Every other trading terminal took the other side. This one took yours.
The design choices that answer the Gambling Parlor Dossier, one primitive at a time.
A companion to “97% Lose Money. The Interface Is Why.”
There is an operator running a 3,000-member Telegram group. He has been doing it for four years. He calls trades. His audience follows. They trade on a major centralized exchange that pays him on a referral schedule that resets every 90 days.
He has watched two cohorts of his best referrals — the highest-volume traders he ever onboarded — hit VIP-3 and convert silently to $0 in his dashboard. He has watched 41% slide to 28% slide to 17% across three quarterly reviews. He has read the exchange’s terms that prohibit him from re-referring his own audience. Twice he has written a support ticket asking why a commission halved between Monday and Tuesday. The first was closed without response. The second got one line back that did not name the change.
That is the surface story. It is not the real one.
The real story is the cohort he stopped hearing from. The traders who hit VIP-3 and then went quiet — not because they were upgraded out of his attribution, but because they were liquidated. The members of his Telegram group who slowly stopped messaging. The handles he can scroll back through and find, three months later, no longer posting.
The dossier names the population they belong to. 97% of persistent Brazilian day traders lose money. EU retail CFD accounts lose at rates of 74% to 89% across jurisdictions. The Hyperdash/BeInCrypto sample of 1,000 Hyperliquid traders: 13.5% profitable. The interface is the mechanism. The pattern is on the record from Munger, Buffett, ESMA, FINRA, the SEC, the Chicago Fed, and thirty years of peer-reviewed research.
The bridge he built carried his audience into that population. The terms he was reading — the 90-day reset, the VIP-3 cutoff, the re-referral prohibition — were the mechanical expression of the same architectural property that produced the 97%. Volume is the metric the platform monetizes. Survival is the metric his audience needed. The interface is the surface on which the misalignment is expressed.
This document is for him. It is about the audience.
What this document is, and is not
The dossier closes on an architectural question: what would an interface look like if it were designed for trader survival rather than volume? This document answers that question — concretely, primitive by primitive — for the Wickery terminal.
It is not a claim that the design produces a specific profitable-trader rate. No one can promise a number like that. The 13.5% Hyperliquid baseline is what the volume-aligned interface produces; what the survival-aligned interface produces is a question to be settled empirically over time. The architectural claim is narrower and more falsifiable: each of the following design choices removes a specific, documented mechanism by which the standard interface converts trader cognitive bias into platform revenue.
The dossier names four properties a risk-first interface must deliver: clarity, control, loss reduction, and aggregate-risk visibility. The dossier also names six UX primitives through which the volume-aligned interface expresses the rentability pattern. Below is how each one is answered in the terminal that ships at partner.wickery.xyz.
1. The fee architecture: receipts nobody can edit
The economic precondition has to be true before the interface can be honest. A terminal designed for trader survival is only viable when the operator running it earns more from surviving traders than from doomed ones. The fee architecture makes that condition true.
When a partner deploys Wickery, they become a Hyperliquid builder in their own right. The branded terminal lives at partner.wickery.xyz. Builder fees route to a joint multi-sig wallet, controlled via Hyperliquid’s native multi-sig with the partner’s signature required for every withdrawal. Wickery cannot access the partner’s wallet. Neither party can unilaterally alter the fee record, the revenue split, or the attribution.
Every fee event is a public, permanent, cryptographically-signed transaction on Hyperliquid’s ledger. Every unit is verifiable by transaction hash.
The fee record cannot be tier-reset. It cannot be VIP-3-cutoff. It cannot be silently re-attributed to “came from ads.” It cannot be clawed back. Nobody can edit a blockchain transaction after the fact.
The inversion is structural, not cosmetic.
A CEX affiliate’s earned commissions exist because the exchange’s backend says they do. A Wickery partner’s earned fees exist because the chain says they do. The first is a row in a private database. The second is a row in a public one. The exchange can change its row.
Nobody can change Wickery’s.
That property is the precondition for everything that follows. With the receipt immune to revision, the partner’s incentive is freed from the platform’s quarterly review and reattached to the only variable that still matters: how long the audience continues to trade. The interface is what determines that.
2. The four deliverables
The dossier names four properties a risk-first interface must deliver. Three of them appear in scattered form across the retail trading universe. The fourth is absent everywhere.
2.1 Clarity: every entry is its own trade
The trader who entered at $100, added at $90, and again at $80 sees three trades, three risk profiles, three exits. Not a single blended position with a friendly averaged cost basis of $90 that hides the cost of the two later decisions.
The partner-facing copy states the rule:
“Every market entry is a separate trade. You cannot ‘blend’ your mistakes — you learn from every decision.”
That is the honesty-of-data argument in one sentence. A trader cannot review what they cannot see. A trader cannot learn from what they cannot review. The blended-position convention is the interface mechanism of the unreviewable trade. The per-entry convention is the interface mechanism of the reviewable one.
