Positive Reinforcement Can Kill
The news media recently reported the death of a man in a gaming café. He had been playing League of Legends for 23 hours. Police suspected that “a combination of tiredness, lack of movement and the cold weather could have caused blood clots and a heart attack.” He literally gamed himself to death. A high frequency of positive reinforcement and rewards is a critical feature of video game design. Player skill and advancement is shaped through progressively more challenging levels and a schedule of reinforcement so dense that it often can be measured in reinforcers-per-second. Not only was this man so absorbed by his gaming experience that it turned into a 23-hour gaming marathon and ultimately his death, but others were so absorbed that nearly 9 hours passed before anyone noticed he had died.
The café was full of gamers earning points and leveling up. Herrnstein’s matching law [1] helps explain this kind of persistence during intense gaming sessions: the rate of behavior in a situation is proportionate to the rate of reinforcement available for that behavior. The availability and density of reinforcement from his gaming were so great that other behaviors were shut out. The gamer who died couldn’t pull himself away from his game, and the other gamers were so focused on their games that they didn’t notice the corpse in the room. Aubrey summed it up in his book, Other People’s Habits when he said, “behavior goes where reinforcement flows.” This unfortunate story provides some reminders about positive reinforcement:
- Positive reinforcement is neither good nor bad – Positive reinforcement strengthens the behavior that it follows. It makes the behavior more likely to occur again regardless of what the behavior is. Positive reinforcement affects unhealthy, unproductive, unsafe, and unethical behaviors just as effectively as it does healthy, productive, safe, and ethical behaviors. Whether or not positive reinforcement produces a desirable result depends on whether or not the behavior being reinforced is something that you want or don’t want. Effective behavior change strategies often require eliminating positive reinforcement for what you don’t want and adding positive reinforcement for what you do want. This application of the matching law gets reinforcement flowing to the right behavior.
- You get more of what’s being reinforced – Regardless of your intentions and what you ask for from others, the behaviors getting the most relative reinforcement will be the most likely to occur. Virtually all work environments have uncontrolled sources of consequences that pull behavior in different directions, sometimes encouraging what you don’t want and discouraging what you do want. Two primary sources of this reinforcement are natural consequences (e.g., work gets done faster or easier) and what peers and managers say and do in response to behavior. If you’re getting too much of the wrong behavior despite clear expectations and demonstrated capability, look to the consequences to find the source of the problem.
- Positive reinforcement can suppress other desired behaviors – When positive reinforcement is much more readily available for one behavior or task than for others, you might get much more of that behavior than you want. This can start in well intended ways such as reinforcing behavior that leads to high productivity. However, if all of the reinforcement is directed toward increasing productivity, safe behavior and behavior aimed at ensuring quality may take a back seat. In extreme cases, addictive behavior can develop when the reinforcement availability and density for that behavior is so high that it suppresses other desired behavior. When reinforcing behavior, be mindful of the range of behavior you’re looking for so that your reinforcement of one behavior does not become detrimental to other desired behaviors.
[1]see Performance Management: Changing Behavior that Drives Organizational Effectiveness