Descending from Altitude: What Can the Braves Expect from Mike Hampton?

As it has been well-publicized by the media, both in and outside of the Colorado area, Mike Hampton‘s time in the Mile High City was anything but successful. In his two years with the Rockies, Hampton posted a won-loss record of just 21-28, along with a major league-high ERA of 5.75-more than half a run higher than the next closest starter, fellow Rockie and member of the Great Changeup Experiment, Denny Neagle.

And yet, as 2003 regular season gets under way, whose name do we see listed in the Atlanta Braves starting rotation? None other than Michael William Hampton. Acquired over the off-season in one of the most complicated deals of all time not involving Herschel Walker, Hampton joins perhaps the most vulnerable Braves rotation in a decade-a unit that has already seen its star have a brush with mortality, and an off-season signee go down until July.

What can be expected of Hampton, though? Members of the media have spent more than their fair share of time waxing philosophical on the situation, with most coming to the conclusion that gambling on the 30-year-old lefty is a worthwhile risk. Granted, this might very well be the case, as Hampton was among the winningest pitchers in the National League before signing with Colorado in 2001. Fellow Rockies refugee Darryl Kile was able to make the transition from 5,200 feet above sea-level to Busch Stadium without missing a beat. If he could do it, why can’t Hampton?

But what does history tell us? Well, if one takes “history” to mean the host of statistically comparable pitchers who have appeared over the past 130 years, then Nate Silver’s PECOTA projection system is as far as you’re going to have to look. According to PECOTA, Hampton should be good for roughly 160 innings of slightly below-average baseball in 2003-an ERA of 4.93 with a strikeout-to-walk ratio of nearly one-to-one.

However, in many ways, PECOTA’s assessment of Hampton is incomplete. After all, for the past two seasons, Hampton has spent more than half his starts in the most eye-poppingly ludicrous hitting environment in the history of the world. And if you don’t think that has an adverse effect on your skill set, you’re probably going to want to contact Jeff Cirillo and have a nice, long chat.

Now, in order for us to fill in the gaps that PECOTA ignores, the first thing we’ll have to do is look at former Rockies pitchers who have managed to escape from the clutches of Mile High, and make a living throwing a baseball elsewhere. Is there a trend we can identify among pitchers leaving Colorado who go on to pitch for other teams? Does playing at altitude appear to have a carry-over effect on a pitcher’s ability to perform? Below is a list of the top 15 pitchers in Rockies history, ranked by their innings pitched with the organization.

Pitcher             IP    ERA  PRAR/IP
--------------------------------------
Pedro Astacio    827.1   5.43    0.151
John Thomson     610.2   5.01    0.160
Kevin Ritz       576.1   5.20    0.154
Jamey Wright     541.1   5.57    0.100
Armando Reynoso  503.0   4.65    0.203
Brian Bohanon    471.0   5.83    0.119
Curt Leskanic    470.0   4.92    0.219
Darryl Kile      420.2   5.84    0.112
Mike Hampton     381.2   5.75    0.045
Steve Reed       369.2   3.68    0.327
Roger Bailey     356.0   4.90    0.160
Marvin Freeman   337.0   4.91    0.175
Denny Neagle     335.0   5.32    0.128
Darren Holmes    328.0   4.42    0.244
Bruce Ruffin     321.0   3.84    0.324
                 -----           -----
AVERAGE          456.2           0.169

First, an explanation of the stats. Innings pitched (IP) and Earned Run Average (ERA) you know. Pitcher Runs Above Replacement per Inning Pitched (PRAR/IP) you’re probably not quite as familiar with. PRAR is a statistic developed by Clay Davenport that measures the number of context-neutral runs that a pitcher saves above replacement level. It’s both park- and defense-adjusted, with higher numbers meaning more runs saved per inning. Dividing that number by innings pitched simply turns it into a rate stat, like OBP.

Getting back to our topic, though, how have these pitchers done since leaving Colorado? Below is a breakdown of each pitcher’s performance, through 2002, after appearing with the Rockies.

                          -- Season After --       ----- Overall -----
Pitcher         Team      IP    ERA  PRAR/IP       IP     ERA  PRAR/IP
----------------------------------------------------------------------
Pedro Astacio    HOU    28.2   3.14    0.349     220.2   4.57    0.141
                 NYN   191.2   4.79    0.109
John Thomson     NYN    54.1   4.31    0.072      54.1   4.31    0.072
Jamey Wright     MIL   164.2   4.10    0.212     488.2   4.73    0.127
Armando Reynoso  NYN    91.1   4.53    0.121     545.2   4.77    0.136
Curt Leskanic    MIL    77.1   2.56    0.478     146.2   3.07    0.395
Darryl Kile      STL   232.1   3.91    0.245     544.1   3.54    0.259
Steve Reed      SF/CL   80.1   3.14    0.508     323.1   3.40    0.319
Darren Holmes    NYA    51.1   3.33    0.292     174.0   4.03    0.218
                       -----           -----     -----           -----
AVERAGE                108.0           0.219     312.1           0.205

Granted, it’s not the most sophisticated way to evaluate performance, but for an exercise such as this, it’ll do. As you can see, the column on the left reflects the performance of former Rockies pitchers in their first year away from Coors-or in the case of Pedro Astacio, the first year-and-a-half. Predictably, runs allowed are down across-the-board, with every pitcher improving upon his Elevated ERA™ by at least half a run.

