Outside the Box Thinking: The Theory of Constraints & The Devils

On Twitter, Justin Bourne was asking for a non-hockey book recommendation and I suggested The Goal by Dr. Elihayu Goldratt. It describes and introduces The Theory of Constraints, a method of process improvement for a goal-related process, through a light novel. While the method and the novel focused on a manufacturing site, Goldratt does make it clear that it could be applied beyond the factory. The principles can be applied personally or in other businesses. And, as appropriate for this blog, potentially a hockey team like the New Jersey Devils. A reader, Martin Devon, asked whether I wrote about it before with respect to hockey and even had some suggestions as to how it could apply. I don’t think I did and if I did, it would have been years ago. So let’s explore it today. This is some food for thought as we are about to begin pre-preseason, preseason, and soon regular season hockey.

The Theory of What Now?

While I am not a trained or certified person in this mnethod, I can give you a high-level overview of what the Theory of Constraints is and is not. It is a method to drive process improvement. It is not a method that will fix everything. It is a method that will highlight a major issue of a system and allow you to focus on improving that issue to make the system better as a whole. It is not a method that will necessarily solve or eliminate the issue. It is a method of continuous improvement; it is designed to . It is not that effective if you do it once or twice and then never do it again.

The whole thing is built around the saying: “You are only as strong as your weakest link.” That is the constraint. A process is going to be limited by whatever is holding it back the most, so to speak. A constraint is not just personnel or a specific process. It can just as easily be a company policy, a team’s way of thinking, or the equipment involved. If it is something that limits what the organization is trying to accomplish even if nothing is going wrong, then it can be considered to be a constraint. Some examples come to mind:

• Let’s say you have a widget company. It takes 15 minutes to process a widget order, 10 minutes to pick the materials for the widget order, an hour to make the widget, 20 minutes to inspect the widget, and 5 minutes to pack up the widget for shipping. My goal is to have an efficient process. Yet, it is currently limited by the time it takes to make it because it is the longest. The Theory of Constraints would identify that and work to improve that part of the process to improve the business as a whole.
• Let’s say you are a coach on the staff of the New Jersey Devils and you’re putting a forward line together. You want the line to dump the puck in on zone entries and have Miles Wood chase it to gain possession. Your want one or both of his linemates to be in the offensive zone when he does try to win the puck in case he does; or in a place to forecheck if he does not. If the other two linemates cannot keep up with Wood or get into a positions to support Wood, then they are the constraints for that strategy and the line. The goal is to have a line that plays this way for Wood and successfully win zone entries. By the Theory of Constraints, the coach would recognize this and consider what could be done to mitigate that to make that line more functional on the ice.

Applying the Five Focusing Steps

There is more to it than just identifying the constraint and doing something about it. As the Theory of Constraints is a continuous improvement method, its application consists of five steps - The Five Focusing Steps, as its called - that cycle around each other. The site, Lean Manufacturing, has an excellent overview from a manufacturing perspective of what it is and how it can be applied. This is a Devils blog, so let’s go at the five steps with one of Devon’s examples: a breakout.

The goal of the breakout in hockey is to move the puck from the defensive zone to the neutral zone with possession. This is often done in the face of some forechecking pressure by the opposition.

Step 1: Identify the constraint - One of the constraints of the Devils’ breakout was their personnel on defense. Not everyone on defense had the same skill or confidence at moving the puck up ice. For example from last season, the pairing of Will Butcher and Ben Lovejoy worked well in general. However, the recently-retired Lovejoy struggled to pass the puck effectively. He would rather move the puck to Butcher or just clear it away. This meant that when they were on the ice, the team was reliant on Butcher making it work. While he was good, opposing teams could target this and cut off the breakout.

Step 2: Exploit the constraint - This really means to address the constraint with what you have. As much as you would like to, you cannot bring in a new player from another team in the middle of a game. So what can you do? You can switch Lovejoy off for another defenseman if the play has enough time and space for a change. You can have a forward draw deeper and act in the role Lovejoy would have. You can run a different play, which does happen when plays breakdown in a fast and fluid sport like hockey. You can even have Butcher switch to Lovejoy’s side and hope he can run the play there. The idea is that if the constraint is whether Lovejoy has the puck, then the staff has to minimize that in the interim.

