:format(webp)/cdn.vox-cdn.com/uploads/chorus_image/image/23639463/170249276.0.jpg)
Editor's note: Please welcome Eric to staff, formerly known as the fantastic commenter FGChoo.
Today marks the one-week anniversary of the Dexter Fowler trade rumor, which means it has been six days since most people forgot about the Dexter Fowler trade rumor. Yet, of all the free agents and trade targets connected to the Mariners in recent weeks, Fowler is the most likely to be in a Mariners uniform in 2014.
First, the obvious reasons:
- The Mariners would like to improve their team.
- The Rockies would like to improve their team.
- The Mariners would like a right-handed hitting outfielder.
- The Rockies would like a cost-controlled second baseman and pitcher.
- The Rockies have a right-handed hitting outfielder. His name is Dexter Fowler.
- The Mariners have several cost-controlled second basemen and pitchers. Their names vary depending on how they are spelled.
So far so good, but every team has at least one player another team could use and pairing them together is easy enough. We need something more substantial:
- For the Mariners, Fowler’s salary is financially pleasing compared to outside alternatives.
- For the Rockies, Fowler’s salary is financially displeasing compared to inside alternatives.
Fowler is scheduled to make $7.35 million in 2014 and will be arbitration eligible in 2015 before hitting free agency. From the Rockies perspective, two more seasons of Fowler will cost around $14 million more than they are scheduled to pay center field heir apparent Charlie Blackmon, who finished 2013 on an absolute tear. Blackmon might be a step down in terms of overall WAR, but the Rockies could use the extra cash to make up the difference.
From the Mariners perspective, that price tag and modest two-year commitment must look dirt cheap compared to the outside alternatives and they would have two years to decide if Fowler is more than a bridge to 2016. If Fowler simply repeats the average of his last three seasons in Colorado, he is a 2.3 WAR center fielder. The Mariners entire outfield combined for 0.0 WAR in 2013 and of that group, only Michael Saunders, Abraham Almonte and Dustin Ackley are guaranteed to return, and it’s debatable whether any of those players are worthy starters on a roster built for contention. In addition, the minor league system has nothing to offer in terms of immediate help.
It makes sense that Fowler is being dangled by the Rockies, and even more sense that the outfield-starved Mariners are interested. However, making sense of what Fowler might look like as a Seattle Mariner is not so simple.
Our own Ashley Varela touched upon the mysteries of evaluating Fowler as a Mariner last week and we have no idea how he might respond to playing 100% of his games in more context neutral environments. Projecting any Rockies hitter away from Colorado is tricky. Here is why:
<table>
<tr>
<td>Hitter</td>
<td>Yrs</td>
<td>wRC+</td>
<td>Road</td>
<td>Road%</td>
</tr>
<tr>
<td>Non-Rockies</td>
<td>2008-13</td>
<td>100</td>
<td>96</td>
<td>.96</td>
</tr>
<tr>
<td>Rockies</td>
<td>2008-13</td>
<td>98</td>
<td>89</td>
<td>.91</td>
</tr>
</table>
Road% illustrates the difference between road wRC+ and overall wRC+. The disparity between Rockies hitters and non-Rockies hitters makes sense considering Rockies hitters gain a huge advantage by playing 50% of their games at Coors Field. In addition, the average Rockies hitter is at a greater disadvantage on the road than the average non-Rockies hitter, but more on that later. First, let’s see how Fowler’s Road% compares to the average Rockies hitter:
<table>
<tr>
<td>Hitter</td>