The pitchers’ context-neutral performances are more surprising. Both in the season immediately after leaving Colorado and overall, former Rockies make a noticeable improvement on their PRAR/IP scores-0.050 of a run in the “season after,” and 0.040 of a run overall. Translated to 200 innings over the life of a season, that’s 10 runs of improvement in the year after leaving Colorado, and eight runs of improvement, per season, going forward, that cannot be attributed to a park illusion.

In other words, pitchers who leave Colorado don’t just go to look better, thanks to more friendly home parks. They actually get better, with regard to their physical ability to perform.

Of course, we must tread lightly in determining the significance of this conclusion. While the sample of innings we’ve used is reasonable in size, the number of pitchers-eight-who’ve left Colorado and made appearances with other teams certainly obscures our findings. Nevertheless, it’s probably safe to say that there isn’t a negative effect on a pitcher’s performance after leaving Colorado, which is, in all honesty, good enough for me.

OK, so we can expect Hampton to perform a little better this season than he did last season with the Rockies, simply because he’s changing home addresses. Big deal. Hampton’s problem wasn’t really Coors Field last year, anyway. This will come as no surprise to those of you who pay close attention to the box scores, but Hampton not only sucked at home in 2002, he sucked on the road as well. Take a look:

2002      IP    H    R   ER   HR   BB    K    ERA   BAA
-------------------------------------------------------
Home      69   79   48   44    7   35   36   5.68  .247  
Away     109  149   87   78   17   56   38   6.44  .288 

Yes, you’re seeing that correctly. Despite pitching close to 40% of his innings in the best hitting environment on the planet, and pitching in the NL West-the division which claims three of the best pitchers’ parks in all of baseball in Dodger Stadium, Qualcomm Stadium, and PacBell Park-Hampton was worse on the road last season in almost every possible way.

So what does this mean? If Hampton was worse on the road last season than he was at home, does that negate the positive effect we would normally expect to see from a pitcher escaping from Colorado? Maybe. Since 1987, 25 different pitchers have “accomplished” this same feat-that is, posting a higher ERA on the road than at home, while their home Park Factor was above 105. Roughly three-quarters of those pitchers performed worse in the following season. Take a look:

Season 1

                                   Home           Away      Overall
Pitcher          Year  Team      IP    ERA      IP    ERA   PRAR/IP
-------------------------------------------------------------------
Mike Moore       1987   SEA   109.1   4.12   121.2   5.25     0.108
Pete Smith       1988   ATL   125.0   3.46    70.1   4.09     0.200
Mike Bielecki    1989   CHN   114.1   2.83    98.0   3.49     0.235
Roger Clemens    1989   BOS   105.2   2.90   147.2   3.29     0.316
Frank Viola      1989   MIN   128.0   3.52   133.0   3.79     0.313
Mike Bielecki    1990   CHN    75.2   4.88    92.1   4.97     0.083
Mike Harkey      1990   CHN    85.1   2.43    88.1   4.08     0.276
Char. Leibrandt  1990   ATL    87.0   2.59    75.1   3.82     0.271
Roy Smith        1990   MIN    73.1   4.30    80.0   5.29     0.085
John Smoltz      1990   ATL   124.0   2.76   107.1   5.11     0.259
Kevin Tapani     1990   MIN    74.2   3.38    84.2   4.68     0.207
Frank Viola      1993   BOS    91.2   2.26    92.0   4.01     0.272
Erik Hanson      1996   TOR    96.2   5.31   118.0   5.49     0.093
Pat Hentgen      1996   TOR   152.1   3.19   113.1   3.26     0.330
Armando Reynoso  1996   COL    88.0   4.81    80.2   5.13     0.202
John Thomson     1997   COL    93.2   4.32    88.0   5.11     0.180
Glendon Rusch    1998   KCA    85.0   5.66    65.2   6.17     0.103
Steve Trachsel   1999   CHN    98.0   5.14   107.2   5.93     0.117
Chris Holt       2000   HOU   113.2   4.91    93.1   5.88     0.145
Steve Parris     2000   CIN    90.2   4.07   102.0   5.47     0.145
Brad Radke       2000   MIN   123.1   3.94   102.1   5.05     0.260
Mark Redman      2000   MIN    76.0   4.14    75.1   5.38     0.264
Masato Yoshii    2000   COL    67.2   5.85    99.2   5.87     0.131
Tony Armas Jr.   2001   MON    79.1   3.97    85.0   4.87     0.224
Chad Durbin      2001   KCA   109.2   4.27    69.1   5.97     0.123
                              -----          -----            -----
AVERAGE                        98.2           95.1            0.202

Pretty, isn’t it? And here’s the following season for each pitcher, to complete the comparison. You’re going to want to keep your eye on the two columns on the far right, as they display how much better or worse each pitcher was in Season 2 than in Season 1.

