in liverpool'4-0 win against Barcelona 2019 Champions League Semi-Finalsstriker Divock Origi scored what was later named the club's greatest goal.'130 years of history. Right-back Trent Alexander-Arnold wins a corner kick for his own half, puts the ball inside the corner arc, pretends to walk away, then suddenly turns around and hits a fast low kick straight at Origi, who hits the ball over the top. I kicked it. Left corner of the net.
While soccer doesn't usually have predictable situations, corner kicks are different in that they are highly reproducible and teams spend hours training for them. Alexander-Arnold-Origi's goal was clearly the product of practice, and even a small advantage in the approach strategy to a corner kick can make a big difference in a team's tournament result.
This is where artificial intelligence (AI) could potentially play a role, although it is generally believed that it will have only a limited impact in sports, perhaps outside of board games such as chess and individual sports such as track and field. There is a gender.
deep mind's TacticAI
Google's DeepMind has introduced TacticAI, an AI system that can provide experts with tactical insights, especially regarding corner kicks, through predictive and generative AI.
A corner kick is awarded if the ball crosses the goal line after touching a player from the defending team. The average Champions League game takes around 10 corner kicks, but predicting their outcome is complicated given the randomness of gameplay by individual players and the dynamics between players. .
However, despite the limited availability of gold standard data on corner kicks, TacticAI shows promise by using a “geometric deep learning approach” that helps create more generalizable models. states that they have achieved good results.
predictive and generative
DeepMind developed and evaluated TacticAI in a multi-year research collaboration with experts at Liverpool Football Club.
Identifying the key patterns of tactics carried out by rival teams and developing effective responses is at the heart of modern soccer. According to Veljkovic, TacticAI incorporates both predictive and generative components, allowing coaches to effectively sample and explore alternative setups for their players. Select each corner kick and choose the one with the highest predicted chance of success.
Sports such as soccer are dynamic areas for AI development because they feature real-world, multi-agent interactions with multimodal data. While TacticAI fundamentally demonstrates the potential of assistive AI technology to revolutionize sports for players, coaches, and fans, advances in AI for sports are also expanding beyond the field, from computer games and robotics to traffic coordination. could also be translated into practical models in many fields.
Using AI in corner kicks
Velikovic'The team at DeepMind first published a paper called Game Plan. In this paper, he considered why AI should be used to assist with soccer tactics, and took examples such as the analysis of penalty kicks. Then, in 2022, we developed the “Graph Imputer” to demonstrate how AI can be used in prototyping predictive systems for downstream tasks in soccer analytics.
(B) How TacticAI handles a given corner kick. All four possible combinations of reflections are applied to the corners and fed into his TacticAI model at the core. These interact to compute the final player representation and can be used to predict outcomes.
(Source: DeepMind)
The latest version, TacticAI, is touted as a “complete AI system.” According to the DeepMind team, the service is built to address his three main questions:
- What happens in the tactical setup for a particular corner kick? For example, which player is most likely to receive the ball and whether to attempt a shot.
- Once you play the setup, can you understand what happened? For example, have similar tactics worked well in the past?
- How can I adjust my tactics to achieve a certain result? For example, how should I change the position of my defensive players to reduce the chances of a shot?
This tool represents corner kick situations as a graph, with players represented as nodes and player relationships as edges. A neural network then operates on this graph and updates the representation of each node. This allows us to model interactions between players that may be more important than their absolute distance.
Why especially corners?
In a sense, the emphasis on corner kicks is strategic. The model started by trying to predict aspects of open play in soccer matches, but there are pitfalls. Even if a model gives an indication of the current state of open play, it does not necessarily mean that a coach can meaningfully act on it in the moment. Shouting instructions during open play can confuse players and cause them to interfere with the opposing team's tactics.
Corner kicks are a good place to utilize AI tools to develop your strategy. The main reason for this is that a corner kick is the moment when the game effectively freezes. Corner kicks are always started from the same kind of position at the corner of the pitch, giving players the opportunity to score immediately. Corner strategy is also usually decided long before players actually take to the pitch, so there is no confusion on match day.