<td>Yrs</td>
<td>wRC+</td>
<td>Road</td>
<td>Road%</td>
</tr>
<tr>
<td>Non-Rockies</td>
<td>2008-13</td>
<td>100</td>
<td>96</td>
<td>.96</td>
</tr>
<tr>
<td>Rockies</td>
<td>2008-13</td>
<td>98</td>
<td>89</td>
<td>.91</td>
</tr>
<tr>
<td>Fowler</td>
<td>2008-13</td>
<td>104</td>
<td>92</td>
<td>.89</td>
</tr>
<tr>
<td>Fowler</td>
<td>2012-13</td>
<td>114</td>
<td>97</td>
<td>.85</td>
</tr>
</table>
This is where Fowler begins to lose some of his shine. Whether Fowler has enjoyed an above average boost from Coors Field or has struggled more at lower elevations (likely a combination of the two), neither scenario works in his favor as we begin to translate his numbers to a more context neutral environment. And before anyone assumes that above-average hitters gain an even greater advantage from playing at Coors, here is Troy Tulowitzki:
<table>
<tr>
<td>Hitter</td>
<td>Yrs</td>
<td>wRC+</td>
<td>Road</td>
<td>Road%</td>
</tr>
<tr>
<td>Tulowitzki</td>
<td>2006-14</td>
<td>121</td>
<td>117</td>
<td>.97</td>
</tr>
</table>
Now, about the average Rockies hitter facing a greater disadvantage on the road than the average non-Rockies hitter . . . this is due to a phenomenon known as the Coors Hangover. As it pertains to hitters, there are two sides to the Coors Hangover and neither involves the Banquet Beer. One side of the Coors Hangover is what happens to Rockies hitters when they arrive in Colorado and adjust to pitches with less movement, and the other side is what happens when they leave Colorado and adjust to pitches with more movement. These road-to-home and home-to-road transitions happen 25-30 times per season. The latter is more extreme because the movement of every pitch is suddenly amplified and the competition is already acclimated to such conditions. In each case, batters perform below their actual true talent level for a hypothetical period of x plate appearances. Further complicating matters is the fact that batters are human, and every human is different, and sometimes even the same human is different depending on what he ate for breakfast, so x will be non-linear on an individual level.
Those poor Rockies hitters. You can’t take them seriously at home or on the road.
If you haven’t guessed by now, the Coors Hangover is an absolute bear to tackle. My entire weekend was spent seducing reams of retrosheet data and it only left me left me with more questions than answers. While there does appear to be a Coors Hangover rate for both home and away, there are two primary issues as they concern Fowler.
First, the rates are generic. They are averages for all Rockies hitters. Fowler is not an average Rockies hitter. His extreme Road% tells us that. Second, the rates tell us nothing about how a Colorado hitter responds when he joins a new team and no longer has to deal with the effects of the Coors Hangover. For that, we need before/after comps of hitters with a substantial number of plate appearances who have played for Colorado in the humidor era and elsewhere during their prime years. There are five of those.
I feel bad for dragging you all this way only to show you a dead end. And because I feel bad, I will share something I found interesting that actually does play in Fowler’s favor. Here are the aforementioned five players who left Colorado in the humidor era:
<table>
<tr>
<td>Player</td>
<td>Age</td>
<td>Team</td>
<td>PA</td>
<td>wRC+</td>
<td>Road</td>
<td>Road%</td>
</tr>
<tr>
<td>Matt Holliday</td>
<td>26-28</td>
<td>COL</td>
<td>2003</td>
<td>144</td>
<td>122</td>
<td>.85</td>
</tr>
<tr>
<td>Matt Holliday</td>
<td>29-31</td>
<td>OAK/STL</td>
<td>1861</td>
<td>148</td>
<td></td>
<td></td>
</tr>
</table>
