Everything on this site is computed from third-party data. None of it is generated here, none of it is exclusive, and all of it is public. This page names every provider, says what is taken from each, and is honest about how each one fails.
Crypto prices and market data
CoinGecko supplies the entire crypto side: prices, market capitalisation, 24-hour volume, and the hourly price history behind every chart and every factor. The public API is used, optionally with a free Demo key. CoinGecko sends permissive CORS headers, so your browser may fetch it directly; where the caching proxy is installed the server fetches once and serves everyone from that copy.
Alternative.me provides the Crypto Fear & Greed Index shown in the summary tiles. It is displayed as context only and is not an input to any score. Note that it measures the whole crypto market; there is no Solana-specific version of it, so the Solana desk shows the same market-wide number.
Which tokens count as Solana
The Solana desk does not maintain its own list. Every request it makes carries CoinGecko's
solana-ecosystem category, so the board is exactly what
CoinGecko's
own Solana ecosystem listing contains at that moment — SOL itself, the DeFi and
infrastructure tokens built on the chain, the memecoins, and the stablecoins and bridged assets that
settle on it. Nothing is filtered in the page.
That means the membership question is CoinGecko's to answer, not this site's. Tokens are added and removed on their schedule, a token can be tagged for more than one chain, and a bridged asset appears under the ecosystem it was bridged into. If you think something is missing or wrongly included, the fix is on their side; tell us and we will pass it on.
The long tail here is genuinely risky. Below the first hundred names this board contains very small, very new tokens. Reported volume on some of them is unreliable, liquidity can vanish, and a technical score computed from a week of prices says nothing about whether a token is a going concern. Treat a high score on an unfamiliar Solana token as a prompt to investigate, never as a finding.
ETF prices
Stooq is the primary source for fund prices, delivered as daily OHLCV rows in CSV. When Stooq returns nothing for a ticker, the request falls through to Yahoo Finance’s public chart endpoint for the same symbol. A fund is only dropped from the board when both come back empty.
Fund prices are end-of-day closes, not live quotes. The board shows the last completed trading session and names it in the status line. Nothing on the ETF desk reflects what is happening in the market right now, and none of the prices are adjusted for dividends or capital-gains distributions.
The universe is a fixed list of roughly 124 US-listed funds, chosen for liquidity and for covering the categories people actually compare: US broad market, sector, style, size, bonds, commodities, international, emerging markets, dividend, and thematic. It is a curated list, not the whole market, and it is maintained by hand.
News
Headlines come from eighteen public RSS and Atom feeds, published by the outlets themselves. Each story shows the publication’s name, the headline, a short excerpt where the feed supplies one, and a link to the original article. Nothing is rehosted or reproduced in full, and no feed is modified beyond stripping markup and de-duplicating.
| Crypto | Markets and ETFs |
|---|---|
| CoinDesk | ETF Trends |
| Cointelegraph | Nasdaq — ETFs |
| Decrypt | Nasdaq — Markets |
| The Block | WSJ Markets |
| Blockworks | MarketWatch |
| CryptoSlate | CNBC Markets |
| Bitcoin Magazine | CNBC Investing |
| BeInCrypto | Yahoo Finance |
| Bitcoin.com News | Investing.com |
The two sides are merged round-robin by topic rather than by pure recency. Crypto outlets publish far more often than markets outlets, and a straight chronological merge buried the ETF stories entirely. Inclusion is not endorsement: these are widely read publications, they disagree with each other constantly, and several of them cover assets they have an interest in. Read them as sources, not as verdicts.
Publishers who want their feed removed or an attribution corrected should write in; such requests are actioned rather than argued with.
The short desk: your own model
The short desk is the one part of this site whose data comes from neither a market vendor nor a
news publisher. A model on a separate machine writes its predictions to a database there; a small
read-only service publishes one view of it over HTTPS behind a bearer token, and
short.php on this host fetches, caches and serves it. The database is never exposed
to the internet, and no database credentials exist on the web host.
Three things are read: the latest ranked predictions, the run metadata that defines what “collapse” means for that run, and the live calibration — every resolved prediction bucketed by what the model claimed against what happened. Per-name outcome rows stay on the database host; the site derives the track record by summing the buckets.
Because it is your model, its accuracy, its coverage and its update schedule are yours too. Nothing on this page validates it. What the page does is show you its record so you can judge it.
How long data is cached
Every upstream response is cached on disk so that a hundred visitors cost one request. The practical effect is that what you see can be slightly behind the provider, by at most:
| Data | Provider | Cached for |
|---|---|---|
| Crypto prices, global stats, Fear & Greed | CoinGecko, Alternative.me | 45 seconds |
| Solana ecosystem board | CoinGecko | 45 seconds |
| Crypto and Solana price history | CoinGecko | 15 minutes |
| ETF board | Stooq, Yahoo Finance | 15 minutes |
| ETF price history | Stooq, Yahoo Finance | 6 hours |
| Headlines | 18 RSS feeds | 10 minutes |
| Short desk board and calibration | your model's database host | 15 minutes |
| Short desk per-ticker history | your model's database host | 1 hour |
If a provider fails while a cached copy exists, the stale copy is served rather than an error, and the page says the data is stale. Showing you the last known good number is more useful than showing you nothing — but it does mean a long provider outage can leave old figures on screen.
When sources fail
They will, and the failures are visible on purpose:
- A few funds missing from the board. Normal attrition. ETFs close, merge, and get renamed, and providers drop the symbol as soon as it stops trading. The board loads what it can and names what it could not in the status line, rather than failing outright over one dead ticker.
- A feed not responding. Outlets change feed URLs, rate-limit by geography, or go down. The news header reports how many feeds answered; the rest of the stream is unaffected.
- Crypto rate limits. The free CoinGecko tier allows a limited number of calls per minute per IP. On a busy minute the cache absorbs it; if it cannot, the page falls back to the last good response.
- Provider disagreement. Crypto volume and market capitalisation are computed differently by different aggregators. A number here may not match your exchange, and neither is necessarily wrong.
Attribution and terms
All trademarks, publication names, ticker symbols, and fund names belong to their respective owners. VirtualSun is not affiliated with, endorsed by, or sponsored by CoinGecko, Alternative.me, Stooq, Yahoo, any news publisher, any fund issuer, or any exchange. Each provider’s data is used under its own public terms; if you intend to build on this data yourself, read those terms rather than relying on their appearance here.
How these inputs become a score is set out on the methodology page.