Baseball has always been a game filled with tradition. The theatrics of opening day, the ceremonial singing of Take Me Out to the Ballgame, the hallowed walls of the Baseball Hall of Fame…it’s a pastime steeped in nostalgia. There has traditionally been a degree of “this is how it’s always been done” within the game.
As we know all too well in the education field, the “this is how it’s always been done” mindset is often paired with a resistance to change. Education is a field that everyone knows a little about because everyone has experienced schooling at some point in their lives. If a drastic change is made that makes school look different than it did when a student’s parent was a kid, that parent often tends to get uncomfortable and uneasy with the change.
Take summer vacation, for instance. The American school calendar was built on top of the agrarian calendar, when family farms needed their children home during the summer to help in the fields. Now, all these years later, we still follow a traditional calendar that has our students taking a lengthy break during the summer months, often leading to a mental “summer slide”. Why do we do this? Because we’ve always done it. It’s ingrained in our society.
But we have data that clearly showcases the problem. Research has shown that, on average, student achievement declines over summer break by about one-month’s worth of learning in reading and math. The evidence is there. The generally-accepted school calendar has a direct negative impact on reading and math achievement, the very areas that our country consistently cites as the biggest failures of the American school system. Yet, for the most part, the calendar remains unchanged. Summer vacation exists, as it always has.
Baseball was stuck in a similar rut for years. The sport was played in a specific way and, aside from small diversions, nothing really changed. Major League Baseball teams constructed a roster containing 5 pitchers who were designated as “starters” and who were tasked with pitching every 5 days and lasting as long as their effectiveness could hold out. For years, teams employed a “closer” whose job was to pitch the final inning of a close game. The batting lineup consisted of a “leadoff man” who was fast and good at stealing bases and a “cleanup hitter” who could bat 4th and potentially hit a home run to bring in the runners who got on base before him. It was very infrequent for a team to drift away from these conventional strategies. Little consideration was given to approaching things differently.
At the turn of the 21st century, something happened in baseball that began to change the tide. The Oakland A’s, one of the poorest teams in Major League Baseball, hired Billy Beane as their general manager, tasked with putting together a team that could compete with financial giants like the New York Yankees. Beane knew that they weren’t competing on the same playing field. Teams like the Yankees had resources to sign the most prolific power hitters, the hardest-throwing fastball pitchers, and, in general, the most popular players in the league. Many of Oakland’s young talent would even bolt for these top clubs as soon as they reached free agency.
Beane and the Oakland A’s had to try something different. They had to find a most cost-efficient way to win. To do this, Beane embraced the underappreciated and often underutilized side of baseball: the data.
Historically, the stats that mattered have been things like
- Average (AVG) – A calculation of hits divided by at-bats, while not accounting for walks or hit-by-pitches
- Runs Batted In (RBI) – The number of times a batter’s actions result in a run being scored
- Home Runs (HR) – The number of times a batter circles the bases and scores in one at-bat
- Earned Run Average (ERA) – The average number of earned runs a pitcher allows per 9 innings
- Strike Outs (K) – The number of times a pitcher throws three strikes to a batter, resulting in an out
- Errors (E) – The number of times a field flubs a play that should have resulted in an out, as determined by the official scorer.
- Stolen Bases (SB) – The number of times a base runner advances without the aid of a hit, walk, or error.
These were the stats that were discussed during broadcasts and placed on the back of baseball cards. Players who excelled in these stats were often rewarded financially. So, in order to be competitive, Beane had to find value outside of these stats that would result in wins.
What happened next was chronicled by financial journalist Michael Lewis in his 2003 book, Moneyball: The Art of Winning an Unfair Game. Beane utilized an approach known as “sabermetrics” to uncover hidden aspects of the game and construct a winning team with less money. Basic sabermetric stats that are now commonplace in the modern game include
- On-Base Percentage (OBP) – A measure of how often a batter reaches base per plate appearance, this time incorporating hits, walks, and hit-by-pitches.
- Slugging Percentage (SLG) – A measure of the total bases earned per at-bat, not just singles.
