The Super Mario Effect: Punishment isn't Free
Mark Rober, a former NASA engineer, ran an experiment on 50,000 people. He gave them all the same coding puzzle but he split up the people into two groups. When Group One failed they saw the message, "That didn't work. Please try again." Group Two saw, "That didn't work. You lost 5 points. You now have 195 points. Please try again."
The no-penalty group had a 68% success rate and the group averaged 2.5 more attempts before completing the puzzle than the other group. The group that lost points had only a 52% success rate and there were far fewer total attempts.
Same puzzle, same difficulty, same failures along the way. The only variable was whether failure cost something. That one change was enough to cut the success rate and willingness to try.
Rober named this the Super Mario Effect, because it explains something every kid already knows intuitively: nobody quits Super Mario just because Mario keeps dying.
Kids will run Mario into the same pit forty times and experience plenty of emotions doing it but their motivation does not falter. The failure alone isn't enough to destroy motivation.
But when you add a penalty, the whole dynamic changes.
In the video game example, as in the puzzle, we know that failure will eventually destroy motivation because failure naturally comes with penalties. It costs you time, effort, assets, and maybe even reputation. And those unavoidable losses do add up. If someone fails long enough, they will quit trying.
However, when leadership layers penalties on top of failure it eliminates persistence and destroys the willingness to take a chance.
The core takeaway for leaders should be that the act of punishing failure in a task where success isn't assured will have the same effect as punishing the act of trying.
Think about what a failure means for someone who is genuinely attempting a task. They aimed at an outcome and missed. The only thing missing was capability: skill, information, or maybe just practice. That's it. There's no motivation problem to solve, because they were already motivated enough to attempt it even knowing they might fail.
So what does punishment accomplish in that moment? It can't build the missing capability. Punishment carries no instructional content. It doesn't tell someone what to do differently next time. All it does is attach more pain to failure.
And since attempting can't be separated from failing in a motivated individual, you end up punishing the effort itself. You're not disciplining bad performance. You're disciplining the decision to try.
That's why admonishing a good-faith attempt is never neutral, it's corrosive. It's the reason Rober's second group didn't just fail at a higher rate, they stopped trying sooner. People learn that trying is the risky part, not failing. The rational response is to try less, hide attempts, or wait until success is guaranteed before acting at all. You get quieter failure, not fewer failures.
This isn't to say that you should remove consequences from real negligence. It's correctly diagnosing what a failed attempt is telling you. If someone was trying and came up short, you have a capability gap and capability gaps close with help and/or more trying. They do not close with a penalty.
Punishment has a place. It belongs where there was no real attempt at all or where the gap wasn't capability but choice.
Confuse the two, and you'll spend your energy punishing your best people for the crime of trying, then promoting those who didn't try at all. And your organization will run like the lower performing side of Mark Rober's experiment.
Accountability doesn't have to make people afraid to fail.
The right system can reinforce standards, close knowledge gaps, and help people improve without turning every mistake into punishment.
Comments