<table>
<tr>
<td>Player</td>
<td>Age</td>
<td>Team</td>
<td>PA</td>
<td>wRC+</td>
<td>Road</td>
<td>Road%</td>
</tr>
<tr>
<td>Chris Iannetta</td>
<td>27-28</td>
<td>COL</td>
<td>649</td>
<td>96</td>
<td>68</td>
<td>.71</td>
</tr>
<tr>
<td>Chris Iannetta</td>
<td>29-30</td>
<td>LAA</td>
<td>652</td>
<td>109</td>
<td></td>
<td></td>
</tr>
</table>
<table>
<tr>
<td>Player</td>
<td>Age</td>
<td>Team</td>
<td>PA</td>
<td>wRC+</td>
<td>Road</td>
<td>Road%</td>
</tr>
<tr>
<td>Clint Barmes</td>
<td>29-31</td>
<td>COL</td>
<td>1453</td>
<td>78</td>
<td>64</td>
<td>.82</td>
</tr>
<tr>
<td>Clint Barmes</td>
<td>32-34</td>
<td>HOU/PIT</td>
<td>1318</td>
<td>71</td>
<td></td>
<td></td>
</tr>
</table>
<table>
<tr>
<td>Player</td>
<td>Age</td>
<td>Team</td>
<td>PA</td>
<td>wRC+</td>
<td>Road</td>
<td>Road%</td>
</tr>
<tr>
<td>Yorvit Torrealba</td>
<td>28-30</td>
<td>COL</td>
<td>946</td>
<td>75</td>
<td>70</td>
<td>.93</td>
</tr>
<tr>
<td>Yorvit Torrealba</td>
<td>31-33</td>
<td>SD/TEX/TOR</td>
<td>1000</td>
<td>89</td>
<td></td>
<td></td>
</tr>
</table>
<table>
<tr>
<td>Player</td>
<td>Age</td>
<td>Team</td>
<td>PA</td>
<td>wRC+</td>
<td>Road</td>
<td>Road%</td>
</tr>
<tr>
<td>Aaron Miles</td>
<td>27-28</td>
<td>COL</td>
<td>913</td>
<td>67</td>
<td>56</td>
<td>.84</td>
</tr>
<tr>
<td>Aaron Miles</td>
<td>29-30</td>
<td>STL</td>
<td>920</td>
<td>76</td>
<td></td>
<td></td>
</tr>
</table>
Note that only the aging Clint Barmes failed to improve his overall wRC+ after leaving Colorado. Also, three of the five hitters share a Road% similar to Fowler. In other words, there is hope that Fowler could improve his overall wRC+ with the Mariners. On the other hand, even small samples are offended by the smallness of this sample.
Now that we have established the overall picture of Dexter Fowler as somewhat cloudy, let’s zoom in a bit more. Without the benefit of HitF/X data, a more stylistic approach is required. Here are Fowler’s R/L splits with spray data:
<table>
<tr>
<td>2011-13</td>
<td>Vs</td>
<td>PA</td>
<td>BB%</td>
<td>K%</td>
<td>ISO</td>
<td>BABIP</td>
<td>wRC+</td>
</tr>
<tr>
<td>Fowler</td>
<td>R</td>
<td>1120</td>
<td>12.6%</td>
<td>25.7%</td>
<td>.173</td>
<td>.357</td>
<td>109</td>
</tr>
<tr>
<td>Pull</td>
<td>R</td>
<td>264</td>
<td></td>
<td></td>
<td>.386</td>
<td>.406</td>
<td>240</td>
</tr>
<tr>
<td>Center</td>
<td>R</td>
<td>234</td>
<td></td>
<td></td>
<td>.197</td>
<td>.360</td>
<td>144</td>
</tr>
<tr>
<td>Oppo</td>
<td>R</td>
<td>183</td>
<td></td>
<td></td>
<td>.108</td>
<td>.289</td>
<td>72</td>
</tr>
</table>
<table>
<tr>
<td>2011-13</td>
<td>Vs</td>
<td>PA</td>
<td>BB%</td>
<td>K%</td>
<td>ISO</td>
<td>BABIP</td>
<td>wRC+</td>
</tr>
<tr>
<td>Fowler</td>
<td>L</td>
<td>465</td>
<td>12.9%</td>
<td>16.1%</td>
<td>.135</td>
<td>.356</td>
<td>119</td>
</tr>
<tr>
<td>Pull</td>
<td>L</td>
<td>124</td>
<td></td>
<td></td>
<td>.073</td>
<td>.341</td>
<td>96</td>
</tr>
<tr>
<td>Center</td>
<td>L</td>
<td>115</td>
<td></td>
<td></td>
<td>.197</td>
<td>.394</td>
<td>168</td>
</tr>
<tr>
<td>Oppo</td>
<td>L</td>
<td>88</td>
<td></td>
<td></td>
<td>.268</td>
<td>.326</td>
<td>143</td>
</tr>
</table>
Fowler is a switch-hitter and his reputation for being better against lefties shows up here. However, there are some surprises. The K% against righties is much higher, but so is the quality of contact. Not that we should confuse a Coors-inflated .173 ISO for legitimate power, but there is something to like about Fowler’s weak side. Also, Fowler’s strength is hitting the ball to right field from both sides of the plate, a characteristic that would conceivably play well at Safeco. So how has Fowler fared away from Coors Field?
