 Dean Little. Dean is the sports science coordinator for the New York Mets. Hopefully baseball will be back soon. Also spent two years, more than three, or more than two years with the Memphis Grizzlies as the lead performance analyst in sports science and reconditioning expert. Master's degree in the field of study in high-performance sport from Australian Catholic University. Welcome Dean. Thank you for being with us. Good afternoon everyone. I'd just like to start off by thanking the Steadman Clinic in the US OPC medical team for putting together such a great conference and a really good line-off of speakers. I hope everyone's doing really well, and they're safe at home listening to this and my thoughts go out to your families who are out there working on the front line as medical staff or essential workers. Today, I'm going to be talking about the return to performance and how player tracking technology may help us close the gap. I'm hoping it fits in nicely after Trent and Carlos's fantastic presentations. I'm currently working as the New York Mets sports science coordinator and work with a superb medical and performance stuff. I just wanted to note that the views expressed in this presentation are my own and don't represent the New York Mets organization. For the purpose of this brief presentation, I decided to focus on one injury type, ACL, and create some new tools with publicly available NBA data. We all focus on simulating the demands of the game in our rehabilitation plans, but I'm hopeful that advancements in tools like the ones that I've built here help us take a step further in the granularity of the game demands and improve the specificity of our training design. I wanted to demonstrate how we can take this jumbler numbers here and translate it into relevant insights for players and coaches so they can optimize the return to performance and create successful outcomes for our athletes. In end stage rehab, no two athletes are the same. As practices move towards a multi-disciplinary and personalized criterion based decision making approach, teams have really had a tough time still balancing the risks of early return to play with the rewards of team success and athlete success with early reintegration. Despite the increased level of care and new technology available, there is still a very high incidence of re-injury in sport. And one of the more frequently cited explanations for this is the inadequate preparation for the demands of the individual sport as re-injury often occurs within the early return to play window or late rehabilitation stages. As a result, it's important for us to remember that no two athletes are the same, their strengths and weaknesses differ, their roles and responsibilities differ and it is important to tailor their programs for their unique skill demands. In an effort to understand if the athlete has returned to performance, we really need to understand what their pre-injury peak performance level was so we can reverse engineer it. With the recent advances in player tracking technology and understanding, our focus has now turned to increasing the detail of information we have so breaking down the components of an individual's game and their skill sets. There's obviously been an abundance of research to support approaches to understand the individual's physical demands of a game from a load perspective and from an intensity perspective. And as the price of 3D motion capture systems has come down, teams are beginning now to explore the individual's mechanics in a similar way that research has within the laboratory setting. And then finally it's important to understand the technical and tactical demands of an individual within the game. Ultimately in this case we want to collect data to understand the quality of the execution of a player or technique under the environmental constraints. With all of this information we can begin to understand the individual's technical and physical demands within the game which allows us to create position or play a specific drills and tests. This highly individualized approach really allows for a very specific replication of the athlete's role within the team and their most challenging periods of play. And this is crucial for conditioning our athletes to return to peak performance. With respect to player tracking technology there's obviously many options available on the market. These include optical such as second spectrum or stat cast which are generally used league wide in the case of the NBA and the Major League Baseball. And then teams often opt for wearable GPS or local positioning tracking systems and these can be used for either or both of competition and practice to track players. Typically these tools provide us insight on the volume of activity or load that an athlete completes such as distance per game by speed zone or cumulative load estimates. As well as information on the athlete's absolute intensity or duration specific intensities such as peak distance per minute and also work to rest ratios. These measures have been revolutionary not only in rehabilitation and return to play decision making but also understanding the physical demands of our sports to improve our overall preparations and obviously then research into associations with injury risk. That being said in my experience these measures are generally difficult to relate to players and coaching staff given their lack of sport specific context. These guys spend the majority of their careers working on technical and tactical aspects of the sport. Fortunately some professional sports have event tagging systems which we can combine with that player tracking data to give us a layer of context. Typically these tags provide us timestamp technical actions such as passing and rebounds for basketball or base running and pick off attempts for baseball. These actions can be used in sequence to provide tactical insights of location of actions and overall team strategy. But what do we do if the data that we need is not available within the tracking system that's been provided? As often coaches will ask questions so we can't answer immediately. A strategy that can be used to solve for this is to create your own custom event models of the Royal X-Wide data. Another option is to build out notational tools to log things that you might not be able to code from that X-Wide data. And this would be for things like play calls, drills, technical actions in practice and other data that's not yet currently available such as jumping which you can see here in the example. So if we bring it back to ACL return to play where do we start? How do we come up with our ideas on what we might want to track? In this example I went through and I first started looking at the risk factors within the wider literature and then those that were demonstrating recent injuries. So I went and reviewed all of the NBA in-game ACL test since 2012. And even though it was a small sample the mechanisms were in keeping with the wider literature. The majority of these occurred at speed as cuts when the ball side leg was the cutting leg and this was often in transition. The other common mechanism was landing on a single leg after a finish at the rim. With this information in mind we can begin to compile some measures to consider for ACL return to performance. These factors may include the area of play on the court. In this ACL sample 87% of the injuries occurred within the paint either on a drive and cut or a finishing maneuver. The animation on the right highlights a passage of play with an automated detection of a drive into the paint. You can see the player of interest with the orange ball comes off a screen and attacks downhill. He turns from a black to a red dot once he enters the zone and makes a shot. This event is now tagged within our data set under drive paint entries and we can use this to analyze drives per game or per rotation or per play call. We also want to individualize the speed and deceleration risk factors. We can create personalized speed and deceleration bands using a method of one-dimensional