Why your leaderboard dies after the fifth repeat winner

It’s a pattern I see all the time in the web analytics for businesses that run competitions, loyalty programs, or community challenges. You launch a shiny new leaderboard. Engagement spikes. People are refreshing the page like it’s the last five minutes of the footy. Then, week three hits, and the same two names are sitting at the top. By week five, the comments section is a ghost town. What happened? You didn’t change the prize, the design, or the rules. The only thing that changed is that people realised the game was already over.

The answer isn’t in your code. It’s in the psychology of how we process repeated outcomes. Your leaderboard didn’t die because it was boring. It died because it accidentally triggered a cognitive shutdown that makes us stop caring about anything that feels predetermined.

The Dopamine Tax on Certainty

Let’s talk about the brain’s reward system, because that’s where your leaderboard lives. The late, great behavioural psychologist B.F. Skinner famously demonstrated that rats respond most vigorously to variable-ratio reinforcement — a reward that comes after an unpredictable number of responses. That’s why a slot machine is more compelling than a vending machine. The vending machine gives you a chocolate bar every single time. The slot machine gives you a win sometimes, and the not knowing is what keeps the lever being pulled.

Your leaderboard, in its current form, is a vending machine. The same person wins, the same way, every time. The outcome is certain. And here’s the kicker: certainty is neurologically expensive to maintain interest in. When the brain predicts an outcome with 100% accuracy, it stops releasing dopamine in anticipation. Dopamine isn’t about the reward itself; it’s about the prediction error — the gap between what you expected and what you got.

When you have a repeat winner, the prediction error for everyone else drops to zero. There’s no surprise, no "maybe this time" moment. The leaderboard becomes a static document, not a game. And a static document is just a PDF with extra steps.

Loss Aversion is a One-Time Trick

Here’s where Kahneman and Tversky come in. Their work on loss aversion showed that losses hurt roughly twice as much as equivalent gains feel good. That’s why you see people grinding for hours to protect a 10th-place spot rather than pushing for 1st. But here’s the subtle part: loss aversion only works when the loss feels imminent.

When the same person wins week after week, the perceived loss for everyone else shifts. It’s no longer "I might lose my spot." It becomes "I’m definitely losing to them." That’s a fixed, unavoidable loss. And the brain doesn’t agonise over unavoidable losses. It just disengages. That’s why your engagement curve doesn’t just flatten — it falls off a cliff.

I saw this play out with a client in Melbourne who ran a monthly photo contest for a local tourism site. The first month, 240 entries. The second month, 180. By the fourth month, they had 40 entries, and the same photographer had won three times. The prize was excellent — a weekend away. But the competition was dead. The photographer wasn’t just winning; they were winning in a way that made everyone else feel like they were donating their time.

The "Arbitrary Expert" Problem

There’s another layer here that’s specific to how we build these systems. Most leaderboards are based on a single metric — votes, points, or time spent. That’s fine in theory, but it creates what I call the "arbitrary expert" problem. The person who wins isn’t necessarily the best; they’re the one who figured out the loop.

Let me give you a concrete example from a SaaS community I helped build a few years back. We had a weekly "helpful answer" leaderboard. The same three power users dominated every week. They were genuinely helpful, but they’d also gamed the timing — answering within the first 30 minutes of a post going live, before anyone else had a chance. They were rewarded for speed, not quality. New users saw the board, assumed they’d never catch up, and stopped trying.

The fix wasn’t a bigger prize. The fix was changing the reward schedule.

Designing for Unpredictable Outcomes

So how do you build a leaderboard that doesn’t self-destruct? You need to borrow a page from competitive game design, specifically the concept of handicapping and dynamic difficulty adjustment. The goal is to keep the race close, not to crown a champion early.

Here are three practical approaches I’ve seen work in Australian small business and community contexts:

1. Rotate the Metric, Not Just the Prize

Don’t have one leaderboard. Have a weekly leaderboard that changes the scoring criteria. Week one is "most votes." Week two is "most creative entry." Week three is "most improved score from last week." This creates a variable schedule of what "winning" means. Skinner’s rats would approve. The repeat winner from last week might not even be competitive this week because the skill set is different.

2. Introduce a "Comeback Multiplier"

This is a classic mechanic from racing games. If you’re in the bottom 10% of the leaderboard, your points for the next task are multiplied by 1.5x or 2x. This makes the race feel winnable for everyone, not just the top three. It also creates a fascinating dynamic where the leader has to work harder to stay ahead, because their points don’t get the multiplier. You’re essentially using a variable-ratio schedule on the entire field, not just the winner.

3. Use "Decay" on Top Scores

This one’s a bit ruthless but incredibly effective. Points earned in week one are worth 100% in week two, 80% in week three, and 60% by week four. This means the leaderboard always reflects recent performance. The early leader can’t just coast. They have to keep playing to maintain their position. This is how professional sports ladders work — you don’t get to keep last season’s premiership points. It keeps the prediction error high because the standings are in constant flux.

The Friction of Near-Misses

There’s one more psychological lever worth pulling: the near-miss effect. Research on this is fascinating — it shows that when we almost win, our brains treat it as more rewarding than a clear loss. In fact, near-misses activate the same reward circuitry as actual wins, just at a lower intensity.

Your leaderboard can manufacture near-misses. If you have a public "top 10" list, make the 11th position visible and highlight how close they are to the cut-off. If you have a points system, show the exact gap between 10th and 11th place. This creates a micro-goal that’s achievable in the short term. "I’m only 15 points behind" is a far more powerful motivator than "I’m 1,000 points behind first place."

I had a client in Brisbane who ran a fitness challenge with a leaderboard. They added a simple line under the top 10: "You are 2,340 steps away from entering the top 10." Participation in the mid-tier segment jumped by 40% the following week. No new prizes. No new features. Just a clear, immediate, close target.

Building for the Long Game

The takeaway here is that your leaderboard isn’t a static object — it’s a dynamic system that needs to be tuned like any other piece of software. The moment you set it and forget it, you’re relying on the same person to keep winning, and you’re asking everyone else to accept a loss that feels permanent.

The next time you’re about to launch a leaderboard, ask yourself: what happens in week six? If the answer is "the same thing that happened in week one," you’ve already built the kill switch. Instead, think about how you can make the race interesting, not just the finish line.

Start by adding a comeback multiplier. Then rotate your scoring metric. Then add a decay rate. You don’t need to do all three at once — just pick one and watch how the behaviour shifts. The goal isn’t to crown a winner. It’s to keep the game alive long enough for people to care about the result. And that, more than any prize, is what keeps them coming back.