the Dodgers) when the young starters experience growing pains. Just as important, more offense means more chances to tweet about tacos and that one word conversation stopper: “Coors” (even during road games, of course).

What else could you ask for?