Automated tuning of computer-implemented games

1. A method comprising:

defining a performance metric indicative of player performance in a computer-implemented game having a set of tunable parameters comprising multiple in-game parameters that are variable by a game administrator to change gameplay difficulty;

defining a performance target that represents respective target values for the performance metric at one or more points of progress in the game, the performance target comprising a target curve that represents respective target values for the performance metric at a plurality of different points of progress in the game; and

in an iterative tuning operation performed by one or more computer processor devices configured to perform the iterative tuning operation based at least in part on the performance target, determining a suggested value set for the set of tunable parameters, each iteration of the iterative tuning operation comprising:

executing a gameplay simulation comprising automated playing of the game by an artificial agent in accordance with a candidate set of parameter values for the set of tunable parameters;

determining a simulation result that represents respective values for the performance metric at one or more points of progress in the gameplay simulation; and

automatically modifying the candidate set of parameter values based at least in part on a comparison between the performance target and the simulation result for the gameplay simulation.

2. The method of claim 1 , wherein the performance metric is defined is a function of a plurality of performance measures indicating different respective aspects of player performance in the game.

3. The method of claim 2 , wherein the plurality of performance measures include level progress indicating progression of the player through different levels of the game.

4. The method of claim 2 , wherein the plurality of performance measures include experience points earned by the player.

5. The method of claim 2 , wherein the plurality of performance measures include in-game currency earned by the player.

6. The method of claim 2 , wherein the plurality of performance measures include revenue generated due to in-game purchases by the player.

7. The method of claim 2 , wherein the plurality of performance measures include the number of instances of performance by the player of one or more specified in-game actions.

8. The method of claim 1 , wherein the simulation result comprises a result curve that represents respective values for the performance metric at a plurality of different points of progress in the gameplay simulation.

9. The method of claim 8 , wherein the automatic modification of the candidate set of parameter values is based at least in part on a comparison between the target curve and the result curve.

10. The method of claim 9 , wherein the comparison between the performance target and the simulation result comprises minimizing a difference between respective endpoints of the target curve and the result curve.

11. The method of claim 9 , wherein the comparison between the performance target and the summation result comprises minimizing an in-graph area between the target curve and the result curve.

12. The method of claim 1 , wherein the set of tunable parameters comprises a priority matrix that defines sequential relationships between a plurality of different in-game actions, thus providing one or more predefined sequences in which respective in-game actions are to be performed for achieving corresponding in-game benefits.

13. The method of claim 12 , wherein the executing of the gameplay simulation for at least some iterations of the iterative tuning operation comprises executing the gameplay simulation in accordance with a candidate priority matrix, each corresponding automated adjustment operation comprising automatically adjusting the candidate priority matrix based at least in part on the comparison between the performance target and the simulation result for the gameplay simulation.

14. A system comprising:

an input interface configured to receive

a definition of a performance metric indicative of player performance in a computer-implemented game having a tunable game configuration that is changeable by a game administrator to change gameplay difficulty, the performance metric being a function of a plurality of performance measures indicating different aspects of player performance in the game; and

a performance target that represents respective target values for the performance metric at one or more points of progress in the game, the performance target comprising a target curve that represents respective target values for the performance metric at a plurality of different points of progress in the game; and

an optimization engine configured to determine a suggested game configuration by performing an iterative tuning operation based at least in part on the performance target, each iteration of the iterative tuning operation comprising:

executing a gameplay simulation comprising automated playing of the game by an artificial agent in accordance with a candidate game configuration;

determining a simulation result that represents respective values for the performance metric at different points of progress in the gameplay simulation; and

automatically adjusting the candidate game configuration based at least in part on a comparison between the performance target and the simulation result for the gameplay simulation.

15. The system of claim 14 , wherein the simulation result comprises a result curve that represents respective values for the performance metric at multiple different points of progress in the gameplay simulation.

16. The system of claim 15 , wherein the comparison between the performance target and the simulation result comprises minimizing a difference between respective endpoints of the target curve and the result curve.

17. The system of claim 16 , wherein the comparison between the performance target and the summation result comprises minimizing an in-graph area between the target curve and the result curve.

18. The system of claim 14 , wherein the set of tunable parameters comprises a priority matrix that defines sequential relationships between a plurality of different in-game actions, thus providing one or more predefined sequences in which respective in-game actions are to be performed for achieving corresponding in-game benefits.

19. A non-transitory computer-readable storage medium having stored thereon instructions that, when executed by a machine, cause the machine to perform operations comprising:

enabling definition of a performance metric indicative of player performance in a computer-implemented game having a set of tunable parameters comprising multiple in-game parameters that are variable by a game administrator to change gameplay difficulty;

enabling definition of a performance target that represents respective target values for the performance metric at one or more points of progress in the game, the performance target comprising a target curve that represents respective target values for the performance metric at a plurality of different points of progress in the tame; and

enabling a suggested value set for the set of tunable parameters by performing an iterative tuning operation based at least in part on the performance target, each iteration of the iterative tuning operation comprising:

executing a gameplay simulation comprising automated playing of the game by an artificial agent in accordance with a candidate set of parameter values for the set of tunable parameters;

determining a simulation result that represents respective values for the performance metric at one or more points of progress in the gameplay simulation; and

automatically modifying the candidate set of parameter values based at least in part on a comparison between the performance target and the simulation result for the gameplay simulation.