To all devoted, and even casual Astros fans everywhere, let me proclaim: Read this book! Learn it. Live it. Love it.
Even if you think you already know everything there is to know about the Astros, this book will inform you. If you love the Astros, it will inspire you.
It has much to teach any other student of baseball as well. But if you are the general manager of another major league baseball team, I advise: Please don’t read this book.
For anyone interested in business management done right; this book should be required reading.
And futurists concerned that one-day artificial intelligence might possibly supplant human intelligence should take heart from the story of Astroball. Because, as author Ben Reiter, the prophet who predicted the Astros victory in the World Series in 2017 for Sports Illustrated back in 2014, says, the essence of Astroball is not the primacy of pure, cold, statistical analysis over human intuition. It is their integration.
What is Astroball. And What it is Not.
In Reiter’s words, “The Astros believed over reliance on Big Data could lead to middling and one-dimensional baseball teams...They synthesized human observations into their probabilistic models and made humans responsible for triaging and sometimes rejecting the results...Success is not a matter of man or machine, but man plus machine, as long as man remains in charge.”
When I started reading this book, I expected it to be like some book that should be called Beyond Moneyball, meaning that Astroball would be Moneyball squared, much more data, much more sophisticated data, even more reliance on data, perhaps nothing but data. Except for that last point to an extent, that all describes the Astros and Astroball. The Astros are definitely a data-driven franchise. But this is also true of many other teams in baseball by now as well.
What makes Astroball unique isn’t the reliance on data, but the integration of everything positive about human intellect, human genius, human inspiration, human intuition, the gut, as the author calls it, into the data calculus, and also using that calculus to ruthlessly exclude all the human negatives, human bias, prejudice and other forms of stupidity, as Moneyball does. Astroball endeavors to use everything good about human consciousness which helps us arrive at the truth, while eliminating everything wrong with it, to improve upon both pure statistics and human insight.
Moneyball taught us that human judgment, the gut, or the eye, or whatever you call it, is very often misleading or just wrong and seldom predictive. Computers and the cold laws of probability must guide decision making. The first users of the Moneyball model quickly developed a competitive advantage over teams that relied only on scouting, eyeball tests and traditional statistics to guide decisions. But when all the teams input the same data and get computerized results based on the same laws of probability, the competitive advantage disappears.
Astroball incorporates this aspect of Moneyball. It even subjects the judgments of scouts to computer analysis.
But Astroball also stresses that baseball players are humans, not machines, and that they don’t always behave in a probabilistic manner, and that it takes human insight to find those outliers that defy the odds. And doing so is the secret to excellence.
Astroball incorporates the kind of uniquely human intuitive insight that can discover or develop talents that computers might miss. The incorporation of the “gut,” properly filtered, into a single, unified, computerized analytical framework, along with the most sophisticated objective data available, is the secret that gives the Astros a competitive advantage, both in scouting and player development. Astroball adds human genius to the analytic capabilities of computers.
Of course every team still employs scouts who use their experience, their eyes, and their guts to make decisions. But what often happens is that when an organization has to make decisions, the analysis of the sabermetricians often comes into conflict with the gut of the scouts, leaving the front office at war within itself, having to choose heart or head. Jeff Luhnow experienced this at St Louis, and interestingly, often chose the gut successfully. Astroball incorporates the numbers and the intuition into a single analytical process, satisfying to both heart and head, scout and statistician.
Moneyball assumes baseball decisions can all be reduced to pure probability, like whether to play another card in blackjack. Astroball is also highly probabilistic, but acknowledges the existence of human, perhaps unquantifiable but still real factors, “growth mindset” for example, that are also important considerations. And at this time in baseball history Astroball says the inclusion of these intangibles in the decision making process is often the difference between having an average team and having that extra something that makes a champion.
So, Astroball IS Beyond Moneyball, but in a qualitative sense, not just quantitatively.
I am tempted to summarize every chapter of this book it is so rich in interesting detail and insights. But I want you to read it so I will leave out a few spoilers.
