K2Tekh has developed digital cards for Winline players of the Media Football League based on artificial intelligence

2025

Creating a Digital Card System

K2 Tekh, together with the Winline Media Football League, created a digital card system that replaced subjective expert assessments with objective analysis using AI. This system automatically updates the ratings of football players after each match based on real statistics and provides coaches, analysts and fans with an accurate tool for assessing the current form of players and making tactical decisions. K2 Tekh announced this on August 28, 2025.

The project was implemented using the computing power of PAK-AI from K2 NeuroTech (a division of K2Tex specializing in AI infrastructure and supercomputer clusters).

Digital cards operate on a hybrid scoring system. It combines expert assessments according to the Delphi method with statistical analysis through the Glicko-2 rating model, widely used in e-sports.

Cards collect full player statistics, and a dynamic rating allows you to monitor the progress or decline of a football player's form during the season. The ranking takes into account not only statistics, but also the role of a football player on the field. For example, goals and strikes are most important for the striker, and for the defender - tackles and power struggle. The result is an independent and objective assessment of players, allowing you to correlate parameters and changes in the rating depending on the player's game on the field.

The preparation of the rating consisted of two stages. At the first stage, independent experts evaluated all players according to six criteria: media, strikes, attack, assists, defense and physics. These assessments formed the basis for computer model training.

Then, using the computing resources of the system based on PAK-AI, machine-based training, expert estimates were compared with the real statistics of the players. Thus, K2Tekh specialists created a mathematical model that learned to objectively evaluate football players.

At the second stage, the trained model began to work autonomously. After each match, she analyzes the statistics and automatically updates the player ratings.

Player digital cards will be useful as an analytical tool for assessing players' current form, strengths and weaknesses, making tactical decisions and forming a roster.

Statistics gamified football. Digital cards help visualize the characteristics of each player - from accuracy of passes to efficiency in defense. It's like looking into a coaching summary and understanding the true reasons for wins and losses, "said Nikolai Osipov, president of the Winline Media Football League.

{{quote 'Even in sports, we move away from routine operations that used to be calculated manually and were often subjective. Artificial intelligence allows you to create models that not only analyze various data sources, but also reveal hidden patterns and dependencies. As a result, we get an accurate, objective and dynamic assessment tool, which is updated after each match, giving out actual metrics, - commented Svyatoslav Smirnov, head of the division K2 NeuroTech. }}

Creating a Digital Analytics

K2 NeuroTech has created a digital analytical platform - a personal account for Winline matches of the Media Football League, which allows coaches and teams to receive advanced data on the match and training in real time. The system processes more than 30 game parameters and visualizes tactical models in the web interface, turning "raw" data into application recommendations for strategy and individual training. K2 Tekh announced this on July 21, 2025.

The platform is based on a three-dimensional collection and processing system, information which, using, algorithms artificial intelligence converts the actions of players on the field into visual analytics: tactical schemes, individual indicators and team statistics. Data for personal account are processed on the PAK-AI software and hardware complex. PAK-AI is a ready-made system for rapid integration AI into an enterprise IT environment that combines a hardware layer (, servers graphics accelerators,) storage systems with a single web management portal.

It is a premium solution that provides the convenience of a cloud AI platform, while allowing you to comply with regulatory requirements in terms of information security and technological independence. Thanks to powerful computing resources (IaaS) and specialized software (PaaS), which includes tools for working with artificial intelligence and data processing, PAK-AI provides visibility and usability, guaranteed performance, high speed and depth of analysis. This allows the system to process huge amounts of data in real time and identify the smallest details: from the total run and speed of players to the accuracy of passes, interceptions and scoring chances.

The platform is not just a collection of data, but an intelligent analytical tool that processes hundreds of game parameters in real time. My account consolidates this data, turning it into easy-to-read insights and recommendations for coaches and players. Such an analytical tool can be adapted for any areas where operational and accurate solutions are important, "said Alexey Zotov, head of IT infrastructure at K2Tekh .

Key features of the personal account:

• Tactical Analytics:
  • tracking of line of defense ruptures and penetrations into the penalty area by zones and time of the match;
  • visualization of attack directions and team activity by game segments.

• Personal player statistics:
  • fitness data, including overall mileage and actions in attack and defense;
  • technical and tactical indicators, such as shots on target and the effectiveness of standards.

• Team game widgets:
  • analysis of possession, accuracy of passes and success of defensive actions on the field.

