By now most sports fans have heard of the “Moneyball” approach to building a winning sports team, and have probably also noticed the increasing number of statistical categories used today to measure a player’s overall talents and abilities. In fact, some professional teams today have entire departments that do nothing more than number crunch and interpret data, with the goal being to use the most advanced analytical techniques to assemble a winning team. On paper, the analytical approach looks like the way to go, but does this method really provide better insights than simply having human experts “eyeball” talent?
Understanding statistics
Statistics are an incredibly powerful tool to use in countless social/behavioral situations, and experts who use statistics are often able to provide greater depth and breadth when trying to understand human thinking and behavior. Using statistics can also sometimes help us better predict the future, especially when stats are used in empirical research studies by academic experts trained in using and understanding scientific data. Still, statistics are not perfect, and even the best, most controlled studies always risk extraneous variables interfering with findings.
Statistics have been used in sports for many years, and when this kind of data is combined with human observations and expertise it can lead to a more thorough, comprehensive understanding of people and situations. The problem occurs, however, when people use statistics exclusively and at the expense of human experts — this over-reliance on statistical data can sometimes miss characteristics not easily measured by traditional means.
Where “Moneyball” approaches miss
Front office professionals use analytics as an aid to speed up the process of building championship teams, but there are many facets of human thinking and behavior that are either poorly measured by use of stats, or not measured at all. For example, the following characteristics are often missed, misinterpreted, minimized, or ignored when relying on stats:
- Heart/motivation. Statistics do very little when it comes to a player’s motivation and his or her desire to prove others wrong.
- Team dynamics. Often statistics are collected in lab situations, not taking in to account real-life situations like how a player will interact with the personalities of individual players on the team, or even the coach.
- Potential future growth. How a player measures today may not be how he or she does tomorrow, and we also don’t know the potential bandwidth for future success/failure based on limited statistical measures.
- Errors with data for prediction purposes. While a great 40-time might catch the attention of football scouts, assuming the time alone is a strong predictor of future running back success may not be of much predictive validity, if any at all.
The impact of extraneous variables.
An extraneous variable in science is described as undesirable variables that influence the relationship between the variables that an experimenter is examining. A couple examples of variables that aren’t usually factored into analyzing future athletic potential are injuries, and athletic performance post big financial contract. With injuries, they are unplanned and can be bad enough to force an athlete into an early sport retirement. Athletes who sign big financial contracts might see their motivation decline because they have been awarded guaranteed money and no longer need to perform their best. Extraneous variables are difficult to control or predict, meaning we should use caution when jumping to conclusions relating to cause-effect relationships.
Analytics, or experts?
Back to the basic question of the impact of analytics in sports the best my advice is to continue to rely on expert appraisals (i.e. coaches with history in the sport), while using analytics to potentially “round out” the profile. The danger occurs when the opposite happens and teams over-rely on stats at the expense of using sound human expertise and judgement. There’s obviously an allure to what data can provide, especially if it means finding better answers at less cost of time and money. On the other hand, direct observation, intuition, wisdom, and critical thinking are terrific “human tools” not to overlook.
Final thoughts
Even though we hear more about analytics at the pro sport level than anywhere else, it’s only a matter of time before user-friendly technologies will be seen more regularly at the interscholastic and youth-sport levels. Remember, human beings aren’t robots, meaning even the best tests won’t perfectly capture every facet of human thinking, emotions, and behaviors.
drstankovich.com
