In reading Joe Maddon’s Book, The Book of Joe,
My take is my own, I am not trying to put words in Joe
Maddon’s voice. I am writing to figure out what I think about the subject. All
interpretations and reasoning, as well as errors are mine unless noted.
heu·ris·tic: involving or serving as an aid to learning,
discovery, or problem-solving by experimental and especially trial-and-error methods
·
heuristic techniques
·
a heuristic assumption
also: of or relating to exploratory problem-solving
techniques that utilize self-educating techniques (such as the evaluation of feedback) to improve performance.
I have consistently
used heuristics throughout my career as an engineer. One reason that people in
the STEM world uses heuristics is because heuristics in the sciences are
usually buttressed and scaffolded by the natural laws of sciences. Natural
constraints have already been woven into the heuristics. The problem with using
heuristics to solve problems that are brimming with human interactions is that human
decision making is nonlinear and random, which adds an infinite order of
complexity to decision making. Using heuristics in a situation that is nonlinear
and random could be compared to fitting a square peg in a round hole: the
solution may not only be erroneous, but more importantly, it may exacerbate the
problem exponentially.
Heuristics
can be described as a System 1 response, per Kahneman and Tversky
Another
consideration is that in using heuristics, we are often using a general
solution, something that we have cobbled together from numerous experiences
into a single heuristic. Contexts are not usually considered when using
heuristics. In short, we are using the generic case to solve problems that are specific
and demand unique situations; that is part of the attraction of using heuristics,
the assumption that good enough is good enough. We assume those pesky details to
go away because we don’t have enough time to delve into the granularities.
In the
pre-Moneyball era of sports, gutfeel and intuitions were the ingredients that
make up heuristics. As such, those were the heuristics that statistical
analysis was used to dispel. Analytics was the gutfeel and intuition heuristic
buster.
In Joe
Maddon’s case, he was an early adapter to statistical analysis. He embraced analytics,
which was the reason that he was hired as a major league manager. The analytics departments of the three teams
he managed churned out reams of data for him to use. In time, however he came
to the realization that analytics is just one tool, one very potent tool, but
just A tool to use.
In the
short time that Maddon worked as a major league manager, a relatively short
tenure when compared with his long tenure working as a coach and a scout, the
accepted decision making emphasis had
turned itself on its head. Because of Moneyball, managing became completely
about numbers and following the algorithmic routine of processing the number,
do what the number says, and never question the numbers. Those who manage had
to defend every decision by reciting the all-important analytics.
One
phrase that I often use to describe people’s attitudes towards sacred cows is: the
biggest and most vocal anti-establishment rebels will often become the biggest new
establishment dogmatist after the ancient regime has been overturned. Those who
are most passionate about their beliefs have skin in the game, which make them
the most vociferous and fierce about their beliefs, so much so that they become
as intractable as those who they had deposed.
In many
ways they suffer from the sunken-cost fallacy: they have bought into their beliefs
that they feel they must not betray their side of the argument that they become
zealots. They will automatically dispense
with debate and become just as entrenched and rooted in their heuristics as
those they had opposed in the old establishment. This is what I see in the debate between the analytics
proponents and the intuition proponents. Because of this dedication to the sunken
cost fallacy, we have a dichotomous
situation that is reflective of the ethos of our society.
Maddon
has an interesting, and I believe, sane approach to all this. He thinks that
the analytics department is generating too much information, information that is
so overwhelming in volume while also containing too much noise. The usual
safeguard, in terms of statistical correlation thresholds have been ignored because
of the amount of data generated by data mining, ANY correlation is deemed significant.
In other words, the decision makers are chasing after noise. His other concern
is that in the flood of information, the decision maker must be able to discern
quickly, and accurately; which pieces of information are pertinent, and which
are extraneous; an impossibility given the minuteness of the time scale and the
immensity of the amount of analytic information. Indeed, he wrote in the book
that even as he is a proponent of analytics, he wants less information when it
comes to crunch time: when the decision maker has the most at stake, or when
the situational pressure and stress is at its maximum. He explains the reasons
for his preference by pointing out the following: the sample space for the data
from high pressure and stress situations are minimal, not enough to meet the
statistical significance threshold for decision making; he also believes that people
behave differently when under high pressure and stress, therefor any data that
is collected will be uncertain, noisy, and not predictive .
He
makes the point of saying that he mostly manages with his instincts during the
most critical situations because he does not believe that more data is helpful,
instead too much data is more detrimental than helpful. This is where he makes
an interesting observation. He thinks of his instincts as thinking in advance;
that is, the accumulated experiences with different situations, in different
contexts, and with different people are all a part of his thinking in advance.
I suspect that the thinking that he does prior to a World Series game has been
scrubbed of any heuristics from his experience as he is mindful of their
presence and the fallacious decision that may result when he employs those
heuristics. He states that he prepares for those situations by reviewing the strategy
and tactics that are unusual and unexpected, just in case.