<table>
<tr>
<td>2011-13</td>
<td>Vs</td>
<td>PA</td>
<td>BB%</td>
<td>K%</td>
<td>ISO</td>
<td>BABIP</td>
<td>wRC+</td>
</tr>
<tr>
<td>Away</td>
<td>R</td>
<td>552</td>
<td>10.9%</td>
<td>27.7%</td>
<td>.129</td>
<td>.344</td>
<td>84</td>
</tr>
<tr>
<td>Away</td>
<td>L</td>
<td>219</td>
<td>14.2%</td>
<td>19.2%</td>
<td>.110</td>
<td>.342</td>
<td>105</td>
</tr>
</table>
That small flame of hope for Fowler against righties has just been doused. As we discussed earlier, a portion of this can be attributed to the Coors Hangover effect but it does not account for a whopping 25% drop in ISO and wRC+ at elevations more comparable to Seattle. Fowler is responsible for most of that. The good news is that while facing lefties on the road, Fowler retains all but 12% of his wRC+ which is more in line with the average Rockies hitter (9%). While we have no idea how he might respond to a more stable environment, evidence suggests that Fowler could maintain the majority of his overall wRC+ against lefties while losing more than half against righties. If those rates are indicative, here is what our new Fowler projection looks like:
<table>
<tr>
<td>2014</td>
<td>Vs</td>
<td>PA</td>
<td>BB%</td>
<td>K%</td>
<td>ISO</td>
<td>BABIP</td>
<td>wRC+</td>
</tr>
<tr>
<td>Fowler</td>
<td>R</td>
<td>456</td>
<td>11.5%</td>
<td>27.0%</td>
<td>.144</td>
<td>.349</td>
<td>92</td>
</tr>
<tr>
<td>Fowler</td>
<td>L</td>
<td>190</td>
<td>13.2%</td>
<td>16.9%</td>
<td>.129</td>
<td>.353</td>
<td>115</td>
</tr>
<tr>
<td>Overall</td>
<td>B</td>
<td>646</td>
<td>12.0%</td>
<td>24.1%</td>
<td>.139</td>
<td>.350</td>
<td>99</td>
</tr>
</table>
Let me clarify that I trust Steamer’s projected overall wRC+ of 111 more than I trust these numbers. However, we also have evidence that suggests Fowler’s bat may disintegrate following a move to the American League West, and we have almost no evidence to the contrary.
This is not a bad player, and he becomes a pretty good player with solid base running and defense in center field. Unfortunately, Fowler is one of the most inefficient base stealers in baseball (66% success rate) and his defense in center falls somewhere between bad and Vernon Wells, depending on how you feel about UZR’s injustice toward Colorado outfielders. Among all centerfielders since 2008, Fowler’s -40.6 UZR ranks 2nd to last between Matt Kemp and the previously mentioned Wells.
The total package of Dexter Fowler resembles an opposite-handed Michael Saunders, aesthetically and statistically. Both are 6’4" striders with nicknames to match (Condor vs. Daddy Long Legs) and the parallels carry all the way through their peripheral numbers. The Mariners could use a player like that, but the short-side of an outfield platoon doesn’t have to cost $15 million and a handful of prospects, either. Besides, after paying that price, the Mariners might be inclined to pencil Fowler into the starting lineup almost every day, essentially pairing Saunders with a mirror image of himself. Getting excited about one Saunders is challenging enough. Prepare to multiply that challenge by two because the possibility is very real.