By Zac Pross and Marcus Colby
Organizations are now gathering more data on their personnel than ever before. This creates the potential to yield greater insights, but large data sets can also bring complexity and be hard to manage. Archetypal analysis is a method for reducing data sets down to their core descriptive features, allowing you to tell better stories about the qualities, skills, and abilities of the personas in your organizations. And as we will explore, while archetypal analysis is useful in assessing individuals, it also gives leaders new and interesting ways to put teams together.
Archetypal analysis is similar to clustering, which is a popular method in machine learning and data mining. But unlike clustering, which is characterized by the “average” data point of a given cluster, archetypal analysis looks to the extreme points in the data set to create certain personas based on the extreme performers in a specific field. Examples would include LeBron James in the NBA, Alex Morgan in soccer, and Tom Brady in football. Once you’ve identified the archetypal/extremal athletes in a data set, you can begin relating and comparing all other athletes to them – where comparisons are made between zero and 100 percent.
So we could say that a certain basketball forward is 30 percent like James, a soccer striker is 40 percent like Morgan, or a football quarterback is 20 percent like Brady. And given that it is natural for us to compare athletes against the best of the best (or the worst of the worst), archetypal analysis presents itself as a useful method for descriptive statistics.
Archetypal thinking enables teams to compare their entire existing roster against single archetypes or to create composite profiles based on the combination of several personas. Archetypes can also be used when scouting opposing teams to better prepare for playing against them, or for talent identification when evaluating potential trade targets and draft picks.
Whatever the desired application, archetypal analysis reduces the size of the data being collected and allows you to observe individuals in new and unique ways. Instead of struggling through mounds of information when trying to describe, evaluate, and compare team members, archetypal analysis can distill your thinking by forcing you to interpret performance through the lens of only the best and worst performances. In this article, we’ll explain how this technique could be utilized within the Smartabase athlete management system (AMS) to improve team training, selection, and cohesion in sports, the military, and business.
Maximizing Team Sports Training
Sometimes in a team sport training setting, coaches want to put the whole roster through their paces in conditioning drills, wish to pit starters against reserves, or need to bring together entire position groups to work on set plays, tactics, and strategy. These are all necessary and effective techniques, but when it comes to choosing small sub-units to perform drills that emphasize specific physical capacities, groups may be selected arbitrarily or without a data-driven justification, potentially leading to mismatches in ability that reduce the overall quality of the outcomes. When players are grouped by their position categories – such as receivers in football, guards in basketball, or backs in rugby – there can be a wide range of skill sets and abilities even though the athletes play similar roles during competition.
Grouping players based off their actual data sometimes could provide a fresh, outside-the-box way of thinking that transcends traditional methods and shakes up the way players practice. Archetypes could be created and utilized to match up athletes of similar abilities, helping individuals and everyone involved in a drill get more out of it. Doing so could also introduce a sense of novelty that keeps players engaged and, on their toes, because the session isn’t going the way they’d typically expect it to. You might still end up with groups containing multiple players from the same position group because they have similar physical capacities, but the way they were put together is still more novel and compelling than the way they’re normally separated by the coaching staff.
This can be particularly useful when considering how to bring similar players together in complex, multifaceted drills. Some practice sessions involve simple expressions of speed or jumping ability, while others might involve the use of multiple skills. A player may be good at one element but poor at two others, while their teammate is solid in all three areas. If these athletes are put together, the drill may yield a diminished learning opportunity because they’re mismatched. In this instance, archetypal analysis would allow the team’s sports scientists to look at multiple variables in the data and then condense these down into three or four variables that make it easier to compare the players. The coaching staff could then put them together in groups of similar abilities so that there’s more competition and parity when doing certain drills. The same approach could be applied to selecting mini teams for small-sided games in practice and inter-squad scrimmages.
Enhancing Special Forces Team Cohesion
During Special Forces selection processes – such as for Army Rangers, Navy SEALs, and Air Force pararescue specialists – instructors seek out candidates who are versatile, well-rounded, and can be effective under pressure in just about any environment. But when it comes to specific missions, there can be highly specialized requirements and some operators are better suited to the required roles than others. Prior experience and commanders’ preferences certainly come into play but having access to archetypes could also be beneficial in selecting the best mix of personnel for a specific mission.
The high-consequence nature of Special Forces operations means that team cohesion is non-negotiable. Being able to utilize an approach like archetypal analysis enables commanding officers to maximize the chances of team-cohesion by composing teams in a data-driven way. Each operator is a multifaceted person with many different skills and attributes; archetypal analysis enables team leaders to reduce complicated profiles down to just a few key variables. This makes it straightforward to assemble groups comprised of individuals who have similar, or, in certain cases, dissimilar archetypes to suit the aims of a particular mission. Doing so could simplify and expedite commanders’ decision-making in situations when time is critical, and outcomes are of high consequence.
Improving Client Project Management
In any business organization, a key component of project management is bringing together the right group to team up on clients’ projects. Sometimes this is simply based on which staff are available, who has the capacity to take on more work and making sure that there’s at least one of each role to cover all the bases. Archetypal analysis can provide a more thoughtful approach to selecting a cohesive team that will best meet a client’s needs. On the other end of the equation, using customer data to create archetypes allows organizations to assign personas to their clients so that they can be grouped into distinct buckets more easily. Once the characteristics of each type of customer has been defined, it’s then easier to match these with a team selection process based on choosing well-qualified staff who will be complementary to one another on the project.
Archetypal analysis doesn’t involve creating new information per se – it’s always been there in the data your organization has already captured. Creating archetypes in Smartabase just involves distilling data down into several descriptors that are simpler, more manageable, and allows you to think about team selection and cohesion in different and useful ways. From the perspective of a sports coach, military commander, or business manager, utilizing archetypes shouldn’t be expected to always provide definitive answers. Rather, it’s a trigger for coming at a problem with a new perspective, asking different questions, and finding a new lens through which to look at group dynamics.
If You Enjoyed This Article, You Might Also Like…