Abstract: Techniques are described herein for facilitating the placement of props on maps by an automated prop placement tool that makes use of a trained machine learning mechanism. The machine learning mechanism is trained based on one or more training maps upon which props have been placed. The machine learning mechanism may be trained to suggest placement based on (a) spatial rules relating, (b) prop-specific rules, (c) prop-to-fixed-object distances between props and map structures, and (d) distances between props. Once the machine learning mechanism is trained, the prop placement tool may be provided as input (a) map data that defines a target map and (b) prop data that specifies the set of target props to be placed on the target map. Based on this input and the machine learning mechanism's trained model, the prop placement tool outputs a suggested placement, for each of the target props, on the target map.

Abstract: Systems and methods for improving initial results of a reinforcement learning model are described herein. In an embodiment, a server computer initiates a reinforcement learning model for a modeled system. While executing the reinforcement learning model, the server computer computes a first result value for a particular action using the reinforcement learning model and a second result value for the particular action using a heuristic separate from the reinforcement model. Based, at least in part, on the first result value for the particular action and the second result value for the particular action, the server computer performs the particular action. The server computer determining a result from performing the particular action and updates the reinforcement learning model.

Abstract: A method, apparatus, and stored instructions are provided for selectively incorporating feedback from a remote audience into live content. The feedback is incorporated into one part of a segment at the same time that another part of the segment is played to an audience. A feedback manager may receive feedback from audience members on a first part of the segment while a second, later part of the segment is being played to the audience by a client machine and while a third, even later part of the segment is being generated by a content manager. The feedback manager may then compare feedback data to stored criteria to determine whether or not to change the playing content.

Abstract: Methods, computer-readable media, and computing devices are described herein for restoring gameplay by replaying past inputs. During gameplay, capture logic captures participant inputs that were received to change states of game objects. The capture logic stores the captured information in an electronic record of period(s) of the gameplay. After the period(s) of gameplay, replay logic receives a request to initiate subsequent gameplay beginning from a point in time from a period of past gameplay. The replay logic causes the subsequent gameplay to begin from the point at least in part by loading saved state information and replaying the captured participant inputs.

Abstract: Techniques are described herein for a gaming system that maintains an internal score and an external score for each user. The system maintains the internal scores to reflect demonstrated user skill. The system uses the accurate internal scores to facilitate communication between client computing devices associated with users that have been selected to participate in gameplay instances based on the internal scores. The external scores maintained by the system displayed to the users. Because the external scores are not used for matchmaking, the external scores may be adjusted in any way without requiring internal consistency. For example, to combat gameplay stagnation, the gameplay system periodically resets maintained external scores. For a time, comparisons between the reset external scores do not accurately reflect the users' rankings within the game. However, as the users participate in gameplay, the external scores return to levels that are similar to the users' internal scores.

Abstract: Systems and methods for modifying three-dimensional digital items to fit different character models are described herein. In an embodiment a machine learning system is configured to compute a shape and size of three-dimensional digital objects to fit a second character model based on the shape and size that the same three-dimensional digital objects have to fit a first character model. A server computer receives particular input data defining a plurality of particular input vertices for a particular input three-dimensional digital object fit for the first character model. In response to receiving the particular input data, the server computer computes, using the machine learning system, particular output data defining a plurality of particular output vertices for a particular output three-dimensional digital object, the particular output three-dimensional digital object comprising the particular input three-dimensional digital object fit for the second character model.

Abstract: Techniques are described for a multiple-phase process that uses machine learning (ML) models to produce a texturized version of an input image. During a first phase, using a pix2pix-based ML model, an automatically-generated image that depicts structured texture is generated based on an input image that visually identifies a plurality of image areas for the structured texture. During a second phase, a neural style transfer-based ML model is used to apply the style of a style image (e.g., a target image from the training dataset of the pix2pix-based ML model) to the texture image generated at the first phase (the content image) to produce a modified texture image. According to an embodiment, during a third phase, the generated texture image produced at the first phase and the modified texture image produced at the second phase are combined to produce a structured texture image with a moderated amount of detail.

Abstract: A method, apparatus, and stored instructions are provided for selectively incorporating feedback from a remote audience into live content. The feedback is incorporated into one part of a segment at the same time that another part of the segment is played to an audience. A feedback manager may receive feedback from audience members on a first part of the segment while a second, later part of the segment is being played to the audience by a client machine and while a third, even later part of the segment is being generated by a content manager. The feedback manager may then compare feedback data to stored criteria to determine whether or not to change the playing content.

Abstract: Methods, computer-readable media, and computing devices are described herein for restoring gameplay by replaying past inputs. During gameplay, capture logic captures participant inputs that were received to change states of game objects. The capture logic stores the captured information in an electronic record of period(s) of the gameplay. After the period(s) of gameplay, replay logic receives a request to initiate subsequent gameplay beginning from a point in time from a period of past gameplay. The replay logic causes the subsequent gameplay to begin from the point at least in part by loading saved state information and replaying the captured participant inputs.