Thirty years of retail-trading literature documents what happens when traders cannot see their own bad entries. Chague–De-Losso–Giovannetti’s finding of “no evidence of learning by day trading” among Brazilian futures traders — the population in which 97% lose money — is partly a finding about the interfaces those traders used.
You cannot learn from a blend.
A trader who can see their own bad entries is a trader who can become a better trader over time. A trader who sees only a softened average is a trader for whom each month’s mistakes are invisible by the next month’s review. The clarity primitive is the interface mechanism by which the audience’s review of its own performance becomes honest enough to compound into improvement.
2.2 Control: no “average down” button
The standard interface offers “average down” as a one-tap primitive. By silently merging fills into a single re-computed cost basis, it removes cognitive friction at the exact moment the disposition effect would otherwise be acted on — the moment a trader is adding to a losing position.
Shefrin–Statman (1985) named the bias. Odean (1998) measured it. Traders sell winners at a rate roughly 50% higher than they sell losers, costing the average account 3–5% per year in foregone returns. The standard retail interface operationalizes the bias at the UI layer.
The Wickery terminal does not offer “average down” as a primitive. There is no button. There is no flow. There is no silent merge.
Each new entry is a new decision. A trader who believes the original thesis still holds and wants to add at a worse price takes a new position with its own entry, sizing, and stop. The platform does not silently merge the two. The platform does not hide which entry was a mistake.
The cognitive load of the second decision is preserved because the interface refuses to absorb it into the first.
The disposition effect is a property of the trader, not the platform. The Wickery terminal does not eliminate it. It stops actively converting it into platform revenue.
2.3 Loss reduction: isolated entries, not just isolated positions
Clarity and control compound. When each entry carries its own stop-loss and its own P&L, a single bad trade cannot pull the others down. The terminal’s design rule —
“Isolated margin: every position is an island. One bad trade will never liquidate your entire account”
— extends, in the per-entry model, to the per-entry level. A losing entry closes against its own stop. The winning entries continue against theirs.
Isolated margin protects one position from another. Isolated entries protect one decision from another. The gap between the two is where most retail traders lose most of their money: a winning thesis stopped out at the price of a subsequent losing entry’s contamination.
The contamination is the mechanism. The isolation is the prevention. The trader who is not contaminated is the trader who is still in the audience next month.
2.4 The fourth deliverable: total current risk
Now the metric the dossier names and no major retail terminal surfaces.
Every trader on every retail perpetuals terminal in the world right now is carrying a number they cannot see. The number is the answer to a single question.
If every open position goes against me and every stop and liquidation triggers, what is the maximum dollar amount I lose from this exact moment forward?
The dossier surveys the platforms. Binance. OKX. Bybit. Bitget. Robinhood. Plus500. IG. eToro. CMC Markets. Hyperliquid’s own native interface. MetaMask Perps. Phantom. pvp.trade. Hyperdash. HyperTracker. ASXN. Dexly Explorer.
None of them surfaces the number.
They surface unrealized P&L. They surface margin ratio. They surface account equity. They surface liquidation prices per position. They do not surface the one number a rational trader is actually trying to answer about their own portfolio.
The Wickery terminal surfaces it.
Total current risk is a single number, visible at all times, that aggregates the worst-case dollar outcome across every open position. Long BTC with a stop at -$400. Long ETH with a stop at -$200. Short SOL with a stop at -$300. Total current risk reads -$900. If the SOL position is on 25× leverage with a smaller margin allocation and the realistic liquidation outcome is -$700 before the stop fires, the terminal reads the larger of the two: -$1,300.
The number moves as the trader opens positions, adjusts stops, and adds collateral. The trader does not compute it. The terminal computes it.
The bias the dossier identifies — narrow framing (Kahneman and Lovallo, Management Science 1993) — is the property of the trader that the absence of this metric exploits. Traders evaluate risks one at a time. They systematically misjudge the portfolio’s actual risk profile as a consequence. They get stopped out at numbers larger than they thought were possible.
The under-estimation is the gambling parlor’s quietest revenue line. It is also the moment when the trader who could have survived stops being a member of the audience.
The interface that makes the worst case visible at all times is the interface that lets the trader make the cut before the cut is made for them.
This is clarity completed. The per-entry P&L is the most granular view. Total current risk is the most aggregate view. Between them, every dollar at risk is visible. There is no hidden position. There is no buried exposure. There is no “I didn’t realize” between the trader and their portfolio.
The dossier names this deliverable. The Wickery terminal is where it ships.
3. The six primitives, answered
3.1 Celebration of volume → no celebration
Robinhood rendered an animated burst of confetti each time a customer placed a trade. The Massachusetts Securities Division named the practice in its December 2020 complaint as “celebratory imagery tied to the frequency of trading.” Robinhood removed the confetti in April 2021. The settlement extracting that removal cost $7.5 million.
