Imagine a popular restaurant with a group of talented chefs. The owner wants to introduce a new dish — something that is both new and unique but also achievable based on the abilities of his chefs — but he doesn’t know how to go about it.
Let’s say a certain group comes along and offers their help, called ‘food analysts.’ These food analysts take all the recipes the restaurant currently owns and look at the history and current popularity of each dish. They then examine the ingredients and different methods used to put the dishes together.
They find that some dishes are more popular with different types of customers and some during different seasons. They also find that despite the popularity of one dish, some chefs are able to draw more customers because of their cooking style and their individual strengths.
The food analysts come up with a list of ingredients and a list of methods that could help create the popular dish, and give those suggestions to the restaurant owner. The restaurant owner creates various recipes based on the suggestions, and gives them to the chefs who then go on and create the dishes. Back and forth it goes; the chefs experiment with the recipes, and the analysts collect information on which dishes are the most successful.
This is how analytics works.
Now, imagine that the restaurant owner is a basketball coach, and the chefs are his players, and you might start to get an idea of the radical analytic experiment that the SFU men’s basketball team has embarked on this season.
Their final recipe? The dish they want to cook up? A winning team.
The SFU team’s ‘restaurant owner’ is head coach James Blake, and together with ‘food analyst’ Peter Chow-White, the SFU men’s basketball program has been trying to find the secret recipe that will lead them to victory.
Chow-White is the director of the Genetics and Network Analysis (GeNA) Labs and an associate professor of communication at SFU. He was researching the “moneyball” phenomenon applied to basketball when he approached Coach Blake last summer.
Blake agreed to form a partnership with Chow-White and gave him access to the data he needed and offered his assistance. It was a natural development for both parties, and came out of a need to further the understanding of the way basketball was being played.
Before the opportunity with Chow-White, the basketball team did not have an analytics advisor working within the team, and held what Chow-White described as “fairly typical, but limited” amounts of statistical data.
There are different ways of capturing the game, but the advance analytics Chow-White has been investigating go beyond traditional box scores. In fact, Chow-White explained that “basketball analytics is using traditional statistics and transforming them into efficiency statistics.”
Statistics by themselves are not analytics, they are just numbers — until they are placed into a context. To put it simply, advanced analytics provides an in-depth perspective on how the game is being played. The difference is a set of data based on totals versus a set of data based on efficiency levels. The purpose for basketball analytics is to create a deeper understanding of the game on a more concentrated and individual level.
The experience has been a building process for everyone involved. Since its official establishment last summer, the analytic experiment has come a long way, and continues to develop at a steady pace. In a relatively short period of time, the research team have been able to create an advanced analytical program from the ground up.
The partnership with Chow-White has greatly affected the way Coach Blake approached coaching; instead of teaching the traditional methods and focusing on effort alone, it became more about “performance based on statistics”.
The difference was training the team with a knowledge on how to be more efficient, and drawing knowledge based on information about the individual player’s strengths and weaknesses, as well as their abilities when working as a team.
Analytics is really about finding a method that works for a team but also one that hasn’t been done before or it becomes pointless according to Chow-White.
When he began, Coach Blake had two goals in mind for the SFU men’s basketball team. The first was to increase the number of possessions per game, aiming for an average of 100.
Chow-White notes that the average amount of possessions for a team is about 75 per game. Possessions refers to the number of times a team had possession of the ball.
The highest this SFU team has had is 80 possessions per game. With a difference of 33 per cent, the goal is to go beyond the type of game normally played. An increase in possessions means creating more opportunities for the team to score a basket, which in turn increases the point difference and the winning advantage.
The second goal was to implement a high offence, high defence strategy. This means playing a fast-paced game and incorporating a pressing offence while consistently maintaining a strong guard against the opposing team.
The method that Blake has implemented is known as a ‘disruptive structure,’ which aims to use more traps and steals to increase the number of possessions.
While rival teams are trained under a typical program, the goal of the SFU team is to disrupt that patterning. At the same time, the SFU players also need to be able to function well in the moment, because they cannot fall back into a typical patterning. With this in mind, the players are constantly playing against the boundaries of their own comfort zones.They need to be quick while remaining calm and steady in order to sustain the pace of the game.
For Coach Blake, the biggest hurdle is the time it takes to implement such a system, and the challenge of pushing the players to sustain effort and play at a high level consistently.
“It’s not just doing and being comfortable with what you do,” Chow-White explained. “You’re trying to systematically take [the team] out of their comfort zone, over and over and over again, every single game.”
A slower system might still be successful if only a portion of the players were fully committed. However, the faster the game, the harder it is to coordinate as a team. For this reason, the strategy involves having a team that is fully committed to the system and its success. Fortunately, the SFU players have found the new system fun to play.
When Coach Blake and Chow-White first met, they quickly got to discussing the goals of the program. Afterwards, they met regularly, but concentrated largely on their individual roles from their side of the partnership.