- Walks + Hits Per Inning Pitched (WHIP) – A measure of the total number of baserunners a pitcher allows per inning.
- Fielding Independent Pitching (FIP) – A measure isolating outcomes that are directly under the pitcher’s control and are not impacted by their fielders’ effectiveness, such as strikeouts, walks, home runs, and hit batters.
- Zone Rating (ZR) – A measure of how many plays a fielder makes when the ball is hit directly into their position’s zone.
- Wins Above Replacement (WAR) – A measure that tries to sum up a player’s overall value by estimating how many more wins they bring to the team than a replacement-level player would.
These analytical methods were not necessarily new, but they were not in heavy use within the game. The A’s began employing them and success followed, resulting in the popularization within the sport.
Suddenly, conventional wisdom was challenged. Teams began to wonder if 5 starting pitchers was the right strategy. What if they utilized 4 starting pitchers and tried to keep their pitch counts down during each game to keep them fresh? The data showed that, as pitch counts got higher, effectiveness dropped. So why not hand it over to the relievers at this point? And why were they saving their best reliever for the “closer” role? There are other points in the game where the best pitcher in the bullpen might be useful. Why were they saving him for a potential situation that might not even arrive at the end of the game? And why did lineups have to be led off by a speedy batter? Isn’t merely getting on base the more important thing? That’s what the data showed. Why not someone who takes a lot of walks? And why did they have to always bat their power hitter 4th? Why not move someone that valuable up in the lineup, which would result in them seeing more at-bats throughout the course of the season.
Embracing these data-based methods to guide player evaluation, financial decision-making, and in-game strategy challenged tradition within the game. There was a lot of resistance from the old guard and there was ridicule from fans and the media, but the results could speak for themselves. Moneyball became a bestseller, even resulting in a Hollywood adaptation of the story starring Brad Pitt as Beane. It wasn’t long before every team in the MLB opened a data science department and employed Ivy League-educated data analysts to guide important decisions on and off the field.
***
Education is in desperate need of a Moneyball moment. Compared to other industries (baseball included), the education field has lagged behind in utilizing powerful data and evidence to make decisions that will result in stronger outcomes for the students. There is regular talk of data-based approaches to student learning and many data collection methods are employed, but like baseball in the pre-Moneyball era, we still find ourselves embracing tradition.
In 2002, the controversial No Child Left Behind (NCLB) Act was signed by George W. Bush and had a lasting effect on education data within the United States. States were required to administer annual standardized testing for specific grades in all federally funded public schools and grade the schools on their results. There were specific provisions for addressing the achievement gaps present between student groups, which meant that additional disaggregation of the data was required to accurately understand the issues. To communicate all of these new data to the public, the act required states to develop reporting systems that could be easily accessed by parents, community members, and researchers.
It was a monumental event for the education field, resulting in an abundance of new data streams flowing from schools into the public sphere. But it was not without controversy as many decried the impact of overtesting on American children and insisted that students amounted to more than a number. Still, it was a big step toward quantifying education in a way that hadn’t been done before and bringing a new data mindset to the field.
But yet, the education sector lags in data usage compared to other industries. Overworked teachers in underfunded schools don’t necessarily have the time to devote to data analysis. Many schools face challenges with integrating all their various administrative data systems, leading to data that is fragmented and inaccessible to those who need it. There is constant apprehension about student data privacy, with many unsure how or unwilling to tackle it. Colleges of education have also been slow to incorporate data literacy into their curriculums, producing teachers that are not ready to turn the data at their fingertips into concrete action plans. There is also frequent pushback on change, even if it is data-based. Whether it is from the government, the parents, or the unions, there is criticism when changes are attempted within systems that impact as many people as schools do.
As a result, schools tend to fall back to the traditional way of doing things, even though the world is changing around them. As socioeconomic gaps widen, technologies advance, and students continue to underperform, we still see a reliance on classic structures like lecture-based classrooms, percentage-driven A-F grading scales, age-based student groupings, and one-size-fits-all student learning opportunities.
