Dingerball

How different variables correlate to postseason success

Scroll down to content

As we inch closer and closer to the postseason I find myself wanting to write about it more and more. It’s an exciting time to be a baseball fan especially with how close the playoff race was.

So now that the postseason picture is set, let’s look into some numbers that may help us to guess what October may or may not look like. You will be able to see below that I correlated several regular season variables to post season success over the past ten season. So, the time that I will be evaluating will be from the 2008 season to the 2018 season. I chose that range because I wanted to have a large sample size but I also wanted to make sure that the data was current and applicable to today’s game.

The Structure of the Experiment

For each round that a team makes it to, they will receive a point. So for example, if they make it to the wild card round they get one point, if they make it to the divisional round they get two points, if they make it to the championship round they get three points, if the team makes it to the world series they receive four points, and finally if they win the world series they receive five points.

Before we get into the results, I want to make one point. This experiment is flawed. By giving a team a point between one and five to measure their postseason success just isn’t completely accurate. However, these results are still interesting to see. My point is, you can’t draw any hard conclusions from these results but inferences can be made and hopefully you can get a better understanding of how hard it is to predict post season success.

Winning%

r=0.2561

That’s definitely a number that I was expecting to be much higher. Especially since it’s like the deciding factor as to who gets into the post season in the first place. Part of the reason I think that it’s so low is that since the scale of 1 to 5 is so confined, it tends to lump teams together that may vary a little bit more in postseason success. More specifically, it doesn’t account for the difficulty of the opponent that they faced among other factors. Regardless, with all factors considered that’s definitely a lower number than I was expecting. I interpret this that postseason success is just very difficult to predict. If you don’t agree, just wait as just about every correlation I explore will be as low as this one.

Run Differential

r=0.3532

So, as you can see, run differential correlates to postseason success closer than winning% does. I definitely expected that because run differential is a very strong indicator when evaluating how good a team truly is. Take the 2018 Mariners for example, they finished the season with 89 wins and just missed the playoffs. However, they finished the season with a -34 run differential. So, the franchise realized that their success was a result of a fluke and instead of building on 2018, they disassembled their team and entered the rebuilding process to create something more sustainable.

Even though this correlation is significantly closer, it’s still not close enough to make any conclusions about how run differential translates to postseason success. Another indication that postseason success is so difficult to predict.

FIP

r=-0.024

Out of all of the metrics I correlated to “postseason success” this is the value that surprised me the most. Obviously, the negative correlation makes sense because the lower a team’s FIP, the more success they should have. However, it is surprising how insignificant the correlation is. Especially when you compare it to wOBA which you will see in just a moment. The only conclusion that I can come up with to explain this low value is that pitcher usage in the postseason is so different than in the regular season.

In the postseason we see an all out bullpen wars between teams. We also see pitchers start three games in a seven game series (Corey Kluber in the 2016 World Series). Basically, we see teams always putting their best possible pitcher on the mound at every moment which you just don’t see in the regular season and it makes sense as to why. The regular season is a 162 game marathon between teams when they know that not every game can be won. However, the postseason is another animal. It’s a sprint to the finish line with every game meaning more than the next. So, with the season on the line, why wouldn’t you want your best pitcher on the mound even though he just pitched maybe a few days ago. Simply, regular season pitcher usage just isn’t comparable to postseason pitcher usage. Which makes it hard to predict postseason success. Instead of looking at the team’s total FIP it would be more useful to look at the FIP and other metrics of a team’s top starters and relievers. That analysis is more subjective though and is an entire other article. It may be written though, so stay tuned.

wOBA

r=0.2263

Out of all the values measured, this correlation makes the most sense out of all the others. I’m not saying that it actually does make sense but that its value is more logical than all other variables that I measured in this article. wOBA does a great job at evaluating a team’s offensive production and to correlate a team’s average to postseason success, this is not a bad value. If we wanted to make the value higher, it would be better to measure just the wOBA of each of the player’s on the postseason roster. Even a step further, the wOBA of just the projected starting nine for each team because in the postseason we won’t see a lot of turnover in a teams lineup.

It’s interesting to see that wOBA has almost the same impact on postseason success as winning% does. That is definitely something that I would not have anticipated. Then again in order to win, you do have to hit the ball. I’d be interested to see how closely correlated wOBA is to regular season winning%. I think that would most definitely be a number above 0.5 consider they both equally correlate to postseason success in this experiment.

Conclusion

This was an interesting experiment with results that I say were not anticipated. As I mentioned earlier, no conclusions can be made from this experiment because there are several flaws with its validity. Regardless, the results are still very interesting to see and evaluate. I think it really helps to put into perspective just how difficult it is to predict postseason success. So, as the postseason arrives this Tuesday, just know that anything can happen.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s