Abstract: A method, apparatus, and stored instructions are provided for selectively incorporating feedback from a remote audience into live content. The feedback is incorporated into one part of a segment at the same time that another part of the segment is played to an audience. A feedback manager may receive feedback from audience members on a first part of the segment while a second, later part of the segment is being played to the audience by a client machine and while a third, even later part of the segment is being generated by a content manager. The feedback manager may then compare feedback data to stored criteria to determine whether or not to change the playing content.

Abstract: In an approach, a game server records events that occur within a match of a video game played using a plurality of game clients. After the match has concluded, the game server scores the events according to a plurality of criteria corresponding a plurality of play of the game categories. A sliding window is passed over the events in a number of increments. During each increment, the score for each event that falls within the sliding window is aggregated for each of the categories. The game server then selects a play of the game category and determines the top aggregated score for that category. Once determined, the game server sends one or more instructions to the game clients which causes the game clients to display a replay of the events that occurred during the time window increment corresponding to the top aggregated score for the selected category.

Abstract: Methods, computer-readable media, and computing devices are described herein for restoring gameplay by replaying past inputs. During gameplay, capture logic captures participant inputs that were received to change states of game objects. The capture logic stores the captured information in an electronic record of period(s) of the gameplay. After the period(s) of gameplay, replay logic receives a request to initiate subsequent gameplay beginning from a point in time from a period of past gameplay. The replay logic causes the subsequent gameplay to begin from the point at least in part by loading saved state information and replaying the captured participant inputs.

Abstract: A method, apparatus, and stored instructions are provided for selectively incorporating feedback from a remote audience into live content. The feedback is incorporated into one part of a segment at the same time that another part of the segment is played to an audience. A feedback manager may receive feedback from audience members on a first part of the segment while a second, later part of the segment is being played to the audience by a client machine and while a third, even later part of the segment is being generated by a content manager. The feedback manager may then compare feedback data to stored criteria to determine whether or not to change the playing content.

Abstract: Methods, computer-readable media, and specially configured machines are described for hosting an instance of a cross-realm zone that manages interaction among characters from different instances of a virtual world in a massively multiplayer online game. Different zones of the virtual world may support different numbers of realms or different combinations of realms, and some zones of the virtual world may remain as single-realm zones. When a character enters a zone, the character may be added to an instance of a cross-realm zone based on the virtual world for which the character is a member. A single cross-realm zone may handle all characters that enter the zone from a subset of realms, but not characters that enter the zone from other realms that are not in the subset of realms. Characters in a cross-realm zone might not have any affiliation or prior social connection or interaction with each other.

Abstract: Methods, computer-readable media, and computing devices are described herein for restoring gameplay by replaying past inputs. During gameplay, capture logic captures participant inputs that were received to change states of game objects. The capture logic stores the captured information in an electronic record of period(s) of the gameplay. After the period(s) of gameplay, replay logic receives a request to initiate subsequent gameplay beginning from a point in time from a period of past gameplay. The replay logic causes the subsequent gameplay to begin from the point at least in part by loading saved state information and replaying the captured participant inputs.

Abstract: In an approach, one or more player characters controllable by a first player account perform an action in an instance of a game world. In the event that the action meets one or more conditions for creating a nemesis, a computer system controlling the game instance selects a second player account to be invaded by a nemesis. When the second player account plays the game in the same or a different game instance, the computer system determines whether second one or more characters of the selected account meet one or more invasion conditions. If the invasion conditions are met, the computer system generates computer-controlled enemy NPCs for second one or more characters of the second player account to defeat. If the second account defeats the nemesis invasion, rewards are generated for the first and second player accounts. Otherwise, the computer system chooses a third player account for nemesis invasion.

Abstract: Methods, computer-readable media, and specially configured machines are described for hosting an instance of a cross-realm zone that manages interaction among characters from different instances of a virtual world in a massively multiplayer online game. Different zones of the virtual world may support different numbers of realms or different combinations of realms, and some zones of the virtual world may remain as single-realm zones. When a character enters a zone, the character may be added to an instance of a cross-realm zone based on the virtual world for which the character is a member. A single cross-realm zone may handle all characters that enter the zone from a subset of realms, but not characters that enter the zone from other realms that are not in the subset of realms. Characters in a cross-realm zone might not have any affiliation or prior social connection or interaction with each other.

Abstract: Methods, computer-readable media, and computing devices are described herein for restoring gameplay by replaying past inputs. During gameplay, capture logic captures participant inputs that were received to change states of game objects. The capture logic stores the captured information in an electronic record of period(s) of the gameplay. After the period(s) of gameplay, replay logic receives a request to initiate subsequent gameplay beginning from a point in time from a period of past gameplay. The replay logic causes the subsequent gameplay to begin from the point at least in part by loading saved state information and replaying the captured participant inputs.

Abstract: Methods, computer-readable media, and computing devices are described herein for restoring gameplay by replaying past inputs. During gameplay, capture logic captures participant inputs that were received to change states of game objects. The capture logic stores the captured information in an electronic record of period(s) of the gameplay. After the period(s) of gameplay, replay logic receives a request to initiate subsequent gameplay beginning from a point in time from a period of past gameplay. The replay logic causes the subsequent gameplay to begin from the point at least in part by loading saved state information and replaying the captured participant inputs.