Spring Dingers Almost Meaningless

If you are an optimist, the word "almost" probably jumped out at you in the title.

The Mariners have hit more home runs than any other team in spring training so far. The Mariners made a concerted effort to beef up their offense with some power hitters, so the early onslaught seems like a promising sign.

I decided to see if spring training dingers have translated into regular season dingers in recent history. Below are team home run totals in spring training vs. team home run totals in the regular season for the past four seasons, 2009-2012. I stopped in 2009 for no great reason; it gave me over 100 data points and was also the season Jack Zduriencik took over the Mariners.

Here's the data (click for larger image):

Correlation coefficient: 0.18

The data points look more or less like a random blob with perhaps a slight upward trend. The correlation coefficient confirmed what is obvious to our eyes - there is a slight positive correlation (suggesting teams that hit more home runs in spring do hit more home runs in the regular season), but it's impossible to stress slight too much. Barely any of the variation in regular season home runs is explained by spring training home runs in theory.

I got to thinking about the Mariners situation though. They go from playing in the warm, dry air of Arizona to the cool, damp confines of the Pacific Northwest. I wondered if park factors were obscuring a more noticeable trend in the data.


Luckily, we live in an age where we can splice data quickly and easily. Considering only away games in the regular season is a quick way to minimize park factor differences. It is far from a perfect method, given the unbalanced schedules that MLB teams play, but it's good enough for our purposes. If park factors are masking a trend between spring training and regular season home runs, then taking out the home park should push the correlation coefficient upward, even though the result still wouldn't represent the most precise relationship we could concoct.

Here's the data (click for larger image):

Correlation coefficient: 0.15

The data for away games only generated a scatter plot eerily reminiscent of the first one. Upon closer inspection, via the correlation coefficient, it is actually spread even more randomly. That pretty much puts to bed any thoughts that home park factors, and park factors in general, obscured the relationship between spring training home runs and regular season ones.

I am happy that the Mariners are hitting so many home runs early on in spring training, but the data strongly suggests they are no harbinger of things to come. It isn't as if it's a bad thing that the Mariners are hitting so many dingers, but it's not a good thing either. It's just a thing, more or less, and it makes checking my e-mail/folding my laundry/cleaning my sink more enjoyable on a drab Saturday afternoon in March. It just turns that's about all the spring training dingers are good for.

With that said, the home runs are still a positive sign, especially when put in a broader context. There are other reasons to think the Mariners will hit more home runs in 2013 - namely 1) Safeco Field played incredibly cavernous last year, 2) the fences are coming in, and 3) the team acquired some hitters with histories of hitting more dingers than the M's we've grown accustomed to. Maybe none of these factors on their own hold much predictive power, but perhaps together they say something worth noting, since they all point past the fences.