GameDev Protips: How To Use Operant Conditioning To Improve Your Game’s Design

Every game asks players to react to different situations. Unfortunately, since each person reacts differently to various situations, it can be difficult to guide and shape player reactions as they play. In order to remedy this, you’ll have to go through a process of guiding a player’s reactions in order to get them to do what you want them to. This process is known as “conditioning.” Knowledge of psychology makes the conditioning process easier as it can give you insight into players’ behaviors and states of mind.

Reinforcement is a term found in behavioral psychology which is defined as a conditioning model that serves to “reinforce” a certain behavior following a certain outcome or event — operant conditioning. Out of many kinds of ways to shape behavior, I’ll be covering two: positive reinforcement, and punishment, where undesirable behaviors are punished. If you are having trouble understanding this concept, think about rewarding your dog for catching a ball or how you trained them to sit or roll over. You generally rewarded your dog with food and treats for doing what you want. After a while, your dog will do things on command, with, or without treats. This type of conditioning is positive reinforcement.

Similarly, you can reward players with items or points for achieving objectives in game. Positive reinforcement serves as a good conditioning model because people like to be rewarded, rather than punished. You can also use a different way to shape behavior, such as punishing a player by removing items or hindering their progress. A problem arises with punishment’s effectiveness, however, as people have an easier time processing positive stimuli vs. negative stimuli. Furthermore, excessive amounts of negative stimuli tends to make players frustrated and more likely to give up.

One psychologist whose discoveries had implications relevant to game design was B.F. Skinner. Skinner discovered a new condition of positive reinforcement. By training his test subjects of rats to only eat every tenth time they pressed a lever, as opposed to every time, he was able to save money by not having to buy as much food. This also trained the rats to press the lever more often if they wanted to be rewarded. By experimenting with the frequency of the times he rewarded the rats, he was able to explore different patterns of response. You can do the same sort of thing during games to reward players.

You can start out by providing rewards for very basic actions, but have them taper off over time. For example, you can provide a reward for the player navigating an obstacle course in the tutorial, but you generally won’t have to reward them for negotiating a similar course late in the game. Conversely, you want to create an incentive for the player to progress in the game by increasing rewards as they progress during the game. If a player has just defeated a boss, for example, providing them with more positive feedback such as granting access to a new ability will make them want to continue. Rewarding players appropriately will make players feel successful and make them want to continue playing the game.

Important Takeaways: It is important to guide player reactions during games in order to get them to do what you want. This can be achieved via operant conditioning. Positive reinforcement generally tends to work better because people like to be rewarded and remember rewards better than punishments. It is also a good idea to reward players progressively by providing fewer and fewer rewards for simple actions as they progress, while still providing large ones after major milestones in the game, such as boss levels. This will help keep players interested as everyone loves to be rewarded for a particularly difficult feat. Conversely, excessive punishment and hindering progress will make players frustrated and less likely to continue. It can still work as a method of guiding players but it must be used in moderation if at all.