Step 3: Subordinate the constraint - This is where you look at the non-constraints and determine how they function with the constraint - if at all. This is where the breakout play itself is considered. Does it need the right-sided defenseman to make that first pass? Does it allow for different defensemen to switch off, or is it ad hoc depending on who retrieves the puck? Does the play change for different lines in front of Butcher-Lovejoy? Remember that whatever is identified as the constraint is impacting the process. It is important to consider how the system interacts with the constraint and ensure that the system would support what the constraint can and cannot do. In other words, does the breakout play require that both defensemen be able to be at least decent at moving the puck? Or is it structured to “hide” someone who is having issues? This is what this step does.

Step 4: Elevate the constraint - This is the actual improvement part of the constraint. There are a lot of different approaches here. For some players, it could be something to focus on improvement through practice, drills, and video sessions. For others, it could be something that may drive the coaches to adjust or change the breakout play entirely. Or consider a different one for specific pairings and/or lines. The elevation could be more permanent, like what the what the Devils eventually did. They traded Lovejoy to Dallas for Connor Carrick and a pick last season. Granted, this was not really why they did it. The Devils were awful, he was on an expiring contract, and so they sought out assets. Still, it was a solution and that is consistent with this step. Sometimes the best improvement for a constraint is to eliminate it.

As a quick aside, Lovejoy retired in this past week. Congratulations to him for his successful career - it is very difficult to stay in the NHL for 11 seasons - and on going out on your own terms.

Step 5: Repeat the process - This is why the Theory of Constraints method is continuous. Sure, Lovejoy will not be on the 2019-20 team. There cannot be a recurrence in that sense. This does not mean the team’s breakouts are going to be great from here on out. It is important for the players, coaches, and the off-ice staff to continue to monitor breakouts for other issues. There could very well be more constraints, such as where forwards are for these breakouts or how they respond to different forms of pressure. Or even new ones should players decline or players not work out as expected on these plays. It is the responsibility of the organization to confirm whether their improvements to the constraint actually led to improvements. If so, then look for additional constraints to make the process better. If not, then start again with a fresh set of eyes and re-evaluate.

As a second quick aside: The good news for the Devils (and everyone else) is that with player and puck tracking, there could be a lot more data available to truly track how well the Devils are doing on their breakout plays (among other areas). This will make it much easier to find constraints as well as find areas of strength and functional tendencies.

Again, I’m not an expert or a certified person in terms of the Theory of Constraints method. But following the Five Focusing Steps shows that there can be more to it than just finding an issue and cutting it out or changing it. There’s work to be done in the short-term, there’s confirmation of how the process works with the constraint, there’s , and the whole method is broad enough that it can apply in many areas.

It Can Even Work Off the Ice, Too

While Martin suggested breakouts, the first thing that came to my mind with respect to the Theory of Constraints and hockey is scouting. Scouting is a very challenging process. Consider this summary of what is involved:

There are hundreds of kids playing hockey all over the world in many different levels. The goal is to identify which kids playing today could grow up and develop into professional hockey players years from now. Never mind that the differences between being on a stacked team and being on a basement dweller; there are differences in how the leagues play in. As most leagues are not on the level of the NHL in terms of tracking statistics, relying on that to filter out players may not . Further, on-ice stats are not going to reveal how well the player skates and handles the flow of the game, how they react in physical situations, and how they approach certain situations, struggles, and successes. I am a big believer of quantitative analysis but scouting involves quite a bit of qualitative analysis. And this requires both video and in-person observation. Of course, when the scout does this, it is not a guarantee that the player they see is really the player they truly are. They could be on a hot streak, a cold streak, having a bad day, having a really good day, they are under the weather, they are under a lot of stress, or something else. And let us not forget the common statement of “I came to see X and I’m wow’ed by Y.” One could account for this with repeated viewings but resources such as time and personnel are limited. And this is all involved to figure out who is likely to be a NHL player years from now. Which is the goal.

Teams such as the New Jersey Devils have an entire department devoted to scouting. According to the team’s staff page as of August 30, there’s a Director of Amateur Scouting (Paul Castron), an Assistant Director of Amateur Scouting (Gates Orlando), ten Amateur Scouts, six European Scouts, and three Pro Scouts. Scouts are assigned to work with different regions, leagues, events, and, presumably, as-needed to see players. Assuming those in director positions are not making regular trips themselves, that’s sixteen people looking at potential future players. Eighteen if they are. How they are organized, how they process their observations, and how they use their time to scout players is critical to provide management the best decisions for the draft (or even future trades and signings). Efficiency is a big advantage.