Chapkovski–Khapko–Zoican measured the cost of confetti at the population level. Hedonic gamification increases trading volume by 5.17%. The effect is strongest on participants with lower financial literacy.
The Wickery terminal uses no confetti. No achievement badges. No scratch tickets. No lottery-style stock rewards. The closing of a profitable trade is visually neutral: the position updates, the P&L line records, the event is over.
Five point one seven percent does not sound like much. Compounded across an 18-month trading horizon against a population already losing 13.5% of its members to the perps baseline, it is the difference between a trader who survives to month 12 and a trader who does not. It is the difference between the cohort that keeps showing up in the Telegram group and the cohort that goes quiet.
3.2 Attention engineering → no attention engineering
Barber, Huang, Odean, and Schwarz published the cost of attention engineering in the Journal of Finance in 2022. The number: -4.7% average 20-day abnormal returns on the top stocks purchased each day by Robinhood users. The mechanism: “Top Movers” lists, “100 Most Popular” lists, push notifications hyping short-term volatility. The U.K. Financial Conduct Authority’s Occasional Paper 66 measured an 11% trading-volume increase from push notifications alone.
The Wickery terminal has no leaderboards. No “top movers.” No “biggest gainers.” No “trending tokens.” No copy-trading surface. No ads. No pop-ups.
“Distraction-free terminal: no ads, leaderboards, or pop-ups. Just you and the market.”
Notifications are limited to position-risk events. Margin call. Stop-loss approach. Liquidation proximity. They are not price-attention bait.
The 11% volume increase from push notifications exists because, in their standard retail form, push notifications are the volume-aligned interface’s way of pulling the trader back to the chart. Wickery’s notifications pull the trader toward the position that needs attention, not the price that wants attention.
The audience is not pulled into a trade they were not planning to make. They are pulled toward the trade they have already made, at the moment it requires their decision.
3.3 Friction removal → friction restored at the right moments
The Massachusetts complaint named one customer with no investment experience who placed more than 12,700 trades in six months. FINRA fined Robinhood $57 million plus $12.6 million in restitution for approving customers for options trading “based on inconsistent or illogical information.”
Alex Kearns was 20. He was a college student. On a June night in 2020 he opened the Robinhood app and saw a negative $730,000 balance that was a display artifact of an unsettled options spread. He emailed customer service three times. He received no live response.
He took his life that night.
The friction was removed everywhere it pulled the audience deeper. It was installed everywhere the audience needed to ask a question.
The Wickery terminal answers the primitive at the same layer the dossier names it. Stop-loss is a first-class action at order entry, executable in 2 clicks. It is not buried inside an order-modification submenu. The order-entry surface treats risk management as part of the order, not a follow-up to it.
“Risk management is the first step, not an afterthought.”
Fill-blending is refused. Adding to a losing position requires a new trade — sized fresh, stopped fresh, monitored fresh. The “average down” primitive does not exist because it does not exist.
Friction is not the absence of design. Friction in the right places is the design.
3.4 The leverage pyramid → conservative per-entry defaults
Before ESMA intervened in March 2018, retail CFD platforms in the EU offered leverage ratios up to 500:1 on FX pairs. ESMA capped major FX at 30:1 and cryptocurrencies at 2:1. The FCA made the caps permanent in PS19/18 and estimated U.K. retail consumer savings of £267M–£451M per year as a direct consequence.
The offshore crypto perpetuals exchanges that ESMA cannot reach default leverage sliders to 100× and 125×. At 100× leverage, a 1% adverse move liquidates the position.
The Wickery terminal includes position-sizing guidance at the order-entry layer, calculating optimal size from the trader’s capital and risk tolerance. Isolated margin is the default position-layer convention, extended in the “every entry is a separate trade” model so that per-decision liquidation is also per-decision.
Leverage is still the trader’s. The architectural decision is not to lock the trader to a conservative number but to make the per-entry container honest about what that number does. A trader who chooses 50× on a single entry has chosen 50× on that entry — not 50× on the blended position that includes a winning trade.
The risk lives where the decision lived.
A losing entry can be liquidated against its own collateral without exposing the winning entry to the same liquidation event. This is the inversion of the most expensive UX pattern in retail trading: a winning thesis contaminated by a losing add-on. A correctly identified trade stopped out at the price of a subsequent mistake.
The dossier puts a thirty-year academic literature behind the cost of that contamination. The Wickery terminal prevents it structurally — not with a warning, not with a confirmation dialog, but at the position layer.
3.5 Loss-aversion exploitation → per-entry honesty
Fill-blending is the UI mechanism of the disposition effect. The dossier’s §2.5.3 walks through it. A trader holds a losing position and adds at a worse price. The standard interface re-computes a single averaged cost basis. The bad entry vanishes into the friendlier blended average.