The platform allows you to flexibly customize reports and filter data by player, match, and type of action to conduct in-depth analysis of game strategy. Thanks to integration with video analytics, the system automatically generates collections of successful combinations - shooting is carried out on a camera with a viewing angle of 180 degrees.

Personal account is a good solution for our league, which opens up new horizons in the analysis of the game. It gives coaches and players the opportunity to comprehend the match in a new way, identify hidden patterns and quickly adapt strategies on the field. It's like having a personal analyst who is always willing to make valuable recommendations to improve outcomes. We are confident that such support will be an important step towards the success of teams in media football, "said Nikolai Osipov, president of the Winline Media Football League.

Four key metrics in the personal account - attack, defense, possession and physics - are available both for each player and for comparison throughout the team. Thanks to this data, we can easily determine who was the best on the field and at the expense of what. Analytics on numbers does not just tell who won, but shows why, - said Pavel Grigorenko, coach-analyst of the BROKE BOYS team.

Sports analytics is just one example of how AI-tems help streamline processes. In, industries for example, a system for collecting, transmitting, analyzing visualization of data and from sensors installed on equipment makes it possible to accurately predict malfunctions of individual units of processing equipment or identify violations of operating conditions. As a result, the costs of equipment downtime due to unforeseen breakdowns are significantly reduced. Another case digital twin is a warehouse based on the system video surveillances and neuronets that recognizes the movement of employees and all kinds of objects. Up-to-date information about their information status is received accounting system in real time, which analyzes, monitors and allocates resources for effective inventory management. This experience can be replicated on companies of any scale from various. industries

Application of AI for detailed analysis of media football matches

In partnership with the Winline Media Football League K2 NeuroTech has applied AI technology for detailed analysis of media football matches. Now every action on the field is recorded in real time: from attacks and assists to players' fitness performance. Digital tools not only made broadcasts more spectacular, but also opened up access to analytics for teams. K2 Tekh announced this on May 20, 2025.

Artificial intelligence is a game changer in a variety of fields, including sports. Instead of manual analysis, which is time consuming and prone to human error, neural networks process the video stream in real time, capturing each action on the playing field. In partnership with the Winline Media Football League, K2Teh implemented a project in which AI technologies were used for detailed analytics of Winline matches of the Media Football League Cup. Thanks to the products based on IaaS and PaaS, which are part of PAK-AI, it was possible to collect data, pre-process video and track objects. This made it possible to analyze and broadcast information to viewers on the screen in real time.

The PAK-AI-based system processed the video stream from several cameras, automatically recording key events: attacks, assists, selections and even fitness indicators of players. Viewers in real time saw thermal maps of football players' activity, predictive analytics of the outcome of the match and detailed statistics on the halves. This data not only increased the entertainment of broadcasts, but also provided coaches with tools to analyze the effectiveness of tactical schemes.

Modern models of artificial intelligence - whether generative, linguistic, discriminative, or video handlers - together with various machine learning techniques such as reinforcement learning (RL), generative adversarial networks (GANs), and principal component analysis (PCAs) are firmly embedded in our daily lives. Neural networks are able to automatically create video collections of errors and successful actions of players, offer alternative scenarios for the development of attacks and assess the physical condition of the team based on GPS tracking data. This includes not only basic metrics such as possession or number of shots, but also complex tactical patterns. For example, analysis of pressure zones, the effectiveness of pressing or moving players without the ball. Through computer vision and deep learning algorithms, clubs gain access to data that was previously unavailable. Technologies are also used in scouting: neural networks analyze player databases, identifying promising candidates, taking into account the requirements of the coaching staff. In addition, artificial intelligence models take on routine tasks, such as preparing reports.

In the coming years, generative AI and predictive analytics will be key elements of the sports industry. Russian clubs will be able to create interactive content, including personalized broadcasts and automatic match reports, adapted to the interests of viewers. However, the main quality of local solutions, such as PAK-AI, is the high security of data and the possibility of customization to the requirements of each team or league, "explained Alexey Zotov, head of IT infrastructure at NeuroTech K2.

Modern technology, such as artificial intelligence, not only improves the quality of broadcasts, but also makes football an interactive experience for viewers. We create an atmosphere where everyone can feel part of the event, and advanced analysis of the game allows us to deeper understand what is happening on the field, "said Nikolai Osipov, president of the Winline Media Football League.