clock menu more-arrow no yes

Filed under:

What do we know about hitters leaving Coors Field?

New, comments
Isaiah J. Downing-USA TODAY Sports

When the Rockies outfielders first came up in trade rumors in December, CarGo was one of the first ones to come up in rumors. As a result, Mike Petriello of MLB.com wrote about how teams should not be scared of the Coors Field effect.

Allow me to let Mike speak for himself before moving on:

While playing in Colorado obviously helps a player's home performance, evidence shows it may also hurt their road performance. For example, over the past 10 seasons, the Rockies scored 4,596 runs at home, unsurprisingly the most in baseball. But over the same period on the road, they scored 3,089 runs, the fewest in baseball, 151 runs behind 29th-place Houston. In order for both of those things to be true, either Coors Field would have to elevate baseball's worst offense to play like its best, or a middle-of-the-pack team would have had to receive positive effects at home and negative effects on the road -- which seems far more realistic. You can call that "the Coors Field effect."

Since 2002, when the Rockies installed a humidor, there have been 18 Rockies hitters with at least 650 plate appearances in Coors Field. Setting aside several current Rockies and those with limited or nonexistent post-Colorado careers, we're left with five good test cases: Matt Holliday, Dexter Fowler, Seth Smith, Chris Iannetta and Clint Barmes. (Let's acknowledge that we've introduced some amount of selection bias[...])

We're using raw OPS instead of the far superior wRC+ because OPS is not park-adjusted, a flaw that's very useful in this particular instance, [and] turning to OPS+, which accounts for park effects and is set so that 100 is "league average." How did this group do after leaving Colorado?

Overall, this group all did just fine offensively after leaving the friendly environment of Colorado.

Every single one improved, except for Barmes, who basically hit at his established averages no matter where he was.

Petrillo uses OPS+ to compare how only five former Rockies did after they left Coors field, finding an overall improvement among the players as compared to league average. His worry was that wRC+ would prohibit examining true Home and Road splits -- for various reasons, which the article delves into.

However, it should be noted that the same holds true even when adjusting for a neutral ballpark.

Player COL wRC+ Post-COL wRC+
Matt Holliday 133 142
Clint Barmes 71 73
Seth Smith 110 115
Dexter Fowler 103 116
Chris Iannetta 101 108

Part of the Coors effect is the hypothesis that pitches perform differently a mile in the air, and consequently, the hitters have a hard time adjusting to sea-level baseball.

Taking the long view, each of these players not only found their way back to their previous level of performance, but exceeded that upon leaving Coors.

As Petriello notes, five players is admittedly not a huge sample, but it really was all we have to work from. Let's take a look at our only other recent example and see what kind of drop off Troy Tulowitzki had after he was traded from Denver to Toronto mid-season:

Splits PA wOBA wRC+ BB% K% AVG OBP SLG OPS+ HR ISO BABIP
Colorado 351 .349 104 6.8% 20.5% .300 .348 .471 107 12 .170 .351
Toronto 183 .307 91 7.7% 23.0% .239 .317 .380 92 5 .141 .291

Not very encouraging, but not entirely fair. Tulo lifted Toronto's lineup but he would also miss some time due to a cracked shoulder blade after colliding with Kevin Pillar, which could explain the sudden loss in power -- but even then, we are discussing a mere 183 plate appearances, which falls short of the 650 PA analysis above.

Petriello is unable to find evidence that a Rockies hitter falls apart when he leaves Colorado for a full-time role elsewhere, and presumably, the Rays feel the same way.

The Coors Field effect is a very real phantom hovering over a player's performance, but it only haunts one house.