Enhancing strategic defensive positioning and performance in the outfield

Over the past 20 years, professional and collegiate baseball has undergone a transformation, with statistics and analytics increasingly factoring into most of the decisions being made on the field. One particular example of the increased role of analytics is in the positioning of outfielders, who are tasked with tracking down balls hit to the outfield to record outs and minimize potential offensive damage. This paper explores the potential of location analytics to enhance the strategic positioning of players, enabling improved response and performance. We implement a location optimization model to analyze collegiate ball-tracking data, seeking outfielder locations that simultaneously minimize the average distance to a batted ball and maximize the weighted importance of batted ball coverage within a response standard. Trade-off outfielder configurations are compared to observed fielder positioning, finding that location models and spatial optimization can lead to performance improvements ranging from 1 to 3%, offering a significant strategic advantage over the course of a season.

Visual predictions, neural oscillations and naïve physics

Prediction is a core function of the human visual system. Contemporary research suggests the brain builds predictive internal models of the world to facilitate interactions with our dynamic environment. Here, we wanted to examine the behavioural and neurological consequences of disrupting a core property of peoples’ internal models, using naturalistic stimuli. We had people view videos of basketball and asked them to track the moving ball and predict jump shot outcomes, all while we recorded eye movements and brain activity. To disrupt people’s predictive internal models, we inverted footage on half the trials, so dynamics were inconsistent with how movements should be shaped by gravity. When viewing upright videos people were better at predicting shot outcomes, at tracking the ball position, and they had enhanced alpha-band oscillatory activity in occipital brain regions. The advantage for predicting upright shot outcomes scaled with improvements in ball tracking and occipital alpha-band activity. Occipital alpha-band activity has been linked to selective attention and spatially-mapped inhibitions of visual brain activity. We propose that when people have a more accurate predictive model of the environment, they can more easily parse what is relevant, allowing them to better target irrelevant positions for suppression—resulting in both better predictive performance and in neural markers of inhibited information processing.

A Cluster Analysis Approach to Profile Men and Women’s Volley Positions in Professional Tennis Matches (Doubles)

(1) Background: Tennis ball tracking technology allows the aquirement of novel and reliable data about several performance indicators, such as volley positions. This information is key to understand match dynamics in doubles tennis and to better help preparing players for the demands they will face in match play. As such, the purpose of this study was to describe and compare the different types of volley positions in men’s and women’s doubles professional tennis. (2) Methods: Ball tracking data were collected for 46 women (Billie Jean King Cup) and 96 men’s doubles matches (Davis Cup). The variables used were the distance to the net, the distance to the centre of the court and the height of the impact. A K-Means cluster analysis was used to identify in each subsample different profiles of volley locations. (3) Results: The inferential analysis revealed differences in men’s (distance to the net η2 = 0.72, distance to the centre of the court η2 = 0.77 and impact height η2 = 0.63) and women’s subsamples (distance to the net η2 = 0.48, distance to the centre of the court η2 = 0.52 and impact height η2 = 0.51). (4) Conclusions: The results allowed the suggestion of a higher variability in men’s matches, as there were seven different clusters identified, and only four in women’s.

Validation of Player and Ball Tracking with a Local Positioning System

The aim of this study was the validation of player and ball position measurements of Kinexon’s local positioning system (LPS) in handball and football. Eight athletes conducted a sport-specific course (SSC) and small sided football games (SSG), simultaneously tracked by the LPS and an infrared camera-based motion capture system as reference system. Furthermore, football shots and handball throws were performed to evaluate ball tracking. The position root mean square error (RMSE) for player tracking was 9 cm for SSCs, the instantaneous peak speed showed a percentage deviation from the reference system of 0.7–1.7% for different exercises. The RMSE for SSGs was 8 cm. Covered distance was overestimated by 0.6% in SSCs and 1.0% in SSGs. The 2D RMSE of ball tracking was 15 cm in SSGs, 3D position errors of shot and throw impact locations were 17 cm and 21 cm. The methodology for the validation of a system’s accuracy in sports tracking requires extensive attention, especially in settings covering both, player and ball measurements. Most tracking errors for player tracking were smaller or in line with errors found for comparable systems in the literature. Ball tracking showed a larger error than player tracking. Here, the influence of the positioning of the sensor must be further reviewed. In total, the accuracy of Kinexon’s LPS has proven to represent the current state of the art for player and ball position detection in team sports.

T-RexNet—A Hardware-Aware Neural Network for Real-Time Detection of Small Moving Objects

This paper presents the T-RexNet approach to detect small moving objects in videos by using a deep neural network. T-RexNet combines the advantages of Single-Shot-Detectors with a specific feature-extraction network, thus overcoming the known shortcomings of Single-Shot-Detectors in detecting small objects. The deep convolutional neural network includes two parallel paths: the first path processes both the original picture, in gray-scale format, and differences between consecutive frames; in the second path, differences between a set of three consecutive frames is only handled. As compared with generic object detectors, the method limits the depth of the convolutional network to make it less sensible to high-level features and easier to train on small objects. The simple, Hardware-efficient architecture attains its highest accuracy in the presence of videos with static framing. Deploying our architecture on the NVIDIA Jetson Nano edge-device shows its suitability to embedded systems. To prove the effectiveness and general applicability of the approach, real-world tests assessed the method performances in different scenarios, namely, aerial surveillance with the WPAFB 2009 dataset, civilian surveillance using the Chinese University of Hong Kong (CUHK) Square dataset, and fast tennis-ball tracking, involving a custom dataset. Experimental results prove that T-RexNet is a valid, general solution to detect small moving objects, which outperforms in this task generic existing object-detection approaches. The method also compares favourably with application-specific approaches in terms of the accuracy vs. speed trade-off.

A Survey on Object Detection and Tracking in Soccer Videos

"The players and ball are the most important object in soccer game videos and detected them are a challenging task because of many difficulties, such as shadow and illumination, ball size, several other objects look like a ball, often the ball overlapping with players or merged with lines, as well as the ball may be disappear which be hidden in the stadium or flying on air, and similar appearance of players, etc. The detect ball is the first step for tracking in broadcast soccer video. There are several methods of ball-tracking are based on their problem. In this paper, we have discussed different methods of object detection and tracking in the soccer videos which are available in the literature.