# About Arcus Points Lab

Policy updated 2026-10-06. This date does not represent a protocol research refresh.

## Purpose and maintainer

Arcus Points Lab is maintained by [mdgazi786](https://github.com/mdgazi786), identified by the public GitHub account linked to this project. It is an independent research tool and is not affiliated with Arcus or any compared protocol.

The dashboard makes volume, trading costs and valuation assumptions inspectable. AI assisted with research compilation, writing and software development. Automated tests check calculation arithmetic and site behavior; they do not independently verify every source claim. No professional credentials or protocol endorsement are claimed.

## Research standards

Official announcements and documentation are preferred for program rules. Market-data snapshots and third-party OTC reports retain their source, observation date and limitations. Confirmed rules, modeled assumptions, announced allocations and realized payouts have different meanings.

A historical OTC report is not a live executable price or proof of settlement. Allocations and denominators must cover the same reward pool. Missing final denominators remain unavailable. The research-check date is separate from the source publication date and the exact volume measurement cutoff.

## Corrections and contact

Report a suspected error through [a GitHub issue](https://github.com/mdgazi786/arcus-points/issues/new). Include the affected page, claim, proposed correction, supporting source and relevant date. Do not include API keys, wallet secrets or private trading information.

Proposed corrections should be checked against primary evidence. Substantive accepted corrections should update the claim, source notes and dated revision history together. Unresolved claims should remain qualified. There is no guaranteed response deadline.

## Privacy and discovery measurement

Calculator assumptions and cached volume settings are saved in your browser. Using Share scenario puts the selected assumptions in a URL you choose to share. Vercel Web Analytics measures page views and available referrer information; calculator query parameters and fragments are removed before page-view events are sent.

Discovery reports remain private to the maintainer. Public-page access, search indexing, AI referrals and citations in answers are measured separately. Successful crawler-style test requests do not prove indexing, and an AI referral does not prove a citation. Private reports and credentials are not included in the site build.

## Revision history

### 2026-10-29: Volume snapshot, rules and protocol research rechecked

Updated the DefiLlama snapshot to $335.1M over 24h and $2.169B over 7d of perpetual volume (spot $1.91M / $12.13M, aggregator $2.70M / $16.22M); the default competing volume follows it. Rechecked the Arcus points rules: unchanged. Rechecked Variational (still no TGE date), Lighter (LIT at roughly $2.8-3.0B FDV, about $56-60 per Season 1-2 point) and Paradex (reported January 2026 change to a 25% XP airdrop; DIME market data too inconsistent to use). The valuation page now lists what is known: 1,000,000 Season 1 points issued over four weeks to about 18,000 accounts.

### 2026-10-29: Points estimator v0.7: fitted stock boost

The boost for stock, gold and index markets is now fitted together with the weights instead of being assumed: ×1.5 fits sixteen account-weeks best (typical miss 1.19×, worst 1.40×, against 1.28× and 1.60× without it); ×5 fits clearly worse. The separate after-hours boost is left at ×1 because the data cannot tell when the boost applies. The result card now shows how many points $1,000 kept at work for a week earns at the latest level. The small-market claim is marked as weakening.

### 2026-10-29: Week 4: predictions checked, estimator rescaled

Before the week-4 points were looked up, the v0.6 estimator predicted 0.39 and 15.3 points for two measured accounts (actual 0.43 and 18.09) and 0 for two idle ones (actual 0 for both). The model form is unchanged; it is now fitted on sixteen account-weeks (typical miss 1.3×, worst 1.6×) and scaled to the week paid 28 October, which paid about 26% more than the week before for the same activity. Position size now gets a small weight in the fit, so that claim is marked unclear.

### 2026-10-22: Points estimator v0.6: money at work

A fourth, much larger account (about $3,000 holding stock positions all week) showed that v0.5 underestimated large accounts by up to 24×. Refitted on four accounts over three weeks, points now follow money at work: your balance, counted only while a position of at least $10 is open, plus a small part for completed round trips. Idle money, position size, volume, fees and order type add nothing in our data; a small stock boost fits as well as none. The fit is looser than before (about 1.25× typically, 1.5× at worst) because it now covers very different account sizes.

### 2026-10-22: Points estimator v0.5: holding and orders, not money kept

Week 3 (distributed 21 October) showed that money left idle on Arcus earns almost nothing: two measured accounts that only kept money earned 0.01 and 0 points, as predicted before their cards, and an account that doubled its deposit but stopped holding earned less. The estimator now scores holding (the square root of position size × hours) and the number of orders, weighted 40/60, fitted on three weeks and scaled to the latest; money kept and volume get no weight, and market size no longer changes the estimate. The previous estimator overstated plans that mostly keep money on Arcus by several times. Claims were retested on three weeks.

### 2026-10-15: Points estimator v0.4: two measured weeks

Recalibrated the estimator on the second Season 1 distribution (14 October) for the same three measured accounts. Default weights are now 30% OI exposure, 70% retained collateral and 0% volume, fitted on both weeks; the scale is the latest week, which paid about 5× less per dollar kept on Arcus than the first. Claims about fees, market and limit orders, and the RWA boost were retested on both weeks. Fixed the estimate range, which could leave out your own weights under the alternative market-split assumption. Saved and shared plans keep your inputs and adopt the new assumptions.

### 2026-10-15: Arcus volume snapshot and rules rechecked

Updated perpetual volume to $462.32M over 24h and $2.923B over 7d, plus separate spot ($5.57M / $22.61M) and aggregator ($7.03M / $29.29M) figures. The default competing volume follows this snapshot. Rechecked the 250K weekly pool, Wednesday 3pm ET distribution, H1 2027 end and outside-hours boost window: unchanged. The 2.25 bps base taker fee and API rate limits are unchanged. Protocol benchmarks and historical OTC references retain their prior check dates.

### 2026-10-08: Points estimator and measured-account calibration

Replaced the volume-only home page with a beginner-friendly points estimator using live Arcus market data, served through an edge-cached markets endpoint. The estimator (v0.3) is calibrated to three accounts measured from Arcus’s public API using exposure, deposit-time and volume (weights 45/50/5), and circulating claims about fees, market orders and the RWA boost are tested against the same accounts. The earlier volume model remains under Advanced.

### 2026-10-07: Arcus volume snapshot refreshed

Updated perpetual volume to $497.44M over 24h and $2.862B over 7d, plus separate spot and aggregator figures. The default competing volume follows this snapshot. Rechecked the 250K weekly points rule; protocol benchmarks and historical OTC references retain their prior check dates.

### 2026-10-06: Authorship, corrections and discovery measurement

Added the About page, research policy and private discovery reports. This is a site update; the underlying protocol research and volume snapshot were not refreshed.

### 2026-10-06: Performance and accessibility

Self-hosted fonts, cached fingerprinted assets and improved text contrast. Formula and research inputs were unchanged.

[View change 8cc8d88](https://github.com/mdgazi786/arcus-points/commit/8cc8d88)

### 2026-10-06: Web Analytics

Added Vercel page-view analytics. Calculator query parameters and fragments are removed before events are sent.

[View change e556d51](https://github.com/mdgazi786/arcus-points/commit/e556d51)

### 2026-10-06: Initial dashboard and research

Published daily and weekly volume scenarios, six protocol comparisons, dated OTC references, formulas and static research pages.

[View change 1192ba9](https://github.com/mdgazi786/arcus-points/commit/1192ba9)

[Formulas and sources](/methodology)
[Project repository](https://github.com/mdgazi786/arcus-points)