Home runs are a pretty incredible thing when you really think about them. They involve a lot of skill, taking a stick and swinging it at a ball smaller than the size of your fist with enough force to send it 400+ feet through the air. That is a significant feat of strength on its own, and it doesn’t even consider that sometimes that little ball is going 100 mph or dropping eight inches from it’s original flight path. I personally find it hard to believe that anyone could ever hit a home run based on what we know about how a ball can act in the air, but I digress. It happens, and it’s beautiful.
Home runs as we all know are not only insane, but also insanely valuable. They are one of the most impactful events in all of sports, and the single most important event in MLB DFS. Knowing how to look for opportunities that could possibly lead to a home run is a very important thing to study, so when doing some correlation research, I set out to find the metrics most high correlated with the good ‘ol dinger.
Playing MLB DFS Today? Go Premium with DailyRoto
Correlations? More like Borelations
The general thoughts behind sifting through players to find good home run hitting candidates follow an intuitive pattern: Hitters need to hit the ball hard and into the air, while pitchers need to allow hitters to do so. That is a very watered-down version of most research done on each slate, but I wanted to go through the data and find if those were really the things that mattered the most.
Above are the correlation numbers for batters as they are related to hitting a bunch of home runs. A couple of these stats make sense based on what the stat was created to tell us: wRC (weighted Runs Created) is a stat that describes how many runs a hitter is responsible for based on playing time and home runs can create quite a few runs. wRAA (Weighted Runs Above Average) is another one that measures runs associated with a player. in this study, we are going to ignore wRC because it is based on playing time and messes with the sample, so the most important stats for hitters are:
- wRAA (.4937)
- ISO (.4478)
- SLG (.4108)
- OPS (.371)
- wRC+ (.3435)
The main takeaway from this for me: For batters, hitting the ball and striking it hard matter a lot, but it doesn’t matter whether the ball is being hit into the air as much as you would think. FB% and GB% definitely have some correlation here, (.1412 and .1398, respectively) but so long as a hitter is making good contact, they’ll send it out of the park at some point regardless of how often they hit a fly. For hitters, our focus should be on SLG and contact metrics, as the rest will generally follow suit.
But maybe they just get Lucky?
We have looked at what metrics matter the most to hitters when it comes to home runs, but what about pitchers?
I wanted to show this because it illustrates a pretty important idea that I think many people might ignore: home runs are kind of fluky, and when a pitcher gives up a big fly it’s generally not related to something he did, and the thing that matters is what the hitter was able to achieve on their swing. It’s glaring that the largest correlation when it comes to HR/9 for pitchers is a regressive stat in HR/FB, something that we normalize to the mean in application of advanced metrics. FIP and FIP- are the next highest correlations and both of those are HR/FB dependent, while xFIP (which has a very low correlation) normalizes those stats.
If we remove those regressive metrics from the sample, we can come up with something that shows the more intuitive side of the equation:
This shows what we would assume: Pitchers that allow a lot of contact and more importantly hard contact, while avoiding opposite field hits and line drives, will give up more home runs. The highest correlations for pitchers:
- Hard% (.1007)
- AVG (.0761)
- Pull% (.0686)
- FB% (.0579)
- GB/FB (.0524)
Another quick note on this: Pull% as a stat is something that is usually ignored, but it shows to be important. Hitters can generate more power when they pull based on the way they make contact and being in front of the pitch, so avoiding pitchers with higher Oppo% rates is a good way to get a little extra edge.
It’s not the Cheese, it’s the Whole Enchilada
No one statistic or metric alone is ever going to tell you exactly what you want to hear. Chances are, with all things in life, that you are going to need to dig a little deeper to find the answers (and the droids) that you are looking for. SLG and Hard% numbers are a good place to start looking for upside, but you have to combine them with a whole mess of other stats to help reduce your margin of error and really zero in on winning construction.
I have built a model based on the correlations that I’ve showed off in this article in a way to focus entirely on hitters with the upside we want to find in MLB DFS, and the results for the slate are as follows:
(HG= Hitter Grade out of 100)
There are a lot of ways to attack a DFS slate, and of course there is no obvious right or wrong answer when it comes to research. But, when you dig into a specific action and find what really builds into something you are looking for, you develop not only a better understanding of what you are looking for but also a process-based mindset. If you take anything from all of this, I hope you go into everything with a deconstructive way of thinking! And always learn to question what you have been taught.
Expect when it comes to the earth being a globe, it’s not flat. No need to question that.