Sometimes It Is How Far You Hit Them

It’s true that no matter how far you hit a home run, it’s still only worth one run. However, people still want to know who hits them the farthest and the hardest. Using an analytical method and tool called Hit Tracker, I was able to show just that by compiling a complete record of long balls struck during the 2006 MLB season.

During the off-season, it struck me that it might also be interesting to know which hitters made a habit of sneaking the ball just over the fence instead of blasting it into the cheap seats. While a tape-measure homer is likely to be a home run no matter what, the ones that barely clear the wall are very susceptible to the influence of wind and temperature, and sometimes also depend on the leaping ability (or lack thereof) of the outfielders.

These shortest of homers are also the sort of hits that could have turned out very differently in another ballpark, had they only been struck on a different day. I wondered, how much of a player’s home run total might be attributable to good or bad luck, as opposed to the player’s underlying skills and strength?

To satisfy my curiosity, I started by classifying every home run into one of three categories, as follows:

1. “Just Enough,” or “JE”: Any homer which cleared the fence by 10 vertical feet or less, OR whose landing point was within one fence height of the fence (e.g. less than seven feet past a fence that is seven feet tall). This category also includes inside-the-park homers.

2. “No Doubt”, or “ND”: Any homer that cleared the fence by at least 20 vertical feet AND whose landing point was at least 50 feet past the fence.

3. “Plenty,” or “PL”: all other homers.

A few interesting anecdotal pieces of data emerged right away:

  • Jim Edmonds of the Cardinals hit 19 homers in 2006, only one of which was of the “Just Enough” type. Perhaps his own experience at robbing his fellow major leaguers above the fence led him to put a little extra on his own homers, just to be sure.
  • Washington’s Alex Escobar knocked four balls over the wall in only 87 ABs, part of a fairly promising .356/.394/.575 injury-shortened season, but all four of his homers were JE’s that barely cleared the fence. Turn those four into long flyouts (a worst-case scenario, admittedly) and his line would be .310/.354/.391.
  • Among the 19 homers allowed by Cincinnati’s Kirk Saarloos were seven No Doubters but only one JE.
  • Seattle/Philadelphia pitcher Jamie Moyer served up 33 homers in ’06, but only one flew far enough past the fence to reach the “No Doubt” category. Consider Moyer’s “crafty veteran” credentials renewed.

Beyond the anecdotes, it turns out that when you look at the entire league’s data together, analysis of how far home runs carry beyond the fence can provide some interesting insight into how much of a player’s home run production from last year was due to good fortune that might not shine on him again this year.

When we examine the 2006 season data (which is available here), we find that the MLB average breakout of home runs by type was 27% JE, 55% PL and 18% ND. For hitters with above average power, the ratio shifts to 25%/55%/20%, and for the elite sluggers, the ratio becomes 23%/54%/23%, as “No Doubt” homers make up a larger fraction of their output.

A player should hit his home runs according to these ratios, given enough home runs to smooth the results. Unfortunately, even Ryan Howard and David Ortiz didn’t hit enough homers to reliably generate a smooth output, so we need to decide what it means when an unusual split exists among the three home run types.

Because fly ball distances follow a roughly normal distribution, if a player hit a lot of ND homers, that player should also hit a lot of PLs and a lot of JEs, given enough at bats and barring any unusual run of bad luck or frequent encounters with home run thieves like Torii Hunter or Andruw Jones. Similarly, when a player doesn’t hit many NDs, he should not hit many JEs either. This means that the ratio of ND homers to JE homers is a critical indicator.

JE homers are the ones that can swing a players HR totals: A little luck, good or bad, can make the difference between a home run and a warning track flyout. Those who hit a lot fewer JE homers than expected (e.g. Adam Dunn, who hit only two JE homers against 16 NDs) are likely to be luckier this year, and thus tally more home runs. Those who hit a lot more JE homers than expected (e.g. Brian McCann, who hit 11 JE homers and only 1 ND) are likely to suffer worse fortune this year, and end up with fewer homers overall.

The plot below shows the percentage of “No Doubt” homers for every player who amassed at least 300 ABs and hit at least 20 HRs in 2006. The black line is a best fit line; its slope represents the increase in expected ND percentage as home runs per 500 AB increases. Some notable players are highlighted on the plot.


The next plot shows the percentage of “Just Enough” homers for every player who amassed at least 300 ABs and hit at least 20 HRs in 2006. Some of the same names appear on this plot, once again far from the best fit line.


Because of the small sample sizes involved, we should be cautious about drawing conclusions. However, some players’ home run “splits” turned out far enough from the typical ratios that a regression back towards the mean can be expected.

A Hardball Times Update
Goodbye for now.

A metric for describing how closely a player follows the “typical” splits is the ratio of ND to JE homers. The league average is 0.67, and for sluggers with > 30 HR per 500 AB, the ratio is 0.92; we will consider any hitter with > 1.50 to have a strong likelihood of hitting more homers, while anyone with < 0.33 is a strong candidate to hit fewer homers in 2007.

