Sabermetrics of Education
As a baseball fan in Chicago, more specifically a Cubs fan, I have fully embraced the sabermetrics revolution. For readers who are unfamiliar with sabermetrics, the Society for American Baseball Research provides a useful guide with an extensive amount of information. Sites like Fangraphs provide a glossary of all of the terms/stats used beyond the basics like RBIs, home runs, and strikeouts. The basic idea behind sabermetrics is to use extensive amounts of data to find the best possible baseball players for your team and to evaluate players objectively. Theo Epstein, the man behind the recent World Series Champion Cubs team, has become sort of a poster child (poster adult? poster man?) for using these kinds of statistics to build a team. Sabermetrics help get an entire picture of a player instead of just the narrow picture that older, more traditional stats may paint.
A perfect example of a player who is loved by these new stats is Aaron Judge of the New York Yankees. Judge led the league this year in WAR (a sort-of catch-all stat designed to show how valuable a player is to their team), but also lead the league in strikeouts with 208. He struck out over 30% of the time he came up to bat. In years past, Judge might have been passed up, or his success dismissed entirely, because of how often he strikes out, but it’s 2017, and he was regarded as one of the best players in the league.
How does this all connect to education? It’s simple: traditional measures of student success shouldn’t be regarded as highly as they are. Grades and test scores don’t tell the entire picture of a student, especially if those things are looked at independently. One bad test score doesn’t make a bad student. If you look at things in a vacuum, there are some dangerous conclusions you can come to. Educators have the ability to collect, visualize, and analyze an extremely large amount of data about students, but we rarely do so outside of the traditional metrics within common datasets like grades (passing/failing), attendance (present rate), test scores (meets or doesn’t meet expectations), and behavior (# negative incidents). These things are looked at independently instead of together as a whole picture.
There are many things that can affect student performance. Behavior, for one. Some students fail assignments and/or tests due to constant class absences due to spending significant time in the office for behavioral reasons. However…he/she could have been acting out because they come to school hungry every morning. Looking at multiple factors as part of a whole, instead of looking at things individually can really help us better understand and help our students. Students and student data, like baseball players and baseball statistics, are complex and not always simple to fully grasp without digging deep into the numbers and truly knowing the individual. You don’t want to pass up the Aaron Judge who may be at your school just because that student scores poorly on one “traditional” metric.
Looking at all the possible information is something that is being used in all industries — not just sports. Businesses do it all the time and that was a big motivator for the sports world. Sabermetrics, or at least the idea behind them, is becoming something in hockey, football, basketball, and others. Education has lagged a bit behind, but it is catching up. By looking at the whole picture of a student’s data, educators can start to catch up to people in other industries. If you have any questions about how to get started with this, feel free to reach out! We’d love to talk to you about it.
Kevin Shane
Director of Professional Development
kevin@learnmetrics.com
What is your current ed data product ecosystem?
