Introduction

Baseball is one of the oldest sports in the United States and is widely considered “America’s Pastime”. The game has changed over the years, but at its core it remains the same as it was when Major League Baseball (MLB) was founded in 1903 (National League – 1876, American League – 1901). Thousands of baseball players have had the privilege of stepping up to the batter’s box, looking out into the stands, and then trying to hit a ball coming right at them. The majority of us are relegated to watching these players live out our childhood dreams. Not necessarily because we lack the passion, dedication, and focus, but because you also have to be highly skilled at the game of baseball. The old saying “Those who can’t do, teach!” comes to mind when thinking about our own love affairs with baseball, but with a slight twist: “Those who can’t play, analyze!”.

Back in the 1900’s (before the internet) fans had to read newspapers to get caught up on the latest stats and box scores. When discussing the top players of the game and/or how those players are getting screwed by the umpires, it required watching a ton of games yourself and talking with tons of other fans (data sources). No matter what, you’d always have some bias thrown into these discussions so unless you are talking about the dominance of Babe Ruth, Mickey Mantle, Ted Williams, or Nolan Ryan. A fan’s impression of an umpire could be made up by a single call they feel they got wrong.

Fast forward to 2021. It’s absolutely incredible how much historical data is available and even more new data is being captured for the first time in 2021. Fan discussions can be steeped in statistics. The funny thing is that even with all of these advanced analytics and immense amount of data, the age-old debates still rage on. Who is the best player in baseball? What makes a Hall of Famer and who among today’s elite will make it to Cooperstown? If Brian O’Nora wasn’t behind the plate in the bottom of the 10^th inning of Game Four of the 2012 World Series, would Miguel Cabrera still have been called out on strikes? These types of questions incite debate and lively, endless discussion among fans and baseball pundits. Without a ton of research on your own, you may never get a definite answer… until NOW!

As we created BaseballML.com, we wanted to develop new ways to analyze the game of baseball. There are many sites that analyze the game extremely well (and that we couldn’t possibly match), but there are areas that have been underrepresented. Let’s take a look at umpires for instance.

Statistics like strikeout rate, walk rate, and hit rate have been around for a while but how about going further and looking at how umpires actually call a game? Well, we created a tool that allows you to do just that. With this tool, you can directly compare two umpires and get a nice visual representation of their individual strike zone and even run a direct comparison to see how much more likely or less likely one up is to call a strike. Along the same lines, we also built a model to predict the outcome (Ball/Strike) of a given pitch location depending on many different game conditions (including umpire behind the plate). Now you can see whether going back in time to 2012 with an umpire other than Brian O’Nora makes a difference in the 10^th inning of Game Four.

Another discussion that has always fascinated us centers around “Who is the best?” and “Are they a Hall of Famer?” so we decided to tackle this as well. Leveraging some of the incredible work Bill James published, we created a new spin on player similarity exploration to allow interactive exploration of players across different points of their careers. We also built a model to allow you to see an individual player’s progress towards being a “likely” Hall of Famer. This allows you to play back the careers of some of the great players baseball has seen and see when they crossed the “threshold” to becoming a Hall of Famer.

Lastly, we wanted to explore how similar overall teams are created a way to compare two teams across history which leverage individual player similarity within a team atmosphere.

The baseball fan in all of us really enjoyed building all of this because it’s a super fun way to get lost in data and was a great fun setting up the infrastructure. Because we know how much fun we had playing with the data, we also built a mechanism for you, the readers, to download the data we meticulously scraped so that you can play around on your own. Hopefully your inner baseball nerd gets as much out of our creations, as we did while building them!