The
other part of his explanation is that the manager does not and should not
operate in a vacuum when it comes to the humanity of his players. Part of his
knowledge and experience comes from his interaction and history with his
players. He should understand their psychological makeup and he has experienced
their lives during the time that they have worked together while under the
glare of competition. Maddon emphasizes that he makes a point of knowing and understanding
all his players. His “instincts” about the humans that he is working with are
worth much more than the reams of soulless data because decision making with
data is just data, coaching a team of people is an art. This art is the expression
of human instincts when dealing with other humans. Indeed, he makes the
eloquent point that analytics and instincts should be used and applied
according to the situation and context. Data guides strategy, and the art is
belief in the human element, the best solution needs to be both. The magic is
in the proportion of each and how they couple and synthesize into new knowledge.
I
remember when I was learning about coaching, a wise mentor made a distinct
point: we are not coaching volleyball, we are coaching people, volleyball is
just the context. That statement has a myriad of meanings. In one instance, it
is an admonishment for the coach to remember to treat people as people while
coaching; in another instance, it is an admonishment for the coach to trust the
human ability to execute the actions and make the decisions no matter the circumstances.
Volleyball Heuristics
This part was actually fun. I thought about some of the
heuristics that I had experienced either as a coach or as a spectator. These heuristics
are not completely false, nor am I insinuating that these heuristics cannot be
true under the correct context and situations. They are heuristics because they
have been true under specific circumstances, but they are not true generically.
This is obviously not a definitive list.
In Training
· Players will always get better when they play more, so just let them play more and never work on fundamentals.
· Players can only get better by doing an infinite number of identical reps without variation.
· Beginning players can only get better by doing an infinite number of completely random reps without seeing the same conditions twice.
· Players should be able to figure out: strategy, tactics, technique, communications, VBIQ without feedback.
· Drilling with no stated goals is the best way to give players reps.
· Players should only practice their assigned positions.
o Liberos and DS should never hit in practice.
o Middles should never get practice time on setting or passing.
o Setters should only set.
· Extended scrimmages starting from zero is good preparation for play.
· Extended scrimmages with the same scrimmage partner is good preparation for play.
· Introducing extraneous distractions into drills prepares the players for play.
· Slapping the ball is a good starting cue to start the rep, especially for hitter timing.
In Competition:
Many of our competition heuristics come from our need to use
some statistical measure. Usually, we use intermediate metrics since using the
score as a measure is too broad and too final. The problem with using
intermediate measures is that we are measuring an action that is only a part of
game action. We are diverting our attention from the goal, which is to win at
the end of a set or a match. We instead let ourselves get head faked into
chasing measures that are merely indicators and never predictors of success. For
more on this please see Coach Jim Stones excellent article (https://jimstoneconsulting.com/if-coaches-dont-know-goodharts-law-they-should/)
as well as my own explanation: (https://polymathtobe.blogspot.com/2022/07/stats-for-spikes-goodharts-law-and.html)
· The teams will always win when team passing scores are above 2.5.
· The correct serving ace to errors ratio hovers around 1.
· Number of service errors are the best predictors of service aggression.
· Regression to the mean predicts that the trend will immediately correct itself to the positive result after a string of negative results.
· A definite and predetermined distribution of sets to hitting positions is necessary for success.
· Always run middle on a good pass.
· Always set the pins or backrow on a bad pass.
· Hitters should only hit their position.
· Never run a set play off of a dig.
· Never run a set play out of system.
· Only run slide plays behind the setter.
· Setters should tip the ball only on bad passes.
· Setters should always dump the ball immediately after the opposing team gets a dump kill on your team.
· Never serve the Libero.
· Always serve the substitute player that just came in.
· Serving and passing are the only things that matter in the women’s game.
· Blocking and hitting are the only things that matter in the men’s game.
· Defense wins games.
· Playing against weaker teams helps team confidence.
· Playing against stronger teams helps team teaches resilience.
In Summary
Every coach has their own heuristic, they are reflections of
our beliefs and personal philosophies. They are heuristics because every coach
has had some form of success based upon their heuristics. Heuristics only become
detrimental to our coaching when we ignore the basis context of how the
heuristics became heuristics. There is a time and a place for every heuristic-based
decision that we make. If we mindlessly follow our heuristics, we will usually
pay the price. We need to challenge every one of our heuristics consistently
and constantly. It is even better if we had other coaches challenge them for
us, either on or off the court. Professional poker player Annie Dukes, in her
book Thinking in Bets
References
Dukes, Annie. Thinking in Bets. New York:
Penguin, 2018.
Emre Soyer, Robin M. Hogarth. The Myth of
Experience: Why we Learn the Wrong Lessons, and Ways to Correct Them. New
York: Hatchett Book Group, 2020.
Epstein, David. Range: Why Generalists Triumph in
a Specialized World. New York: Riverhead Books, 2019.
Joe Maddon, Tom Verducci. The Book of Joe: Trying
Not to Suck at Baseball and Life. New York: Hatchette Book Group, 2022.
Kahneman, Daniel. Thinking Fast and Slow.
NYC: Farrar, Straus and Giroux, 2013.