During these meetings, the two looked at the game reports that Chow-White provided throughout the week and narrowed down what was essential based on the present state of the team and the current set of data. They discussed how the analytics were being applied, and whether adjustments were needed on multiple levels.
Within a few months, they built a multi-page game report. The report consisted of advanced analytics that were repurposed from previous box score data, in addition to other statistical data that was gathered from the games until that point.
Sometimes they changed their approach in reflection to what the analytics revealed, and realigned their goals based on that.
As Chow-White explained, there is a common assumption that shooting three-pointers will be the deciding factor to winning a game; however, sometimes the winning game might include a lower percentage of three-point shots.
“Developing the technology and making it usable for people is a conversation,” said Chow-White. “There’s a larger process going on here than just simply giving stats to coaches.”
Above all else, Chow-White has focused his efforts on constant communication, which has taken time, patience, and understanding — not to mention a passion for the sport. “The key is communication,” Chow-White explained, “which is why a communication professor is interested in analytics in the first place.”
As the basketball team continues to better realize their potential as a team, the analytical data evolves to becomes richer and more applicable.
It is important to note that analytics is a neutral system that can be used to adapt to the different systems of individual teams. Every team has their own needs, and a coach with their own goals in mind.
Analysts examine the data for common trends; these links are examined for a pathway that is most relevant to the advancement of their goals, and the goals of the team. Coaches can then use this data to further their understanding of the strengths, weaknesses, and potential of a given team.
Coach Blake and Chow-White are two people who undoubtedly care a great deal about the game. But behind the scenes, there is also a research team of six individuals, including students working on the spreadsheet data, as well as computer programmers working with the “data-based technology.” All of them are equally dedicated to making this work, including former Clan basketball player, Scott Hind.
Hind began working with Chow-White at the beginning of the school year. Before that, he struggled with a concussion and fractured ankle, which led him to stop playing after two seasons with the team. Since September, he has been working as a research assistant. Along with a few others, he has regularly met with Chow-White to go over the data he’s collected while watching the games.
For Chow-White and the research team, the main goal is to build a database of information, and to use the data to help the coach and the team see and understand the game from a different perspective — as well as plan for the next game.
The data that is collected explains why things happened the way they did, and what future results can be expected from the individual, as well as the team. This is the sort of information that cannot be found easily by watching the game or reading the box scores alone.
In his words, Hind explained, “a key skill in being an analyst is the ability to communicate the information to Coach Blake in a way that makes it easy for him to digest.”
But conversations go both ways. Just as the analysts need to explain their perspective of the game to Coach Blake, it’s equally important to learn about the way Coach Blake perceives the information; in some situations, he finds a connection and sees something that the analysts wouldn’t know without the intimate knowledge he has with the team and the games they’ve played.
“We’ve [the research team] actually had to adapt our own goals and that’s a discussion [and] a conversation between how the coach sees things and [how] the data is seen.” Chow-White stated.
Hind noted the advantage of having a receptive coach like Blake who was willing to be involved and listen to analytical advice. Throughout the process, Coach Blake’s receptivity and openness has contributed to the positive progression of this analytical experiment.
“There’s a gap in terms of what the numbers are telling me and what the outcome is being,” Chow-White explained. “They are not a top-tiered team but they definitely are not a bottom-tiered team.”
In his words, the team while not a top-tiered team, is definitely not a bottom-tiered team and has what it takes to win and be in the playoffs. The numbers have shown that the team is doing well in some areas but they are not playing at the level yet.
In order to improve as a team, it is important to be as efficient as possible, and this takes practice. Evidently, the team is proving this fact by gradually closing the gap, game after game.
Hind shared how it has been an enlightening experience to look at the game through an analytical lens. After many years of playing basketball, he realised the limitation of knowledge gathered from being on the inside differed from the analytical data gathered from the outside.
In some cases, the numbers told a different story to how the players might have felt. For example, what looked to be a relatively good shot may well have been an inefficient one. This knowledge has given Hind and the rest of the team a better understanding for how the analytics translates to actual game play. In effect, perhaps this is the reason behind the growth of the analytics experiment.
In the end, the progress of the analytic experiment is determined in part by the efficient communication and proper understanding that all those involved have provided. Certainly, the process has been a fulling and well-endeavoured experience for all those involved, and would not have reached the same successes without the coordination and dedication of all the individuals.
For Coach Blake and the SFU team, it’s all about improving on their deficiencies while simultaneously trying to secure more wins. Blake complements Chow-White’s mind for basketball, and stated that “it’s been lots of fun to coach and I just look forward to going to every practice.”
At the moment, the final recipe — the winning team — has not been discovered, but the process of experimenting and looking at the data have created many equally wonderful dishes, each one providing a glimpse of what the final goal would look like.
With an enormous amount of data still left to analyze and the summer to come, Chow-White and Coach Blake hope the advanced analytics program will continue to develop and evolve. The ultimate goal? To create a program that can be transferred to other SFU sports teams, and provide an opportunity for those teams to analyse and improve their gameplay as well.
Bon appetit, SFU athletics.