Gold Digger · An Essay · Screwcap Games
The Free Leaderboard Paradox
Why a company that would very much like your money is giving away its single most popular feature — and why charging for it would have been the expensive mistake.
Freemium orthodoxy has one commandment: find the thing people love, put a wall around it, and sell them the key. It is tidy, it is intuitive, and — for a competitive scoreboard — it is wrong in a specific and instructive way. This essay makes the case for the inversion: a leaderboard that is free, public, and permanent, with the subscription reserved for the analytical depth that turns a score into an edge. We are, for the record, a company that would like your money. That we are arguing against charging for the most popular feature should tell you we found something more valuable than the obvious dollar. We did. The footnotes are load-bearing.
§1 The instinct to gate
Every team that ships a competitive feature meets the same temptation at roughly the same hour: keep the good stuff behind the paywall. The reasoning is clean, conventional, and almost always wrong in precisely the same place.
The case for gating a leaderboard is seductive because it is half-correct. A leaderboard is valuable — people care about it, return to it, screenshot it at dinner. Valuable things, the logic goes, should pull their weight on the balance sheet. Leave the scoreboard free and you have given away your highest-engagement asset for nothing; wall it off and you convert a cost centre into a turnstile, handing every visitor a tidy reason to reach for a card.
This is not a foolish argument. It is the argument. It is also the argument that, in early builds of this trainer, produced a measurable dip in people coming back and an unusual bloom of support emails whose tone travelled the full distance from puzzled to betrayed without pausing at annoyed.
The evidence that changed our minds was not a pristine experiment with confidence intervals you could frame. It was messier than that: a stretch where the scoreboard flickered between public and private depending on which build was live, and a pattern that survived the noise anyway. People who met the leaderboard even once — even for a moment — formed a different relationship to the product than people who never saw it. They came back more. They shared more. They described the game to friends in sentences that began with the scoreboard. Take it away and those behaviours did not gently taper. They fell off a cliff inside a week.
Which raised the question worth the whole essay: why would a feature whose value is so obvious on the spreadsheet feel, in the hand, like a small theft?
§2 Reactance, or: the surcharge on your own reflection
The answer has a name and a 1966 vintage. Brehm's theory of psychological reactance holds that when a freedom people expected to have is threatened or removed, they don't shrug — they mobilise to get it back. It is not merely a bad mood; it is a measurable redirection of behaviour. Tell users that a feature they assumed was theirs now lives behind a wall, and you will not receive a polite upgrade. You will receive resistance, counter-argument, and a quiet but firm reduction in their appetite for the very thing you withheld.
The effect intensifies exactly where you'd least want it to: in domains where people have parked some of their identity. Trading and prediction are nothing but parked identity. A player who has assembled a calibration record inside Gold Digger does not experience the leaderboard as a "feature." They experience it as a credential — a small, hard-won statement about how good their judgment actually is. Charging someone to see a credential they earned is not a pricing decision. It is a status threat with an invoice attached, and people do not respond to status threats by upgrading. They respond by leaving, and telling a friend why.
Fairness research turns the screw further. Fehr & Schmidt documented that people will burn their own money to punish what feels unfair, and Falk, Fehr & Fischbacher showed the urge to punish sharpens when the offender seems to be breaking a norm rather than merely chasing a profit. A returning player who finds the scoreboard newly locked reads it through that exact lens: the studio changed the deal. The thought is not "perhaps I'll subscribe." The thought is "I see how it is" — which is the most expensive sentence in consumer software, because it is always said on the way out.
We did, in the interest of rigour, gate the scoreboard on purpose for a spell — free for one group, subscription-only for another. The locked variant returned a sharply worse week-later comeback rate and a conspicuous rise in messages containing the words "locked" or "hidden." We stopped early. Partly the signal was loud; mostly the whole team had felt queasy about the wall before a single number arrived, and the numbers merely handed our consciences a citation.
A calibration score is not a feature you own and rent out. It is the player's property. You are the custodian. Custodians do not charge admission to the vault.
§3 The scoreboard pays its own rent
A public leaderboard is not charity with extra steps. It is an acquisition channel that bills itself to the network instead of to the player.
Shapiro & Varian wrote the economics of this down a quarter-century ago: for information goods, value rises with the size of the audience while the cost of one more onlooker rounds to zero. A scoreboard that is visible — not merely accessible once you've paid, but visible to a stranger who wandered in — turns every player's post-game breakdown into an advertisement the studio neither wrote nor paid for. It is the same machinery that makes a fantasy draft worth sharing and a poker hand history worth posting: a private performance number quietly becomes a public social signal.
