From 1973, when the American League first adopted the designated hitter, until 2006, the year after Major League Baseball first began suspending players who tested positive for performance enhancing drugs, nine of the highest home run season totals occurred between 1998 and 2006, which was the apex of the so-called Steroid Era.
Just a few short years later, however, home runs declined dramatically. Major leaguers hit 773 fewer home runs in 2010 than they did in 2006. In 2014, they hit 427 fewer than they did in 2010.
Along with the disappearing homers, runs scored also dipped across baseball. Over the period spanning the 19 seasons from 1997 to 2015, all six of the years with the lowest run totals occurred between 2010 and 2015.
What happened next was surprising and mysterious. While the offensive environment remained depressed (the league wOBA was just .318 last season), big leaguers hit 5,610 home runs in 2016, the second-most all time and just 83 shy of the major league record set in 2000.
There are ongoing investigations about whether the ball is juiced. Theories abound as to what, exactly, is causing the home run spike. Whatever the cause, the home run is back, and with a vengeance.
This article was born from a gut feeling, and partially as a follow-up to my colleague Jim Turvey’s recent piece in which he searched for a correlation between wins and losses and park effects. Turvey hypothesized that hitter-friendly ballparks may be helping unexpected teams (like the Yankees and Rockies) succeed in 2017, and that non-hitter-friendly ballparks may be causing preseason favorites (like the Giants and Mets) to scuffle.
Although Turvey wasn’t able to find a particularly strong correlation between the two variables, I wanted to dig deeper. Like Turvey, I had a hunch that hitting home runs is more important now than ever.
To investigate this notion, I looked at team home run totals and team winning percentages in every year since 2010. I calculated the correlation coefficient (R) for each season, which basically just measures the strength of the relationship. A perfect positive correlation equals 1.00, a perfect non-correlation equals 0, and a perfect inverse correlation equals -1.00.
I lumped 2010-15 together in one chart to spare the reader from having to look at eight charts. The lines of best fit are included in the charts to illustrate the relationship between the variables. The correlation coefficients are rounded to two decimal places.
2010-2015: R = 0.39
2016: R = 0.36
2017: R = 0.50
The first thing that jumps out is that the correlation coefficient for 2017 is indeed much higher than it was in 2010-15 and 2016. This initially seems to support the idea that hitting home runs is more important now than it has been recently.
While this is true to some extent, the following chart of the correlation coefficients for each individual season from 2010 to 2017 tells a different story:
Over the eight-year period, the relationship was strongest (0.66) in 2011. It fell all the way to 0.21 in 2014, and that happened to be the year with the fewest home runs this century. However, 2011 was the year with the second-fewest home runs since 2000, and it had the highest R^2 in the observed years.
What does it all mean? The correlation coefficient has increased every year since 2014, to the point where it’s now more than twice as high as it was three years ago, so technically speaking, home runs and winning percentage have become increasingly correlated lately. However, there were similarly high correlations in the years preceding 2014 when home run rates were depressed, so it doesn’t look like the home run is any more important now than it was in the recent past.
Even at the two high points of 0.66 and 0.50, the correlation is not very strong. The mean correlation coefficient for the eight-year period is just 0.40, which suggests a pretty weak relationship.
Further research is needed to better understand the correlation. The methodologies used in this post completely ignored pitching, defense, and other factors. By looking only at home runs hit, teams with extreme hitter- or pitcher-friendly ballparks likely skewed the results.
For example, the Cincinnati Reds play in an extreme hitter-friendly park, and they’ve hit the sixth-most home runs in baseball this year (95). However, they’ve also allowed 103 home runs, tied for the most in baseball. These high figures most likely have at least something to do with their small home park, yet for the purposes of this article, their home park was irrelevant. To more precisely study the relationship between home runs and winning percentage, the variables need to be better isolated. Perhaps it would have been more fruitful to study the correlation between home runs and runs scored.
Working with what we have, it appears that teams don’t need to hit lots of homers to win, even in this record-setting home run environment. Even though the relationship has gotten stronger every year since 2014, to the point where it has more than doubled in that span, it’s still not very strong and we saw similarly high correlation coefficients in the early part of the decade when home run rates were relatively low.
Looking back at the 2017 chart, there are several teams over .500 with average or worse home run totals, and there are several teams are under .500 with above-average home run totals.
While teams like the Yankees have enjoyed unexpected success this year fueled by the long ball, their way of winning isn’t the only one that works. Teams can still succeed in 2017 without hitting heaps of homers.