Path Dependence in Human Performance

Is relying on the past the best way to move forward?

One of the most enlightening aspects of working with athletes that exhibit high levels of performance is the historical details of the training and practice of how they got to where they are in the first place.

When looked at from above, it would be possible to take any two athletes - any two high performers, and it would be easy to see that they may be similar in every way from any measurable performance outcome; however, when looked at in-depth and from a systems perspective these two high performers will be different entirely based on how they got there. How they achieved their current level of high performance is ultimately more critical to our understanding of them than is their high level of performance itself (and the statistics it implies), as this gives us as practitioners qualitative information about their current state and range of behaviours.

This is path dependence. From a high-performance perspective, how the sequencing of training over time based on the outcome of training over time plays a role in athlete behaviour at some point in the future.

What is Path Dependence?

Path dependence refers to a phenomenon where the outcome of a process or system depends not only on its current state but also on the sequence of events that led to that state. In other words, the outcome is "path-dependent" because the choices made at earlier points in the process influence the available options and constraints for future decisions.

Very simply, path dependence refers to the tendency to rely on the past to reconcile the future.

Path dependence can be seen in a variety of fields, from economics to politics to technology. For example, in economics, the path dependence of institutions can lead to situations where certain policies or practices become entrenched, even if they are no longer efficient or effective. In politics, path dependence can explain why certain policies or laws are difficult to change, even when they are widely seen as outdated or harmful. More specifically, it is the use of past practices, behaviours, or decisions to guide future outcomes, instead of relying on or using information from the current circumstances or states even when better possible alternatives are available.

In technology, path dependence can occur when a particular standard or format becomes widely adopted, making it difficult for alternative standards to gain traction. This can create "lock-in" effects, where users are unable or unwilling to switch to a better option because of the high costs associated with changing.

Path dependence is a vastly utilized strategy within corporations and institutions as a way to govern policy and strategic endeavours, often because it is a more cost-effective, less difficult, and less informational approach that saves time and effort on behalf of those involved.

It is easy to observe how the concept of path dependency has taken hold within the world of human performance at all levels, both in the specifics of training for sport as well as in the skill development process that accompanies it. Too often, it is often based on history, on ease of application, and on little relevant information that limits the range of behaviours that can be acquired. Often this occurs in the vain of:“this is the way it has always been done, so this is the way we will continue to do it” or some similar version of this thought process. This leads to some of the major limitations that are currently observed in training for performance.

On the surface, this may seem benign, as it is evident that athletes that perform at elite levels within their sport still occurs; however, it must be considered that this occurs somewhat independently of the training undertaken and whether these athletes are simply statistical outliers (an extreme random event) relative to their peers.

Here is a mini-lesson by NN Taleb on path dependence:

***Side note relative to NN Taleb: if you haven’t read any of the Incerto, I would recommend doing so. Anti-fragile is on our original book recommendation list. For those that have read it, I would highly recommend reading it again, especially if a prolonged period of time has passed since your last reading. The first time through this book, I was frustrated by his use of new verbiage, his historical references, and analogies, having studied systems thinking in its most fundamental way. Five years later, I re-read the book and was blown away by the depth of analysis and understanding of systems that Taleb provides***

The Consequence of Path Dependence

In the above video explanation, Taleb demonstrates that there are different paths that can be taken to get to an endpoint. In his example, there are three, one that oscillates around a fixed point (+1/-1) and two that have large deviations away from a fixed point and then eventually return, one on the positive side and the other on the negative. It is also worth noting that the beginning and ending are of the same value.

Although not articulated in the video, what is worth discussing with respect to path dependence is the eventual outcome of path dependence which can be stated simply as regression to the mean.

Regression to the mean is a statistical phenomenon that describes the tendency for extreme values of a variable to be followed by values that are closer to the mean or average of the variable in subsequent measurements. The pathways in the above video demonstrate this in that there can be very subtle changes concentrated around a mean, or there can be extreme events that will often be followed by normal everyday occurrences that level out the extreme event over time, eventually to the point where the beginning and the end are the same.

Imagine a group of individuals who are measured on some variable, such as height or test scores. If the group is split into two subgroups based on extreme scores (e.g., individuals with the highest scores and individuals with the lowest scores), it is likely that the individuals in each of the extreme subgroups will have scores that are closer to the group average in subsequent measurements, even if they do nothing to change their behaviour or performance.

The mean in high performance does not lead to maximum payoffs, Matthew Effects, or career longevity. Unfortunately, athletes that follow path-dependent training programs will regress to the mean, which in no way provides the opportunity for achieving higher levels of performance.

Real-Life Scenario of Regression to the Mean

Recently we were contacted by an athlete who is finishing his college career and is attempting to turn pro. His Point A from a strength perspective had not changed in his five-year collegiate career. His maximum strength numbers as an 18-year-old freshman were the same as his numbers as a 22-year-old senior! This is a common consequence of strength programs following training that simply checks boxes but does not provide for individual performance effects. It is evident in this one example (of many!) that sticking to a traditional or well-worn path is not always the ideal strategy because it restricts information flow that impacts decision-making and subsequently limits growth.

If you do what you have always done, you will get what you have always gotten.

The Paradox of Path Dependence

There exists a paradox of path dependence and its role in training for high performance that leads many practitioners to become highly resistant to change over time and create noted “lock-in” effects that reinforce positive feedback, further cementing the same path over and over, even if it is not returning the intended outcomes. Unfortunately, this is a common occurrence.

The irony of this path reinforcement is that in high performance, random and unexpected events occur on a regular basis (an athlete “pops” vs. an injury), which ultimately determines the path and the further resistance to change. Herein lies the paradox, as these random events are what ultimately create path dependence and the continued travel along that path; when ultimately using the appropriate feedback, a different path is more appropriate as the existence of random events is not dependent on any path.

This is part of understanding the art of training for high performance. Charlie Francis was a master of knowing when to deviate from a path and not get locked in. He would never allow an athlete to become destined to the results of a path, which arguably allowed his athletes to never regress to the mean and also allowed them to benefit from random events instead of faltering as a result.