And social signals convert. Cialdini and colleagues catalogued the obvious-once-stated truth that people look to the behaviour of others to decide what is reasonable, most of all when the territory is uncertain. Trading is uncertainty wearing a tie. A newcomer who lands on the trainer and sees a live board — real handles, recent scores, a community visibly mid-argument with the market — is not assessing a product brochure. They are watching a room that is already full. No headline we could write outsells the sight of other people having already shown up.
So the subscription is asked to solve a different problem entirely. It does not unlock the scoreboard, because the scoreboard was never locked. It unlocks the anatomy of the score: the instrument-by-instrument breakdown, the regime analysis, the exportable transcript, the calibration curve drawn across time. You are not buying the right to be seen. You are buying the lens that makes what everyone can already see mean something. That is not a rhetorical nicety; it is the entire shape of the value ladder.
Why "free hook, paid depth" beats "paid hook"
The freemium literature, when it isn't being quoted by people who skipped to the conclusion, agrees. The recurring finding is that the optimal free tier contains the engagement driver — the feature that brings people back most often — rather than the analytical feature that commands the highest willingness to pay. The mechanism is almost mechanical: gate the thing that creates the daily habit and you suffocate the very behaviour that would have produced the eventual upgrade. Keep the hook free, charge for the depth, and you maximise the population of people who form the habit that, in time, makes them want the deeper view. You cannot sell a telescope to someone you chased out of the observatory.
§4 So what, exactly, are we charging for?
A fair question, and one we'd rather you ask now than in a one-star review. If the scoreboard is free, the subscription has to earn its keep somewhere honest. It earns it in three places, each with a reason that isn't "because we can."
1 · The Laboratory — depth
Your calibration broken out by instrument, by regime, and over time: where your confidence was justified, where it was bravado, and under exactly which conditions you fool yourself. This is the machinery that turns a single number into a profile of how you actually think. People do not pay for this on day one. They pay for it the day they realise they have a track record worth examining — which is a day that only exists because the scoreboard was free long enough for them to build one.
2 · The Transcript — the receipts
An immutable, exportable log of every call: timestamp, instrument, direction, stated confidence, outcome, Brier score. It does double duty. For the serious player it is a coaching instrument; for anyone who wants to prove their judgment to a mentor, a desk, or a sceptical group chat, it is portable evidence. A scoreboard says you are good. A transcript says you can prove it, and proof — unlike bragging — is a thing people will reliably pay to be able to hand someone else.
3 · The Macro Anatomy — the explanation
The regime primer, the Taylor-rule overlay, the yield-curve dissection — the layer that explains why the market did what it did in the scenario you just survived. The free game is the experience; this is the textbook stapled to the back of it. You can learn to land the plane by feel for nothing. The flight manual costs money, and is, frankly, a bargain next to the alternative way of acquiring that knowledge, which is in public, with real money, at full tuition.
What we will not sell you, at any price
Speed. The subscription will never tighten the loop for paying players — no head start, no faster clock, no one-second edge. A trainer that sells pace in a product built to measure judgment would be quietly poisoning the instrument it rents out. Everyone runs the same loop at the same tempo, forever. We will sell you depth, context, and receipts. We will not sell you the one thing that would make the scores mean less.
§5 Three walls we built, and then demolished
Studios rarely publish their bad ideas. We think that's a missed opportunity, partly on principle and partly because the ideas we reversed are funnier — and more useful — than the ones we kept.
The first build split the historical regimes — 1929, the Oil Crisis, the S&L mess, Dot-com, the GFC, COVID — two free, four behind the subscription, on the high-minded theory that players should "earn" the harder eras. What players actually did was treat the free two as a demo and leave, with paid-unlock conversion sitting below a number we are not legally required to disclose but are emotionally required to call embarrassing. The reroute: every regime is playable, then anonymised for the calibration ladder. The subscription opens the analysis of history, not history itself. History, it turns out, does not take well to being held hostage.
An early version revealed which historical period you'd just traded — but on a delay: two hours for free players, instant for subscribers. The intent was a tidy little perk. The effect was a queue of players begging us to tell them sooner, which meant we had monetised suspense at the direct expense of the no-hindsight rule that makes the whole exercise honest. We deleted the timer. The reveal is now the same for everyone; the subscription buys a richer debrief, not an earlier one. Selling people their own answers back, on a clock, is a business model with a very short runway.
The experiment from §2 — free board versus paid board — was the cleanest failure of the three, and the one we argued about most. The internal debate was whether the backlash was a quirk of our particular audience. We decided it was the opposite of a quirk: finance-adjacent people are more sensitive to status cues, not less, so reactance hits them harder, not softer. The reroute was total. The board is fully public — no percentile gates, no friends-only mode, no velvet rope. If the score is earned, it is shareable. That sentence is now load-bearing.
