In this decision, which might be interesting for game developers, the European Patent Office refused to grant a software patent on a method for controlling an avatar. Here are the practical takeaways of the decision T 3033/18 of August 2, 2021 of Technical Board of Appeal 3.2.04:
The Board in charge summarized the invention of the application underlying the present decision as follows:
3.1 The present invention relates to the control of an avatar (self-role in the words of the claim) having virtual action parts (limbs) in a virtual environment.
Fig. 2 of EP 2 674 204 A1
Here is how the invention is defined in claim 1:
Claim 1 (main request)
A man-machine interaction controlling method, wherein kinetic parts of the user are associated with the virtual action parts of the self-role, such that the kinetic part of the user and the associated virtual action part of the self-role are the same, the method comprises steps of:
1) creating the self-role in the virtual environment, wherein the self-role has more than one virtual action part;
2) associating kinetic parts of the user with the virtual action parts of the self-role, wherein the user is lying or sitting at the operating position;
3) setting enlarging proportion of an action range of different virtual action parts relative to action range of the kinetic parts of the user;
4) performing the actions by the user while sitting or lying at the operating position, and obtaining actions performed by the user while the user sits or lies at the operating position; and
5) enlarging the obtained actions according to the enlarging proportion so as to achieve virtual action of the self-role, and enabling the virtual action parts of the self-role to perform the virtual action according to the actions by the kinetic parts of the user, wherein while the self-role is not sitting nor lying during the virtual action in the virtual environment, movement morphology of the self-role is different and not similar with movement morphology of the user, and wherein the user’s part supported by the operation position is different from the self-role’s part supported by the virtual environment, and further the user remains at the operating position while performing the actions required for controlling the virtual action of the self-role.
Is it technical?
The present application was rejected by the Examining Division due to lack of inventive step. At the appeal stage, the appellant argued that the following two aspects are not disclosed by the closest prior art searched in the first instance:
1) wherein while the self-role is not sitting nor lying during the virtual action in the virtual environment, movement morphology of the self-role is different and not similar with movement morphology of the user, and wherein the user’s part supported by the operation position is different from the self-role’s part supported by the virtual environment, and further the user remains at the operating position while performing the micro-actions required for controlling the virtual action of the self-role and
2) an action range of the virtual action part of the self-role has a plurality of enlarging proportions at different times with respect to an action range of the associated kinetic part of the user (i.e., the same user part).
According to the appellant, each of the two aspects provides for a separate technical advantage:
The advantageous effect of the distinguishing feature 1 is that, since the user’s part supported in the operation position is different from the self-role’s part supported in the virtual environment, the invention does not require the direct mapping of the actions of the user to the actions of the self-role. Furthermore, the self-role is not bound by the virtual environment as the user is by their physical environment (see paragraph  of the application, “the user can control the self-role perform a back bend action as shown in FIG. 3 while performing a punch action.”).
The distinguishing feature 2 has the advantageous effect that it presents more freedom for the user part when controlling the virtual self-role part.
The Board agreed in that the claimed subject-matter implies technical means required for performing the claimed method, e.g. means for obtaining the user’s action in the physical world which are then used for controlling the avatar. However, the Board in charge further explained that aspects as mentioned above relating to gaming rules could only serve the explicit purpose of playing a game. Hence, these aspects could only be considered an abstract, mental construct.
3.2 The Board considers (see T0336/07, reasons 3.3.1) game rules to form part of “[…] a regulatory framework agreed between players and concerning conduct, conventions and conditions that are meaningful only in a gaming context. It is normally so perceived by the players involved, and as serving the explicit purpose of playing a game. As such an agreed framework it is a purely abstract, mental construct, though the means for carrying out the game play in accordance with such a set may well be technical in nature”. Moreover, (see T0012/08, reasons 4.6) game rules “form the abstract formal structure of a game describing the interplay between player actions and the choices offered within the game.”
3.3 A set of game rules thus determines inter alia how game-play evolves from beginning to end in response to player actions and decisions and the goals to be achieved to conclude game-play.
Sine the claimed subject-matter is a mixture of technical and non-technical features, the Board applied the well-known COMVIK approach for assessing inventive step:
3.5 In dealing with such “mixed” inventions, the Board adopts the approach as set out in T1543/06 (Gameaccount), reasons 2.1-2.9, which is based foremost on T0641/00 (OJ EPO 2003, 352). Thus, only those features that contribute to technical character are to be taken into account when assessing inventive step.
As a next step, the Board determined the distinguishing features of claim 1 of the main request as follows:
4.3 Therefore, the subject matter of claim 1 differs from D1 in respect of its features 3 and 5.
The first of these (feature 3), the Board summarises as:
– setting an enlarging proportion of virtual action parts to the user’s kinetic parts means setting the scaling factor by which a user’s performed action is to be scaled up on its avatar virtual action part.
The second of these (feature 5), the Board summarises as:
– Scaling up the obtained [user] actions by the enlarging proportion (scaling factor) to control the the self-role [avatar] in carrying out virtual actions.
– The self role [avatar] is neither sitting nor lying when performing.
– The movement morphology of the self-role is different from and not similar to the movement morphology of the user, and
– the user’s part supported by the operation position is different from the self-role’s part supported by the virtual environment (for example the seated real world user is supported by their buttocks, whereas the avatar is supported by the soles of its feet in the virtual game world, cf. published application paragraph ).
– the user remains at the operating position while performing the actions required for controlling the virtual action of the self-role.
In the Board’s view, however, all the distinguishing features would only be governed by the rules of the game concerned. Specifically, scalling-up a real world movement performed by a user and mapping the movement to an avatar is only a convention of a game and thus cannot belong to a solution of a technical problem:
4.4.1 Taking a certain movement in the real world and scaling it up in the movement of an avatar will be perceived by the player as belonging to the framework and conventions of the game. They will know that, in the game context, when they wish the avatar to make a certain movement of its action part (for example a leg) in the virtual world, they need to move the corresponding part of their body (their leg) by a certain smaller amount in the real world. This scaling up of a real world movement in a game world is a typical game concept. For example, in a notorious joystick controlled game, a small joystick movement typically maps onto a much larger movement of an entity being controlled in the game world and the player understands this to be part of the convention of the game.
According to the Board, the problem to be solved could be formulated as follows:
4.6 This question is to be considered from the point of view of the skilled person – here a gaming software engineer – solving the objective technical problem of modifying D1’s man-machine interaction controlling method to implement the above rules.
Since this would be nothing more than a mere implementation of non-technical features of a game, the distinguishing features cannot contribute to inventive step.
Since similar objections would also apply to the auxiliary requests filed by the appellant, the Board dismissed the appeal.
You can read the whole decision here: T 3033/18 of August 2, 2021