I think the best chapter in the book, and the one that illustrates Reiter’s thesis the best, is Chapter 7, “In Search of Carlos Beltran.” It is not just about Carlos Beltran, the chapter could just as easily have been entitled “In Search of Team Chemistry.”
Around here at Crawfish Boxes there has been some debate about how important Carlos Beltran was last year to the Astros’ success. Was he just a replacement level DH, or were the intangibles he brought to the team a major, if unquantifiable, part of the teams’ success.
According to Reiter the intangibles Beltran brought to the team were indeed the latter, and furthermore, it was precisely for those intangibles that Beltran was brought to the Astros and paid the most of any player on the team. There were cheaper DH’s on the market who figured to hit as well or better than Beltran, but Jeff Luhnow and company wanted Beltran because of his demonstrated love of mentoring and how they believed he would help team chemistry.
Yes. Astroball values team chemistry. According to Reiter the Astros were aware of recent research in this field, including a study by economists with the Federal Reserve which had concluded that there could be a 20% variation in a team’s win percentage from season to season even if that team had the same players, and that 44% of that variation was due to what they called the David Ross effect. David Ross was a second string catcher who kept playing mainly because he was considered a team builder. In his day, the Astros scout Enos Cabell was considered such a player as well.
Other academic analyses showed the importance of organizational unity, the lack of debilitating social “fault lines”, as crucial to an organization’s or team’s success.
As Reiter put it, “one of Jeff Luhnow’s strengths was that despite the data-driven approach that led to the Astros turnaround, he realized that only a fool would ignore the potential value of the shit he didn’t know-the shit for which his data couldn’t account.”
As Luhnow put it, “Just because you can’t quantify it (team chemistry) doesn’t mean it doesn’t exist.”
Point is, creating positive team chemistry, drafting and promoting players who would add to it, using behavioral research to build it, is something the Astros do with intent. It was mainly for this, his mentoring, and his perceived ability to break down potential “fault lines” on the team, that according to Reiter, Beltran was signed. Carlos Correa gave Beltran credit for 7 of his 24 home runs last year. Reiter believes that if you spread that kind of contribution throughout the team, and Chapter 7 was devoted to showing just how much you could, then the $16 million that Beltran received last year was a huge bargain.
The Astros are the tightest, most fun, most loving and most lovable team in baseball and yet driven as well. As Reiter points out, the Astros were aware of research which showed that this kind of chemistry could be worth six wins. Where will you find a 6 WAR player that doesn’t cost a cent?
This book is replete with examples of how the “gut” was incorporated into the “process” to create positive decisions that pure mathematical analysis would have missed. According to pure physical analytics, Carlos Correa, Luhnow’s first #1 overall pick, graded out as the seventh choice in the 2012 draft, and in fact, the Astros found out later, he would not have been picked until #6 by San Diego and their Vice President of Scouting, A.J. Hinch. But the “gut” reports from the Astros’ area scout, and the personal interview of Correa by Luhnow and Enos Cabell, who was there specifically to evaluate character, bumped Correa’s grade to #1. Time has proven this to be correct in spades.
One reason the Astros did not mind too much losing out on Brady Aiken in the 2014 draft (it was owner Jim Crane, not Luhnow by the way, who decided to offer Aiken additional money after the failed physical and Luhnow’s initial low ball offer), was they already knew who they would draft with the #2 compensatory pick the next year. They knew that Dansby Swanson was the obvious computer choice, but as with Correa, they had the generally lesser valued Alex Bregman as their #1 even if no else did due to “gut” factors like growth mindset. (“This kid doesn’t give a shit about anything but playing ball.”) It’s not that these intangibles overrode the analytics, it’s that they were a part of the analytics.
The analytics department advised against the famous Justin Verlander trade according to Reiter. In this case Luhnow overrode that judgement because he knew under the circumstances the team, in August 2017, needed a boost in morale, yes chemistry, as did the city of Houston post-Harvey.