Every wall had the same crack running through it: we were manufacturing scarcity in a domain whose value comes from completeness. A calibration record is not a consumable you ration. It is an asset that compounds. Gate a compounding asset and it feels like theft; gate the tools that make the asset legible and it feels like an upgrade. The line between the asset and the lens on the asset is the line we now refuse to cross — and it is, conveniently, the one line that also happens to be good business.
§6 We are not the first to notice this
The tug-of-war between free engagement and paid depth is old, and the products that resolved it well resolved it in the same direction. A short field guide.
Investopedia's simulator
Full leaderboard, full trade history, entirely free. The money lives in the education and the broker hand-offs. It works because the simulator is the friendly front door to a larger building. Gold Digger stands alone, so our economics differ — but the leaderboard-first logic survives the move.
Fantasy sports
Standings are public within your league and free to see. The social proof is a carrot, never a toll booth: you don't pay to learn your rank, you pay for the tools to improve it — optimisers, projections, the injury wire. That is almost exactly the calibration-lens model, wearing a jersey.
Fantasy golf
Public board; the subscription buys advanced stats, expert picks, and multi-entry play. The scoreboard is the marketing department and the analytics are the product. Of everything on this list, it is the closest structural cousin to what we're doing.
The finance-flavoured simulators
eToro tucks its social network behind a funded account; TradingView keeps ideas and leaderboards open and charges for deeper charting and screening. Between them they demonstrate the rule by obeying and breaking it in turn: the moment you gate the social proof in a money-adjacent product, you throttle the exact behaviour — deposits, retention, word of mouth — you were trying to grow.
§7 How you'll know if we're wrong
A theory you can't lose is a theory worth nothing. This one makes specific predictions, and we'd rather hand you the list than have you find the failures yourself.
| If we're right… | You should see | Because |
|---|---|---|
| People come back | A higher week-later return rate with the board public than without | Visible social proof reactivates |
| People share | A meaningful slice of active players posting breakdowns each week | The score is free marketing |
| People convert on depth, not on the wall | Free-to-paid rising when the paid thing is analysis, not the scoreboard | Habit first, ask second |
| Nobody rages about the board | "Locked"/"hidden" complaints near zero for the leaderboard | Reactance avoided where it matters |
| The board is sticky | Real dwell time on it for players a few scenarios in | A habit is forming |
| Depth sells itself | Players whose calibration is visibly improving reaching for the anatomy | The curve creates demand for the lens |
If the return rate doesn't move when the board goes public, the network-effect story is weaker than we think. If conversion doesn't rise when the subscription gates depth instead of visibility, the reactance story is incomplete. Either result would be unwelcome and, more importantly, true — which is the only kind of result worth designing around.
§8 The fine print we'd rather you read
This whole argument leans on an assumption: that Gold Digger's players behave like the humans in the reactance and choice-architecture studies. We think they do — finance-adjacent people are more status-sensitive, not less — but an assumption announced is healthier than one smuggled, so consider it announced.
There is a second risk we owe you plainly. A public board can curdle into a wall of specialists — a top twenty held by people who have ground a single regime for two hundred hours, quietly demoralising every newcomer who scrolls past. The fix is a design choice, not a pricing one: segment the board by experience — your first twenty calls, your next eighty, your hundred-and-up — so a beginner is measured against other beginners rather than against someone who has made the 1994 bond-market rout their personality.
And one commitment, non-negotiable. Every record on the public board is tied to a handle the player chose, never an identity we harvested; no real names, nothing identifying, without an explicit yes. The social proof works precisely because it is personal and freely given. The day it becomes extractive is the day it stops working — and, not coincidentally, the day it stops being something we'd be willing to sign.
§ References
- Brehm, J. W. (1966). A Theory of Psychological Reactance. Academic Press.
- Cialdini, R. B., et al. (2007). Social influence: Social norms, conformity, and compliance. In The Handbook of Social Psychology.
- Falk, A., Fehr, E., & Fischbacher, U. (2008). Testing theories of fairness — intentions matter. Games and Economic Behavior.
- Fehr, E., & Schmidt, K. M. (1999). A theory of fairness, competition, and cooperation. Quarterly Journal of Economics, 114(3), 817–868.
- Shapiro, C., & Varian, H. R. (1999). Information Rules: A Strategic Guide to the Network Economy. Harvard Business Review Press.
- Brier, G. W. (1950). Verification of forecasts expressed in terms of probability. Monthly Weather Review, 78(1), 1–3.
- Thaler, R. H., & Sunstein, C. R. (2008). Nudge: Improving Decisions About Health, Wealth, and Happiness. Penguin.
- Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus and Giroux.