<-Scope Go to ToC References->
Capitalised Terms have the meaning defined in Table 1 or by other MPAI Technical Specifications whose definition is accessible online. Non-capitalised terms have the meaning commonly defined for the context in which they are used.
A dash “-” preceding a Term in Table 1 means the following:
- If the font is normal, the Term in Table 1without a dash and preceding the one with a dash should be placed before that Term. The notation is used to concentrate in one place all the Terms that are composed of, e.g., the word Client followed by the word Cheating, Data or Prediction.
- If the font is italic, the Term in Table 1 without a dash and preceding the one with a dash should be placed afterthat Term. The notation is used to concentrate in one place all the Terms that are composed of, e.g., the word Engine preceded by one of the words Behaviour, Game State, Physics, and Rules.
All MPAI terms are accessible online here.
Table 1 – MPAI-SPG specific terms and acronyms
| Term | Acronym | Definition |
| authoritative multiplayer game | ||
| Checkpoint | An invisible gate which a car passes through during a race. A car must go through all the checkpoints in order to complete a lap. | |
| Client | ||
| – Cheating | An operation of a client deliberately sending wrong data to the game server. | |
| – Data | CD | Any data sent by the client to the server. |
| – Prediction | A technique used on the client in an authoritative server architecture, to predict a Game State. | |
| Curriculum Learning | A training strategy which starts with simpler examples and gradually increasing complexity. | |
| Engine | ||
| – Behaviour | The Engine that computes the evolution of the Game State by translating the inputs of the players into corresponding actions. | |
| – Game State | A process managing the Game State of the server. | |
| – Physics | The Engine that computes the evolution of the Game State using the rules underpinning the physics of the Environment. | |
| – Rules | The Engine that computes the evolution of the Game State using the rules underpinning the game logic. | |
| Entity | Any object or character within the game world that can interact with other objects or characters. | |
| Environment | The virtual space where the game takes place and the objects populating it. | |
| Environmental Data | Data regarding the Environment: Tile type and Tile ranking | |
| Game | ||
| – Clock Period | The time required by the game engine to update the game state. | |
| – Message | GM | Messages exchanged between the Game State Engine and the Physics, Rules and Behaviour Engines containing information proper of the relevant Engine. |
| – Message* | GM* | Messages that each of the Physics, Rules, and Behaviour Engines-AIs receive from the Game State DMUX and send to the Game State Assembler. |
| – Server | ||
| – State | GS | A data type representing all the information about the game at a given instant. |
| – State DMUX | GS-DMUX | AIM that demultiplexes the GS received from the GS Engine into discrete Game Messages* ( ). |
| Gaming | ||
| – Authoritative multiplayer | A type of gaming where a central system maintains the definitive game state, enforces rules, and resolves player actions to ensure consistency, fairness, and synchronization across participants. | |
| – Multiplayer | A form of video gaming that allows multiple players to interact and compete against each other or cooperate as teammates over a network, either locally or online. | |
| – Predictive Multiplayer | A technique that anticipates player actions and decisions, enhancing the gaming experience by reducing the impact of latency in multiplayer gaming. | |
| Long Short-Term
Memory |
LSTM | A type of recurrent neural network architecture, designed to recognize patterns in sequences of data. |
| Machine Learning | ML | A Process using training data to create a Model able to perform specific tasks such as classification and regression. |
| Model | A Data Type representing an Artificial Neural Network. | |
| Multilayer Perceptron | MLP | A class of feedforward artificial neural network that consists of at least three layers of nodes: an input layer, one or more hidden layers, and an output layer. |
| Non-Player Characters | NPC | Game entities not controlled by the clients’ inputs. |
| Neural Network | NN | A set of interconnected data processing nodes whose connections are affected by weights. Also referred to as Artificial Neural Network. |
| Prediction | ||
| Predicted Game State Assembler | GS-ASS | These predictions are assembled by the Predicted Game State Assembler (GS-ASS), forming the predicted Game State ( ), which is then communicated back to the server for the next iteration of Game State evaluation. Note that, the formats of both, the computed and predicted GSs are the same. |
| Ranking | A position in a hierarchy or scale. | |
| Server Prediction | A technique used on the server, to predict the next Game State without the Client Data. | |
| Spatial Attitude | SA | A Data Type representing an object’s Position, Orientation and their velocities and accelerations. |
| Tile | A portion of the racetrack. | |
| – Ranking | The position of the Tile from the start of the racetrack. | |
| – Type | The shape of the Tile: straight, narrow curve and wide curve. | |
| Time | ||
| – Delay | A Latency effect mitigation technique used to keep the game synchronised in all clients. | |
| – Warp | A latency effect mitigation technique used on the server to evaluate the consequences of a client’s action based on their latency delay. |
