This is the public page of Server-based Predictive Multiplayer Gaming (MPAI-SPG). a standard aiming to minimise the audio-visual and gameplay discontinuities caused by high latency or packet losses during an online real-time game. In case information from a client is missing, the data collected from the clients involved in a particular game are fed to an AI-based system that predicts the moves of the client whose data are missing. The same technologies provide a response to the need to detect whether a player is cheating.

  1. MPAI-SPG Progress report and plans

The MPAI-SPG Reference Model is given by the figure. The right-hand side is a regular Online Game Server with its clients. The Server is composed of a Game State Engine and specific-functionality Engines (e.g., Physics Engine, Behaviour Engine and Rules Engine). Each Game State with the Control data just received are communicated to the left-hand side which is a digital twin of the Online Game Server composed of trained neural networks (NN) capable of predicting the next Game State even in absence of some control data.

The MPAI-SPG standard is applicable to both cloud gaming and online gaming. “Client” is used to indicate both Virtual Clients and Clients.

More details about the operation of the MPAI-SPG subsystem are:

  1. The Controller Data and the Game State Engine enter the Game State Engine-AI.
  2. The Game State Engine-AI produces a set of GM*.

In normal conditions, the set of GM* should be close to the set of GM, because the NN has been trained.

  1. The Controller Data and GM* feed the Game Engines AI’s.
  2. The Game State Engine-AI’s produce estimates of the set of Game Messages (GMp).

In normal conditions, the set of GMp should be close to the set of GM’ because the NNs have been trained.

  1. The green Game State Engine, functionally equivalent to the blue Game State Engine,
    1. Computes the predicted Game State (GSp).

In normal conditions, GSp should be close to GS.

  1. Sends the predicted Game State (GSp) to the Game State Engine.
  2. The Game State Engine
    1. Corrects its GS if
      1. GSp is too different than GS (anticheating).
      2. The Game State Engine was missing some Controller Data.
    2. Sends its current GS (GS) to the Game State Engine-AI.

MPAI-SPG is at the level of Use Cases and Functional Requirements. See the latest MPAI-SPG Status report


If you wish to participate in this work you have the following options

  1. Join MPAI
  2. Participate until the MPAI-SPG Functional Requirements are approved (after that only MPAI members can participate) by sending an email to the MPAI Secretariat.
  3. Keep an eye on this page.

Return to the MPAI-SPG page