The Process is a Process
One reason I found this book inspiring as an Astros fan is that Astroball is not a dogma, but a process continuously subject to revision and improvement. The Astros do not think they have found the perfect “Astro way” for all time and eternity. They are constantly seeking to learn from their mistakes and revise the process and its underlying assumptions accordingly. They intend to continue to innovate and stay at the forefront of baseball science and thus maintain a dynasty. As a fan that is exciting. The management of the Astros are geniuses. Not just rocket scientists, engineers, economists, MBAs (they are) but baseball Renaissance men.
Examples of learning from mistakes: The Astros wanted to trade Dallas Keuchel before the 2014 trade deadline but instead traded Jarred Cosart, but only because the Florida Marlins insisted on Cosart. According to Reiter this would be disastrous near miss taught the Astros that a player such as Dallas Keuchel could be more than the sum of his physical skills. And a player like Cosart could be less.
Another lesson learned was the release of J. D. Martinez. About this most Astros fans I know say something like: “They gave Martinez two years to prove himself. He failed. He had to go. So what if he subsequently became one of baseball’s best hitters. It’s not the Astros’ fault. Stuff happens.”
Turns out the Astros don’t look at it like that at all. Martinez had told Luhnow during spring training that year that he had learned a transformative new swing that winter in Venezuela and that he was just asking for a good chance to prove it during spring training. He only got 18 at bats, mostly pinch hitting. Then he was cut.
As Reiter put it, “It quickly became clear that they had released, for nothing in return, the very thing the process was designed to find: a cheap superstar.” As a result, since then the Astros have developed technology that can measure whether such a claim as Martinez made to Luhnow has an element of truth.
And according to Reiter, the fact that Bo Porter did not give Martinez more chances in spring training, 2014, factored into the decision to fire him. The Astros want open minds.
But the main lesson was this: “The feeling that we are smart is our enemy.” So said Sig Mejdal, Luhnow’s chief cohort since St. Louis days. A major component of the Astros “process” is the humility to know that the process itself must constantly improve and to foster an environment where experimentation with new ideas is encouraged.
One or two quibbles with the book. Much time is spent praising the way the Astros’ “nerd cave” and its advanced metrics have helped pitchers like Collin McHugh, Dallas Keuchel, Charlie Morton, Gerrit Cole, even Justin Verlander, find their full potential. Pitching coach Brent Strom, an iconoclast at the forefront of the revolution in pitching science and brought to Houston by Jeff Luhnow early on, clearly deserves credit for much of the success of the pitching staff, and deserves mention in any book about Astroball. His exclusion from this book is a major oversight.
Reiter tended to gloss over any mistakes or shortcomings of the system. For example, he never asked why did J.D. Martinez need to go to Venezuela to find help with his swing, which even the Astros staff knew was a mess, but couldn’t effectively change? That was early in the reign of Astroball, so one would hope they have better analytics now in the field of hitting mechanics.
Reiter spent a lot of time discussing the Brady Aiken draft. It was a very unconventional pick, and because it turned out he had a flawed elbow, we will never know if it would have been another brilliant success of the process, like the draft of Correa or Bregman. But as it turned out the conventional wisdom was correct: you don’t draft a high school pitcher 1:1. Too much bad stuff can happen before he’s ready for the Bigs, including arm injuries. Maybe the process got too cute on that one. It worked out well in the end though, thank goodness.
In the previous draft the front office made a very conventional pick, Mark Appel, and that didn’t work out too well either. No front office is going to draw 21 every time.
I’ll let Reiter conclude: “Were the Astros run by computer, they would probably never have drafted a high school shortstop out of Puerto Rico, or retained a five-foot-five second baseman, or signed a 40- year old free agent, or traded for an aging pitcher who made $20 million a year...Data could help guide best practices, but it was unwise to confuse those with perfect practices. If people who denied the power of data could no longer compete, neither could those who believed that data alone provided an answer, not a tool. ‘All models are wrong,’ the British statistician George E. P. Box once wrote. ‘But some are useful.’”
“There would, in other words, always be a place for human intelligence, alongside the artificial kind, and not just in baseball. There would always be a role for gut feels.”
Great book. Like the Astros themselves; full of heart and brains. It made me remember why I’ve always loved baseball. I read it in a day.