One way to for the Devils to ensure that their scouting system is efficient would be to apply the Theory of Constraints method. The most notable constraint that I can see is resources, as in time and availability to view players. From there, they can make decisions on how to mitigate it in the short-term as the scouting season has already begun (seriously, the Hlinka-Gretzky Cup happened, World Juniors showcase games happened, 4 Nations happened, preseason games in Europe are happening, the 2020 class has begun). They can confirm whether their procedures and approaches to scouting take these constraints into account and how they do. From there, they can identify improvements. This could mean a streamlined approach. This could mean establishing a sample for players considered notable for multiple looks. This method could even be used to justify to upper management to hire more scouts or designate scouts differently. Lastly, though repeating this method and monitoring their system, they can identify other issues to lead to more improvements. Such as maximizing more of their viewing time or casting a wider net to view more players or even focusing on fewer areas/leagues but with the intention of gaining more information about the players within that area/league. There are many possibilities. Either way, this would be a way for the Devils to take a proactive approach with their scouting - which can tremendously help the team in the long run. This is not to say that it is bad now, but it can be better.

For all I know, they are taking proactive approaches already. Even so, this method can be utilized to justify to upper management what changes they need to approve. While there is no salary cap for staff, obtaining more bodies for a department would typically require someone to sign off on it. Utilizing a systemic method to make the case for more people is likely going to be a stronger argument than trying to explain why. At the least, it would show not only why more resources are needed but also how it would impact the system.

And this could apply to other departments. It could apply to ticket sales. It could apply to how equipment is ordered, maintained, and stored. With a little imagination, the Theory of Constraints could be used to analyze and improve a lot of processes that we may take for granted.

Your Take

The concept of continuous improvement is consistently seen in a lot of pro athletes. Decades ago, the offseason would be just that for players - a time to be off. They would try to get in shape in training camp or even as the season starts. Those days are in the past as it is the job of the pro athlete of today to not only stay in shape all season long, but also work on improving themselves physically, mentally, and in their chosen sport all season long. This is why many players work with trainers in the offseason - and they’ll work out on their own too. This is why some players will get together to skate and even practice. This is why even a solid veteran with a locked-down roster spot like P.K. Subban is going hard in the gym so he can be in the best shape possible for the upcoming season. While they may not be thinking systemically about improvement, they are putting in the effort to improve themselves however they can so they can stay in the league and stay good in the league.

Team organizations should absolutely follow that lead. And they can apply a method like the Theory of Constraints to best identify what to work on, how it can be improved, and continue. The mindset is not that “Oh, no, we will never succeed.” It should be “How can we be better tomorrow?” Even as franchises, most organizations are very competitive and they are looking for advantages to put them ahead of others. Applying methods like this can do that and it can be applied in hockey operations as well as in business. It also establishes a culture of not settling for being “good enough” or being stagnant, a mindset that dooms teams to mediocrity or the basement. And even an organization with the riches like Toronto would benefit; being able to spend a lot for resources may be good, it is better to spend that money wisely to get the most out of the resources.

And it can also apply to the rink as well. As it is, coaches and players are limited by quite a bit during the season. While there are morning skates and video sessions, there is not always time to have full practices in the schedule. Travel takes time away. Having back-to-backs and/or weeks with four or five games in it sets your schedule anyway. And as easy as it is for you and me to say So-and-So is no good at this, the team needs to get a better player - that is not always a viable option during the season. Sure, getting better players and players more suited for a role may help, but applying a method like Theory of Constraints can better identify other issues at play, determine what can be done in the short term, and really confirm whether the best approach to address the constraint is for the GM to acquire a certain player, sign a certain player in the offseason, or even call someone up. And it can be done as part of the analysis that is done from game to game with others assisting and communicating downstream (or upstream) with recommendations.

There are absolutely other methods and approaches to improving systems. Theory of Constraints is one I am (somewhat) familiar with and one of its benefits is its flexibility. It can be applied in a lot of different ways. I just happen to think that it could help the Devils. For all I know, they are doing something like that now and if so, great. Keep it up. But it is this type of “outside the box” thinking that can help make a difference in the long run as the Devils work to attain the heights they once held in the NHL. I want to thank Martin Devon for the suggestion to write about this and I want to thank you for reading. If this was too much dwelling on a systemic approach to improvement or too outside the box, then do not fret - I will write about a more sobering NHL-related topic tomorrow.