"DBT #22" is a Computerized Basketball Training System. Guided by Artificial Intelligence we are designing a system that will take Basketball training to the next level. In this system of training, we start by analysing the athlete that wants undergo basketball training and then build them to reach a professional level of training, grading them on their skill level as they progress; determining their true position and at the same time helping them advance to an All-Star level.
TRI-MODE SYSTEM
To help us achieve the goals that we have set for “DBT #22”, we have created Tri-Mode system. This three mode system includes Learning Mode, Training Mode, and Coaching Mode. With this system we can better understand how the athlete plays and then begin training them and determining at every stage which position they truly play so that in coaching mode we can be able to place the right players together for each offensive or defensive play. We'll start by describing each mode separately to get a better understanding of the role each mode is playing in the training.
Our entire system is build up from this fundamental stage. Learning mode is the stage at which the system is studying how the athlete plays. To create this mode we developing a Low Energy Motion Capturing System that we are designing to break the barrier between the “Virtual World” and the “Real world”. The system begins analyzing the athlete by answering questions such as;
What key factors contribute to Athletes accuracy?
By studying the factors contributing to the athlete's accuracy such as height, strength, energy level, stamina, weather, body temperature, we will be able to know the athlete better and they can also know themselves better in what conditions they play best, what exercises they can do to improve on strength, stamina, and endurance. This important because we have to keep in mind that every athlete is different, to give better advice you have to know your players.What moves is the player able to execute?
Another important part of getting to know our athletes skill level is by analyzing the skills that they are able to perform. We imagine that this system can be built to analyze the data received through the motion capture to compare this data with the stored data that we'll train this system to recognize and also train our players with. This is a challenging part of our system currently because it involves precise and accurate calculations but we hope we can program the system with data within a certain range by repetitive trials from different subjects storing this valuable data into the system. Since each player also plays differently, we also want to analyze successful manoeuvres that are made by each individual player and give them the option of storing their moves by repetitive trials to the Player Profile Management System (PPMS).
How does the player shoot?
By studying how the athlete shoots the system first learns which parts of the court the athlete is able to shoot from, the athletes range in shooting, their body form when they shoot to properly help them correct any mistakes they may be doing, and their accuracy levels.After the “athlete analysis” phase is complete, the system now knows this subject as a “basketball player”. The digital trainer is now ready to condition this player. The trainer monitors the progress of the player while training them perform all the moves in the system step by step and constantly updates the Players profile in the PPMS. In training mode the player undergoes offensive tanning as well as defense strategies. On offense we also train a player on moves that will help them drive past their opponents. This “digital-trainer” will adapt the principles human basketball trainers have used to train the basketball players with such as; Ball-Handling & Dribbling, Defense & agility, Footwork, Offense, Offensive Moves, Passing, Post Play, Press Breaker, Rebounding, Shooting, and Transition & Fastbreak.
With AI aided training, we believe the system will help the players develop these skills much better, because of analysis methods the system will use to develop Players. We imagine this system can be trained by professional athletes and through software updates, it will advance on skills that players can be trained to develop. An average NBA basketball Players career lasts for 5 years, with this system we can store his valuable skills and methods of training and conditioning before he retires, and train the new players with these skills to develop better, stronger, and faster basketball players.
COACHING MODE
In Player coaching mode, the system begins coaching players on gameplay. Once the training process is initialized and the system has updated the PPMS (Player Profile Management System), the Players profile containing records of the each Players exercised skills can be loaded in coaching mode to give the player gameplay strategies and teach them how to apply these skills during games. “Coaching Mode” is a multi-player setting, with this mode the system begins by identifying every player on the field and each players profile is loaded; the coach can then give “Over-Air” pointers to each player based on the position the player can play accurately. The coach will help the player learn to know his surrounding and knowledge him to identify “Open” team-mates who are also able to execute their skills from particularly identified positions. In this mode players are also trained on “team-playing”, they learn passing techniques, court visualization, timing, drive past opponents, crossover moves, and develop the confidence to shoot from spots on the court under pressure in order to master the skills they exercised during training. Most basketball players are able to shoot on an open court, but learning to shoot while under pressure is the focus of coaching mode. Another important part coaching mode helps us to make possible is the problem of identifying the key ingredients for an offensive/ defensive play in the coach’s playbook, In this system the right combinations are automatically put together for an offensive and defensive play to create stronger match-ups during games.
PLAYER PROFILE MANAGEMENT SYSTEM
PPMS is a method of keeping updated records of data relating to every player under training with “DBT #22”. As observed, the PPMS is important in all three modes of the system because it identifies each player. The PPMS is responsible for storing the players necessary information (i.e. Name, height, position), his shot statistics from every point/angle on the field (i.e. goal percentages, jump shot percentage, post-up percentage), and record / identify the players exercised moves. After the system gathers all necessary data about a player, this system can grade the player on average and identify a Teams “Starting Line-up” and bench players who can substitute after the system reads low energy levels from an active player. The PPMS is automatically loaded when a player logs-in and can be utilized in both training and coaching modes. This system is also responsible for identifying the Games MVP (Most Valuable Player) based on outstanding performance during a Teams basketball season, the system also keeps records of total game points, total rebounds, assists, total career points, and other important records. Every Skill/ move in the PPMS is saved and can be named to be identified during both training and coaching mode.
Videogame developers have implemented a similar system of storing a players records including “Player Charts”. With this method of recording data automatically based on a Players statistics, we can easily identify key players for each position and construct the right combinations of them. This system of calculating is fully computerized and requires no external calculations making “player ranks” accurate and “player identification” easier to access. Our training system is currently being developed by utilizing standard and improving non standard softwares/hardware that have been developed by known vendors and third party sources. After a lengthed amount of research, we have come up with a number of hardware and software platforms that we will be developing this project on. This artitle is a summary of the "DBT #22" Digital Basketball Training system introduced by a team at the Internet Technology and Enginering Research and Development Center (ITEC) of Huazhong University of Science and Technology in Wuhan China. Newer concepts blooming from this project have be published in other articles. More details are available at the end of the article.
