01 / COMPARE THE MARKET
What held up
against Bitcoin?
Compare returns, the same BTC dip period, and reported volume. Follow the evidence before forming a hypothesis.
Real CoinGecko observations. Search by name or symbol, confirm the unique asset ID, and use transparent calculations and no automated AI conclusions.
SEARCH / SELECT / COMPARE
Choose your own asset.
Search CoinGecko by name or symbol. Confirm the name and unique ID before scanning; tickers can identify multiple coins.
Already know the exact CoinGecko ID?
Optional curated examples — compare a starter watchlist
KEEP THE COUNTERPOINTS
Measurements invite questions.
- Search finds covered assets; it does not scan the whole market.
- Daily observations miss intraday highs and drawdowns.
- Reported rolling 24h volume is not candle volume.
- One dip period cannot establish persistent resilience.
- These measurements do not establish causation.
BUILD YOUR OWN
Turn a question into a tool.
Search for a confirmed CoinGecko asset, then use the same time window for every asset, and documented calculations. Keep API requests and any credentials server-side.
CoinGecko API ↗Historical chart docs ↗Copyable build prompt
Build a CoinGecko watchlist comparison dashboard. Keep API calls server-side, search CoinGecko assets and fetch each selected asset plus BTC on demand, and cache results for 30 minutes. Compare 7/30-day USD returns and asset/BTC ratio changes, each asset’s return over the same BTC peak-to-trough dip dates, and last-seven/prior-seven median reported rolling-24h volume. Align UTC observations; show gaps, as-of/retrieval timestamps and sources. Include linked CoinGecko attribution and methodology. No synthetic fallback, predictions, raw-data exports, or autonomous AI. Keep credentials out of browser code.
Methodology, sources & data limits
Source: CoinGecko API (active access mode shown above), /coins/{id}/market_chart, USD, 35-day history. All requests originate on the server. This Vercel deployment uses Pro mode only when an authorized server credential is configured through Vercel environment settings. If public access fails or is rate limited, the interface reports the failure; it never substitutes invented values.
Cutoff: preceding UTC midnight. Nearest provider observation within 90 minutes of each midnight; no forward fill. The 7/30-day window requires 8/31 daily prices including both endpoints. As-of labels describe the nominal aligned cutoff, not exact exchange candle closing times.
USD return = last / first − 1. Relative strength = (asset last / first) / (BTC last / first) − 1. Dip: BTC’s most negative subsequent price / preceding running high − 1 in the selected window; the asset uses those same dates. If BTC has no sampled decline, dip results are unavailable. Volume: median latest seven reported rolling-24h observations / median prior seven; zero or incomplete baseline is unavailable.
Search: CoinGecko /search, names and symbols with unique IDs; up to 100 matches displayed, refine broad queries for more. Direct ID lookup is also supported. Search caches for 10 minutes, identity for 60 minutes, histories for 30 minutes. A scan uses at most three upstream calls, with shared BTC history. Explorer limits: 30 new upstream requests per minute and three active requests per server instance, 120 cached entries, no automatic retries. New coins or incomplete histories show unavailable metrics. Optional example-watchlist server cache: 30 minutes, concurrent-request deduplication per server instance, at most eleven upstream requests per refresh, two concurrent requests, no automatic retries; new requests stop after a rate-limit response. Failed refresh retains prior data explicitly as stale and introduces a two-minute retry cooldown per server instance. Cache and cooldown are operational controls, not an account-wide usage guarantee.
CoinGecko attribution is linked above. Raw-data exports are disabled pending applicable license verification. No trading signals, predictions, causal claims or chart image recognition. This dashboard is publicly accessible; provider credentials remain server-side.