Season 2

Pitcher          Year      IP    ERA   PRAR/IP     +/-
------------------------------------------------------
Mike Moore       1988   228.1   3.78     0.245  +0.137
Pete Smith       1989   142.0   4.75     0.063  -0.137
Mike Bielecki    1990   168.0   4.93     0.083  -0.152
Roger Clemens    1990   228.1   1.93     0.487  +0.171
Frank Viola      1990   249.2   2.67     0.348  +0.035
Mike Bielecki    1991   173.2   4.46     0.006  -0.077
Mike Harkey      1991    18.2   5.30     0.054  -0.222
Char. Leibrandt  1991   229.2   3.49     0.174  -0.097
Roy Smith        1991    80.1   5.60     0.037  -0.048
John Smoltz      1991   229.2   3.80     0.196  -0.063
Kevin Tapani     1991   244.0   2.99     0.295  +0.088
Frank Viola      1994    31.0   4.65     0.194  -0.078
Erik Hanson      1997    15.0   7.80    -0.133  -0.226
Pat Hentgen      1997   264.0   3.68     0.258  -0.072
Armando Reynoso  1997    91.1   4.53     0.120  -0.082
John Thomson     1998   161.0   4.81     0.217  +0.037
Glendon Rusch    1999     5.0  12.60     0.000  -0.103
Steve Trachsel   1997   201.1   4.51     0.139  +0.022
Chris Holt       2001   151.1   5.77     0.053  -0.092
Steve Parris     2001   105.2   4.60     0.151  +0.006
Brad Radke       2001   226.2   3.94     0.235  -0.025
Mark Redman      2001    58.0   4.50     0.172  -0.092
Masato Yoshii    2001   113.0   4.78     0.150  +0.019
Tony Armas Jr.   2002   164.1   4.44     0.176  -0.048
Chad Durbin      2002     8.1  11.88    -0.480  -0.603
                        -----            -----
AVERAGE                 143.1            0.198

While there are certainly some notable performance swings from Season 1 to Season 2, overall the variance doesn’t appear to be large enough to render a definitive conclusion. While the majority of pitchers saw a decline after posting a worse ERA on the road than they did at home, the overall difference is simply not large enough to allow us to say that Hampton is likely to do the same. Sure, -0.004 PRAR/IP shows up on the screen, but translated to 200 IP, that’s less than a one-run difference from the average pitcher’s performance in Season 1. Yes, that’s right-one run.

So what does all of this mean? Well, essentially it means that based strictly on historical precedent, the Braves can expect Hampton to be the same pitcher in 2003 as he was in 2002, if not a little better. According to the data we’ve presented, there is some evidence to suggest that pitchers show improvement after leaving Colorado, for no other reason than the fact that they’ve left Colorado. We’ve also shown that pitching worse on the road when you should have technically been worse at home is not a precursor to a decline, and is more likely just a statistical fluke.

Of course, that conclusion also ignores a number of facts specific to Hampton’s profile, two of which are actually addressed in PECOTA’s projection:

• Even before joining the Rockies, Mike Hampton issued a lot of free passes. People tend to forget this because he spent the first seven years of his career in two of the best pitchers’ parks in all of baseball-the Astrodome and Shea Stadium-but Mike Hampton wasn’t really all that good to begin with. In his two seasons prior to joining the Rockies, Hampton struck out a total 328 batters, while walking 201. Does that scream $121 million to you? Yeah, I didn’t think so.
• He keeps the ball on the ground. Outside of Tim Hudson, there really isn’t a starter in the majors who induces more ground balls than Hampton does. Even with an ERA over six last season, Hampton allowed just 24 home runs in 178.2 innings pitched. Granted, that might not seem like anything special to the naked eye, but that’s quite impressive given the total number of runs he allowed (138).

• He’s joining a staff headed by perhaps the greatest pitching coach ever. He might not be able to heal lepers or return sight to the blind, but Leo Mazzone can raise pitchers from the dead with the best of ’em. If there’s one person who could turn Hampton’s career around at this point, it’s Mazzone.

Now, does this mean that the Braves made a wise decision in gambling on Hampton? Given that they’re paying only a portion of his contract, and that they have a legitimate miracle worker in the bullpen, probably. But then again, just because you’ve made a good bet, that doesn’t necessarily mean that you’re likely to succeed. If you don’t believe me, just ask Dan O’Dowd.