clustering into five zones. We would also want to track the transition speed and counts as speed plays a big part in injury risk. And athletes often need a full court to generate top speeds in basketball. Obviously the other risk factor we would love to assess is change of direction. For this we can utilize angular velocity data to develop a new measure. We can then grade these changes of direction by angle of cut and intensity. As you can see in the right, no two athletes are the same. Here we have a big man whose change of direction on offense is limited to the paint and of low intensity but very high frequency. On the defensive end is where he gets his higher intensity change of direction. On the other hand his counterpart who's a guard makes a lot of high intensity change of direction actions on the offensive end around the perimeter. One thing to consider with this is we do have some limitations. We don't know the strict body orientation of the athlete and it also misses deceptive cuts such as a crossover where the athlete actually continues in a north-south direction. We obviously also have to study the validity and the reliability of these measures. Once we have our measures we can begin to build out informative plots such as vectorized clocks for direction at a time point for intensity. We can also build heat maps to highlight the location of high frequency actions or risk factors as demonstrated by high velocity sprinting here and hopefully this improves the knowledge transfer with the athlete and the coach. In my experience I've found it a lot easier to ask a coach to design a half court drill without cutting or paint touches than to ask them to reduce the amount of deceleration and change of direction for an athlete. Once you have your sweeter metrics that you think are important you can begin to review and catalogue all of your old game and training data to find the frequency and intensity of which these events occur. For instance transition efforts per game or quarter, the intensity of decelerations on a closeout drill, paint touches and cuts per play call and most importantly got to relate it back to outcomes such as shooting percentage. Overall this process should allow you to develop informative tools to educate your staff, the coaches and the athlete themselves. So how do we use this information to guide us in our return to play and then onwards to return to performance? Tabanare and colleagues have proposed the control to chaos continuum and have applied it numerous case studies within the literature. This is a highly practical resource that helps you with phasing a constraint based approach to return to play. Restoring workloads, intensity and specific skills and then eventually tactical challenges in keeping with the athlete's injury healing timelines. If we start with high control of risk factors for a session or drill we can build a physical and psychological foundation before graded exposure towards the most challenging periods of play and beyond. In understanding that the return to performance might come after the return to play we need to use these data points to drive continual growth within our training plans. So where do we start? Well once we have the traditional workload and intensity rehab plan set we must promote the early integration of athletes with their coaching staff and teammates but we've got to do it in a safe yet progressively challenging environment. So we can take all the data collected and the drill heat maps and stuff and we can work collectively with the treating practitioner, the coaching staff and athlete themselves to build out an appropriate continuum. With the endpoint in mind we know where we want to head and where we've got to progress but it's about creating a nice stable progression from the athlete's current status. It's got to be hard enough to challenge them but not too hard that it's dangerous or they repeatedly fail. I'd always start by considering the risk factors that we must control for early. So the planes of movement, the self-paced linear drill controlling for cutting and deceleration movements is going to be safer for someone with an ACL injury. Whereas maybe lateral and less acceleration may be safer for someone with an Achilles injury. When involving teammates we want to consider the zones of the court where the athlete may be at risk and the space available as that will take physical demands. The skill acquisition literature can help us guide us in developing appropriate constraints both to achieve the appropriate challenge point to keep the individual motivated and also to grade the progression. Some factors to consider include the pacing of the drill, the work to rest ratio and accumulative fatigue. Temporal pressure such as shot clock or player density as less space requires quicker actions and increased perceptions of pressure. And the other thing is the number of decisions and options available to the athlete as this challenges their mental processing capacity. Obviously some of these factors are potentially dangerous especially early in rehabilitation. But in later progressions even after the return to play we need to be cognizant of these factors to continue to promote growth. Once a decision has been made that the athlete is clear to return to competition, tracking data can be utilized to optimize the initial return to play period to offset the height and risk of injury. From a load management viewpoint we could consider minute recommendations based on player specific predictions that consider the opponent's pace and personal matchups on offense and defense. With respect to intensity we could explore the periods of the game that are most appropriate for the athlete. Typically the fastest play occurs in the first quarter and the fourth quarter is a little slower with more timeouts and shouts especially in close games. Another option is to explore rotation lengths to minimize fatigue and maximize performance. With respect to skill we need to monitor the athlete's current performance given the opportunities they've been given. For instance not taking open drives and potentially find drills and play calls that need to be improved and ones that we can use to instill confidence. Finally through the whole return to play and return to performance process it's important to monitor the athlete's response to the environment and workloads that we're creating for them. There are obviously a lot of assessments out there that can be conducted. These include things like perceptions of workload, soreness and confidence, force plate analysis and musculoskeletal screens as well as measuring the heart rate indices. All of this information should allow us to adjust and modify our training plans to optimize for the athlete's individual needs at the current time. To finish up I just wanted to touch on the importance of the mental impact on return to play. It has been suggested that an athlete's psychological status may actually have a greater effect on the return to play outcomes than physical performance factors. Athletes often have common negative thoughts through this process and these include a fear of failure to reach pre injury performance and obviously a loss of sport specific routine and team contact. So all of this highlights the importance of early reintegration and gradual progression of safe sport specific skills within the team environment. Well thanks for listening to my presentation. I hope it provides some insights for those working in the field as to one possible approach to bridge the gap and highlight some of the data and tools that are available for those who aren't working within the field. Among the many people who take the time to answer my calls in talk sport and analytics, I'd like to extend a special thanks to Dave Taylor, Jared Faust, Joe Boylan and Noah LaRoche for their continued support and for answering all of my persistent questions. For any of you who would like to get in contact with me feel free to reach me via my email or Twitter listed here. I hope you all enjoy the rest of the conference and continue to stay safe.