Clint Hurdle, the data reform-minded former Rockies and Pirates manager and subject of the book Big Data Baseball, often reminded players, “Traditions can be wonderful and meaningful, but also a vision killer.” We, as collective educators of America’s youth, need to take a step back and think about why we do things the way we do them. Does the data support these methods? Does the data even exist to evaluate these methods?
In his book Ahead of the Curve, baseball journalist Brian Kenny examined the issue of bunting. For years, traditional baseball wisdom stated that if the first batter of an inning reaches first base, a solid strategy was for the next batter to bunt. This move would typically result in the batter being thrown out at first, but the runner advancing to second base, where he should be able to more easily score if a future batter gets a base hit. This was a strategic move that became expected and was often seen as the right call during these scenarios. But Kenny examined the data over decades worth of games:
Man on 1st base; no outs
- Runs expected: 0.94
- Percent chance of scoring: 44.1%
Man on 2nd base; one out
- Runs Expected: 0.72
- Percent chance of scoring: 41.8%
The numbers clearly show that the conventional baseball wisdom was not supported by data! The managers were making a decision that actually put their team in a worse position than if they would have just let the next batter approach the at-bat normally. Even considering the possibility of a double play or a poor batter at the plate, the numbers still showed the strategy to be faulty. This data was not a secret. This was not some revolutionary new finding that could change the game. So why did managers keep making this decision?
Kenny referred to the phenomenon as “blame-free failure”. Leaders often see value in doing what is expected because, when failure does occur in these situations, the criticism will not be directed toward the leader’s decision-making. In the book, Kenny recounts the story of game 1 of the 2014 American League Division Series. The Angels, led by manager Mike Scioscia, had the best record in baseball and were matched with the Royals in the playoffs. In the first game, the Angels had their leadoff man on base in the 7th, 8th, and 9th innings. In every one of these innings, Scioscia made the decision to bunt the man over to 2nd base and, in every one of these innings, the team failed to score. They ended up losing that game and, eventually, were swept in the series. Did the media skewer Scioscia for making a decision that was not based on data? No! Scioscia did what had always been done. The truth is, if he hadn’t bunted the player over and they still didn’t score, this is when the fans and the media would have likely asked for his head on a platter.
Applying this concept to education makes one wonder about its prevalence. How often do we cling to tradition because we fear the backlash? When we do this, who are the ones that are burdened by the results? It’s the students; the very people that the system exists to help. We need our Moneyball moment so that the data can speak more loudly. We need a culture of objectivity and trust in the information we have collected. Without it, we’ll be stuck in tradition forever.
***
For our schools to improve, education leaders need to have the courage to embrace change. The best way to do this is to create a robust data system that will uncover truth, and to base all decisions on that data. Obviously there are barriers and bureaucracy that make change hard, but sound data that supports student achievement is difficult to ignore. If we have data that suggest student learning will improve as the result of a change and that change still isn’t being made, then what truly is the purpose of school? Why is everyone gathered in this school building for 8 hours a day and 5 days a week if not to help students learn?
I find myself thinking a lot about how we can apply baseball’s data framework to the education field to explore what lessons we can learn and how we might undergo a similar revolution as the sport. Yes, data does exist in our schools. Yes, NCLB upped the ante of data being collected across the nation. Yes, I’m sure you have at least considered purchasing local screeners, climate surveys, or advanced student information systems. But are you able to answer all the questions you have about your school and are you able to develop next steps that are grounded in what the data says?
There is no perfect number that sums up success. Various services might claim to give schools an overall “grade” that explains their effectiveness, but we all know these marks don’t contain the nuance of the system. For that reason, it is essential to gather and utilize as much data as necessary to guide decisions. Not all data collection is intrusive. Not all data collection has to result in public outcry about overtesting or teacher outcry about methods used for evaluation. Sometimes it’s just having information at the ready and finding new ways to look at it. Additional layers of analysis and comparison can lead to valuable insights on what schools are actually doing and how it is affecting the kids.
