Everyone is aware of the rise of analytics across all industries, whether healthcare, business, education, or sports. The application of analytics has drastically affected evaluation and decision-making processes within organizations and individuals, leading to innovative growth and achievement.
Predictive analytics has allowed for higher success in asset acquisition and completely changed decision-making science. However, as the science and construction of analytics have evolved with new models and statistical development, the application of analytics has somewhat become oversaturated in nature.
Models and analytical evaluations can act as better decision-makers than humans, but humans dictate the inputs and data development. If we become overly reliant on analytics without considering what decisions they are making and the criteria they're using, we lose focus of the intended goal.
I want to touch on how analytics has dramatically changed the landscape of sports, specifically baseball.
Baseball Analytics
For the betterment of baseball, analytics has created a whole new world within the sport. It is hard to remember when advanced baseball analytics didn't have a home in the game. Baseball analytics has transformed the game's player evaluation and development aspects, creating new ways to measure and quantify talent. This is shown by developing statistics such as WAR, WRC, wOBA, etc. We can micro-analyze nearly all aspects of the game but sometimes lose our direction in analyzing the data independent of the presented context.
The evolution of a baseball evaluator and analyst has come a long way as teams are now hiring and seeking talent that can apply data and technical skills to the context of baseball. The issue at hand for all evaluators and teams is creating the correct blend of analytics and scouting. To act skewed to either side is naive in nature and contributes to a flawed evaluation process.
Over my time as an amateur baseball evaluator, I have learned methods to enhance my player evaluation by applying aspects of baseball analytics to scouting itself. In its raw nature, analytics can help quantify or describe something that the human eye can't. Common examples include pitch and batted ball metrics.
Scouting vernacular can be tied to events and characteristics described by metrics, creating a cohesive blend backed up by both sides. This shows how baseball analytics and scouting can be bounced off one another to support claims in player evaluation. I want to question the context of analytics within a player evaluation.
Baseball has come to a general understanding of quantifying good and bad with analytical interpretations, which is an essential step in the process. The next step is contextualizing the analytical evaluations through a scouting eye.
After understanding what's good and bad, we have shifted to weighting statistics on different levels. Essentially, certain aspects matter more than others, when evaluating. Within these aspects, teams construct their player development departments to refine what can be refined. Research and PD have shown it's likelier that a hitter can improve their swing decisions and process rather than their bat-to-ball skills.
While on the flipside for pitchers, pitch design is inherently known as a loud and present science. Teams feel more confident developing pitchers with good base traits who have room to go with their "stuff." In essence, player development can help make players more dynamic instead of skilled. Thus, it is up to teams to develop their organizational scouting and evaluation philosophy to funnel to their PD pipeline.
Tradeoffs are also essential to evaluate within baseball analytics. At what threshold do you deem a hitter's batted ball data acceptable, at the expense of high whiff rates? How much is too much? These questions are asked within departments and evaluators not to get blinded by the story the numbers tell.
Scouts are necessary to validate possible data suggestions as they consider a player's movements, body, age, and intangibles. As in the case of pitchers, if I run a stuff model and a player grades out highly, but said player had had poor reports on his body and mechanics, all while failing to throw pitches over the plate, is he worth the investment? The threshold for leaning into the stuff model at the expense of all other traits is suspect. The next market of scouting and evaluation within teams is leaning into their PD departments to learn what's attainable to help a player refine his skillset and application.
I wanted to write about this topic because, at times, I feel the data is blinding evaluators and shouldn't be used as a lead all resource, while it should be relied on heavily.
Skewing data to fit one's narrative of a player also presents a naive evaluation process. I'm excited to see how analytics continue to transcend baseball and player evaluation itself.
Human inputs and what we deem important, good and bad, will continue to evolve when evaluating talent. The most crucial part is continuing to refine the process, while keeping an open mind to anything.
Remember that data is objective, but the context we choose to interpret and apply it is subjective
.