Experimentation and optimization service

1. A computer-implemented method, comprising:

in an automated machine learning procedure, training a tuning system comprising one or more processor devices with game-related data of a plurality of players of an online game;

in an automated operation performed by the tuning system trained with the game-related data of the plurality of players, creating a game variant model for the online game;

in an automated process that is based on the game variant model and that is performed by one or more hardware processors configured therefor, generating one or more rules for building a respective instance of a game variant based on the game variant model;

receiving an indication that a client computing device associated with a requesting player is requesting access to the online game;

identifying one or more attributes of the requesting player;

automatically generating a custom instance of the game variant based at least in part on:

the one or more rules, and

the one or more identified attributes of the requesting player;

sending, from a server computing device, the custom instance of the game variant to the the client computing device for instantiation; and

at the server computing device, thereafter automatically managing online gameplay in the custom instance of the game variant via the client computing device.

2. The computer-implemented method of claim 1 , wherein:

creating the game variant model comprises:

defining at least one target player activity in the online game;

identifying a plurality of player segments, each respective player segment including players that share at least one common attribute; and

for each of the plurality of player segments, generating a respective game variant model based on promoting a likelihood performance of the target player activity by players in the respective player segment, thereby generating a plurality of game variant models;

wherein the generating of the one or more rules comprises generating a respective custom set of rules for each of the plurality of game variant models; and

wherein automatically generating the custom instance of the game variant comprises:

based on the one or more identified attributes of the requesting player, identifying a particular one of the plurality of player segments to which the requesting player corresponds; and

generating the custom instance of the game variant based on the respective custom set of rules for the particular one of the plurality of game variant models associated with the particular player segment to which the requesting player corresponds.

3. The computer-implemented method of claim 2 , wherein defining the at least one target player activity in the game comprises:

defining a target type of transaction request to be received from players that correspond with a particular player segment.

4. The computer-implemented method of claim 2 , wherein defining the at least one target player activity in a game comprises:

defining a target increase in a number of additional players of the online game, wherein each additional player has a respective social network connection with a current player of the online game.

5. The computer-implemented method of claim 2 , wherein identifying the plurality of player segments comprises:

for each player segment:

identifying previous instances of the at least one target activity that have occurred in one or more game sessions played by one or more players who belong to the respective player segment;

identifying at least one setting of an adjustable feature of the game that was present during the one or more game sessions during which the previous instances of the target activity occurred; and

incorporating the at least one setting of the adjustable feature as part of the game variant model for the respective player segment.

6. The computer-implemented method of claim 5 , further comprising:

upon sending the custom instance of the game variant to the client computing device for instantiation:

receiving game performance data of the requesting player after instantiation of the custom in-game instance; and

incorporating the game performance data of the requesting player into the game related data for the particular player segment.

7. The computer-implemented method of claim 6 , further comprising:

updating the game variant model for the requesting player's player segment based on the game performance data of the requesting player.

8. A non-transitory computer-readable medium storing executable instructions thereon, which, when executed by one or more computer processor devices, cause the one or more computer processor devices to perform operations including:

training a tuning system comprising one or more computing devices with game-related data of a plurality of players of an online game;

by use of the tuning system trained with the game-related data of the plurality of players, creating a game variant model for the online game;

based on the game variant model, generating one or more rules for building a respective instance of a game variant based on the game variant model;

receiving an indication that a client computing device associated with a requesting player is requesting access to the online game;

identifying one or more attributes of the requesting player;

automatically generating a custom instance of the game variant based at least in part on:

the one or more rules, and

the one or more identified attributes of the requesting player;

sending the custom instance of the game variant to the client computing device for instantiation; and

thereafter automatically managing online gameplay in the custom instance of the game variant via the client computing device.

9. The non-transitory computer-readable medium of claim 8 , wherein:

creating the game variant model comprises:

defining at least one target player activity in the online game;

identifying a plurality of player segments, each respective player segment including players that share at least one common attribute; and

for each of the plurality of player segments, generating a respective game variant model based on promoting a likelihood performance of the target player activity by players in the respective player segment, thereby generating a plurality of game variant models;

wherein the generating of the one or more rules comprises generating a respective custom set of rules for each of the plurality of game variant models; and

wherein automatically generating the custom instance of the game variant comprises:

based on the one or more identified attributes of the requesting player, identifying a particular one of the plurality of player segments to which the requesting player corresponds; and

generating the custom instance of the game variant based on the respective custom set of rules for the particular one of the plurality of game variant models associated with the particular player segment to which the requesting player corresponds.