Upside List: (Number of HRs JE/PL/ND, ratio = ND/JE)

Jim Thome: 8/21/13, ratio = 1.63
Travis Hafner: 7/23/12, ratio = 1.71
Carlos Beltran: 6/25/10, ratio = 1.67
Albert Pujols: 8/27/14, ratio = 1.75
Manny Ramirez: 6/18/11, ratio = 1.83
Aramis Ramirez: 6/21/11, ratio = 1.83

Jason Giambi: 3/26/8, ratio = 2.67
Alex Rodriguez: 3/23/9, ratio = 3.00

Very High:
Jonny Gomes: 3/7/10, ratio = 3.33
Adam Dunn: 2/22/16, ratio = 8.00
Mike Jacobs: 1/10/9, ratio = 9.00

Downside List:

Alfonso Soriano: 18/23/5, ratio = 0.28
Josh Willingham: 12/11/3, ratio = 0.25
Nick Johnson: 8/13/2, ratio = 0.25
Brad Hawpe: 8/12/2, ratio = 0.25
Bill Hall: 15/17/3, ratio = 0.20
Nick Swisher: 14/18/3, ratio = 0.21
Justin Morneau: 15/16/3, ratio = 0.20
Jacque Jones: 11/14/2, ratio = 0.18 (though perhaps we should give Jones a break, as one of his JEs was a 445-foot blast to dead center field that just cleared Tal’s Hill at Minute Maid Park!)
Rich Aurilia: 7/15/1, ratio = 0.14

Ramon Hernandez: 8/14/1, ratio = 0.13
Adrian Beltre: 8/16/1, ratio = 0.13

Very High:
Ryan Zimmerman: 9/10/1, ratio = 0.11
Brian McCann: 11/12/1, ratio = 0.09
J.D. Drew: 4/16/0, ratio = 0.00

Notice that the hitters with the highest ratios are mostly guys already known for their prodigious power, so it may seem strange to suggest that there is upside to their 2006 power numbers, but most other sluggers follow the typical ratios fairly well: Howard hit 10/35/13, Ortiz hit 11/31/12, Lance Berkman hit 11/21/13, Frank Thomas hit 10/22/7, Jermaine Dye hit 10/22/12, etc. The hitters on the upside list were unlucky in 2006, and will probably enjoy better luck this year.

The hitters with the lowest ratios are mostly guys who had career years in terms of power: Among the 14 players on the list, only Aurilia, Drew and Beltre did not set or match career highs in homers last year. However, lots of other hitters had career years while conforming to the typical ratio: Matt Holliday hit 8/17/9, Adam LaRoche hit 6/20/6, Raul Ibanez hit 7/21/5, Prince Fielder hit 7/14/7, Craig Monroe hit 4/20/4, etc. The hitters on the downside list, immensely talented hitters all, nevertheless owe a portion of their outstanding performances from a year ago to good fortune (as is usually the case when a player has a great year). They probably won’t enjoy the same luck in 2007.

Thus far we’ve narrowed the field of consideration to those who amassed 300 ABs and 20 HRs, and focused only on the blatant statistical outliers, in order to minimize the likelihood of being misled by small sample size effects. However, a few more names did stand out among those who did not meet the minimums listed above. David Ross (2/13/6), Conor Jackson (1/7/7) and Edwin Encarnacion (1/10/4) seem to have gotten less than their share of JE homers, and thus should do better in 2007. Joe Mauer (10/3/0) and Jhonny Peralta (11/2/0) both made a living in 2006 dropping homers just over the fence; in 2007, we should expect fewer of their long flies to make it out.

Similar lists can be compiled for pitchers, reflecting how some pitchers gave up fewer JE homers than expected, and some more than expected, based on the proportions of the different types of home runs they allowed. Due to the smaller sample sizes, I will stick to a single upside and downside list:

Upside List: (Number of HRs JE/PL/ND, ratio = ND/JE)

Kyle Lohse: 8/7/0, ratio = 0.00
Sean Marshall: 10/9/1, ratio = 0.10
Paul Maholm 9/9/1, ratio = 0.11
Jose Contreras 9/10/1, ratio = 0.11
Joel Piniero 9/12/2, ratio = 0.22
Jason Bergmann 7/4/1, ratio = 0.14

Downside List:

Kirk Saarloos: 1/11/7, ratio = 7.00
Ryan Madson 2/12/6, ratio = 3.00
Scott Kazmir 3/4/8, ratio = 2.67
Taylor Buchholz 4/8/9, ratio = 2.25
Brett Myers 3/20/6, ratio = 2.00

Visit the Hit Tracker site for trajectory analysis of home runs and other batted balls throughout the 2007 season.

Comments are closed.