The losing trade is not closed. The interface has cosmetically improved its appearance.
The Wickery terminal does not present the friendlier blend. Each entry’s P&L is visible at the per-entry level. A bad entry is visible as a bad entry. The terminal does not surface averaging-down nudges. There is no recommended action on a losing position that says “average down at a better price.” The recommended actions are the actions that close it or stop-loss it.
Not the action that absorbs it.
A trader who can see the original bad entry is more likely to cut it than a trader who sees only a softened average. Odean’s 3–5% per year disposition-effect drag is the annuity the standard interface charges for the cosmetic improvement.
The Wickery terminal does not charge that annuity.
3.6 Opacity → on-chain verifiability
The SEC’s December 17, 2020 finding: Robinhood received “unusually high” payment-for-order-flow rates and failed to adequately disclose this to customers. Settlement: $65 million. Plus500’s class-action allegations of asymmetric pause mechanics — service paused selectively to prevent customers from selling profitable options while losing positions remained open. The Chicago Fed’s Letter No. 479: approximately $13 billion in customer outflows from Celsius, Voyager, BlockFi, Genesis, and FTX in the run-up to bankruptcy.
The structural property in every case is the same. Retail customers did not have visibility into the flows the platform was operating.
The Wickery architecture is the inversion at every layer the dossier names.
There is no payment for order flow. There is no opaque routing economy between trader and venue. Trades settle on Hyperliquid’s on-chain order book at 0.07-second block finality. Every order, cancellation, fill, and liquidation is a public on-chain event.
There is no custodial deposit. Users connect their own Web3 wallets and maintain full self-custody throughout. Funds never leave the user’s wallet to enter a platform-controlled commingled pool. The single liability vector that drained $13 billion from CEX customers in 2022 — platform custody followed by withdrawal suspension — is structurally absent.
The trader does not have to trust the operator’s good intentions. The trader does not have to trust Wickery’s good intentions. The settlement layer is the protocol.
4. The design choices, in one place
The terminal can be summarized as the decisions a designer made, one at a time, against the volume-aligned default. Not as outcomes. As decisions:
- Stop-loss is at the order-entry surface, in 2 clicks, before the position opens. The standard interface puts it inside an order-modification submenu.
- Each market entry is rendered as its own position object. The standard interface blends fills into a single averaged cost basis.
- The “average down” primitive is absent. Adding to a loser requires opening a new trade with its own sizing and stop.
- Total current risk — the dollar-denominated worst case across every open position — is a single number visible at all times on the trading surface.
- Position-sizing guidance is rendered at the order-entry layer, computing an optimal size from the trader’s capital and risk tolerance.
- Isolated margin is the default at the position layer, extended to the per-entry layer.
- Notifications are restricted to position-risk events. There are no price-attention notifications.
- There are no leaderboards, “top movers,” trending lists, or copy-trading surfaces.
- There are no confetti animations, achievement badges, lottery-style rewards, or other hedonic markers tied to trade execution.
- Trades settle on Hyperliquid’s on-chain order book at 0.07-second block finality. There is no payment for order flow.
- Funds remain in the user’s self-custody wallet throughout. There is no custodial deposit.
- Builder fees are recorded as on-chain transactions, routed to a joint multi-sig wallet, and verifiable by transaction hash.
Each one removes a documented mechanism by which the standard interface either extracts revenue from trader cognitive bias or obscures the cost of doing so. None of them depends on the others. All of them ship together.
5. Closing
The dossier closes on a question. Whether the people building the next interface read the record before they default the leverage slider to 100×.
This document closes on the answer.
The stop-loss is the first action. Every entry is its own trade. The total current risk is visible at all times. Isolated margin is the default. There is no “average down” button, no confetti, no leaderboards, no price-attention push notifications. Settlement is on-chain. Funds are self-custody. The fee record is signed by the chain.
These are not promises about outcomes. They are decisions about the interface. The decisions are made against the rentability pattern, one primitive at a time, at the layer the trader actually sees.
For the operator from the opening, the question is no longer whether his next bridge will be revoked. The question is whether the audience that crosses it gets to keep trading at month 12. The answer is a different terminal — not a different exchange, not a different fee schedule, not a different referral program. A different interface, with different defaults, in a different place on the screen.
There is no quarterly review. There is no VIP-3 cutoff. There is no confetti.
The other side of the trade was always the platform’s. The dossier names every primitive that holds it there.
This terminal moves it.
References: Gambling Parlor Dossier §1.4, §2.1–§2.6, §6, §7. Wickery technical documentation: wickery-exec-summary-bizplan-v6.md, Your_Audience__Your_Exchange__Your_Revenue_.md, wickery-brand-narrative-v2.md.