10. The non-transitory computer-readable medium of claim 9 , wherein defining the at least one target player activity in the game comprises:

defining a target type of transaction request to be received from players that correspond with a particular player segment.

11. The non-transitory computer-readable medium of claim 9 , wherein defining the at least one target player activity in a game comprises:

defining a target increase in a number of additional players of the online game, wherein each additional player has a respective social network connection with a current player of the online game.

12. The non-transitory computer-readable medium of claim 9 , wherein identifying the plurality of player segments comprises:

for each player segment:

identifying previous instances of the at least one target activity that have occurred in one or more game sessions played by one or more players who belong to the respective player segment;

identifying at least one setting of an adjustable feature of the game that was present during the one or more game sessions during which the previous instances of the target activity occurred; and

incorporating the at least one setting of the adjustable feature as part of the game variant model for the respective player segment.

13. The non-transitory computer-readable medium of claim 12 , further comprising instructions to perform operations including:

upon sending the custom instance of the game variant to the client computing device for instantiation:

receiving game performance data of the requesting player after instantiation of the custom in-game instance; and

incorporating the game performance data of the requesting player into the game related data for the particular player segment.

14. The non-transitory computer-readable medium of claim 13 , further comprising instructions to perform operations including:

updating the game variant model for the requesting player's player segment based on the game performance data of the requesting player.

15. A computer system comprising:

one or more computer processor devices; and

memory having stored thereon instructions executable on the one or more computer processor devices to configure the computer system to perform operations comprising:

training a tuning system comprising one or more computing devices with game-related data of a plurality of players of an online game;

by use of the tuning system trained with the game-related data of the plurality of players, creating a game variant model for the online game;

based on the game variant model, generating one or more rules for building a respective instance of a game variant based on the game variant model;

receiving an indication that a client computing device associated with a requesting player is requesting access to the online game;

identifying one or more attributes of the requesting player;

automatically generating a custom instance of the game variant based at least in part on:

the one or more rules, and

the one or more identified attributes of the requesting player;

sending the custom instance of the game variant to the client computing device for instantiation; and

thereafter automatically managing online gameplay in the custom instance of the game variant via the client computing device.

16. The computer system of claim 15 , wherein:

creating the game variant model comprises:

defining at least one target player activity in the online game;

identifying a plurality of player segments, each respective player segment including players that share at least one common attribute; and

for each of the plurality of player segments, generating a respective game variant model based on promoting a likelihood performance of the target player activity by players in the respective player segment, thereby generating a plurality of game variant models;

wherein the generating of the one or more rules comprises generating a respective custom set of rules for each of the plurality of game variant models; and

wherein automatically generating the custom instance of the game variant comprises:

based on the one or more identified attributes of the requesting player, identifying a particular one of the plurality of player segments to which the requesting player corresponds; and

generating the custom instance of the game variant based on the respective custom set of rules for the particular one of the plurality of game variant models associated with the particular player segment to which the requesting player corresponds.

17. The computer system of claim 16 , wherein defining the at least one target player activity in the game comprises:

defining a target type of transaction request to be received from players that correspond with a particular player segment.

18. The computer system of claim 16 , wherein defining the at least one target player activity in a game comprises:

defining a target increase in a number of additional players of the online game, wherein each additional player has a respective social network connection with a current player of the online game, wherein each additional player has a respective social network connection with a current player of the online game.

19. The computer system of claim 16 , wherein identifying the plurality of player segments comprises:

for each player segment:

identifying previous instances of the at least one target activity that have occurred in one or more game sessions played by one or more players who belong to the respective player segment;

identifying at least one setting of an adjustable feature of the game that was present during the one or more game sessions during which the previous instances of the target activity occurred; and

incorporating the at least one setting of the adjustable feature as part of the game variant model for the respective player segment.

20. The computer system of claim 19 , wherein the instruction set configures the computer system to perform further operations comprising:

upon sending the custom instance of the game variant to the client computing device for instantiation:

receiving game performance data of the requesting player after instantiation of the custom in-game instance; and

incorporating the game performance data of the requesting player into the game related data for the particular player segment.