Help & glossary

What the numbers on this site mean, and exactly how they are computed. Every parameter below is the one the server actually uses — if the code changes, this page changes with it.

Reading a signal

Each tracked asset carries one current call, refreshed twice a day. A call is made of four parts:

The verdict — ▲ Buy, ▼ Sell or ■ Neutral, with « strong » variants when the score is extreme. The same three marks are used everywhere on the site : on charts, ▲ marks a switch to Buy, ▼ a switch to Sell, ■ a return to Neutral.
The technical score — a number from −100 (strong selling pressure) to +100 (strong buying pressure), computed server-side from price history alone. Deterministic : the same data always gives the same score.
The confidence level — 0 to 100 %. Below 50 % the badge is dimmed and dashed on purpose : a shaky call says so instead of dressing itself up as obvious.
The source — ia when an AI weighed the numbers and issued the verdict, technique when the AI was unavailable and the deterministic score alone took over. The site never hides which one you are reading.

The AI never computes anything. Indicators are calculated in PHP on the server; the model only interprets the resulting numbers — a language model is a poor calculator, and giving it exact figures removes that whole class of error.

To see all of this on a live example, open any asset and follow « Why this call? » : the current verdict is broken down into its exact score components, the model's own words, and the news it could see.

How the indicators are computed

The main analysis runs on hourly candles, the last 500 of them — a window of roughly three weeks, wide enough for the 200-period average to be a true exponential one. The same indicators are also computed on 4-hour and daily candles for the multi-horizon read, and a longer lens looks at 30 / 90 / 180-day drifts. All of it comes from public Binance price data, cached by scheduled jobs : opening a page never triggers a computation or a network call.

RSI — Relative Strength Index (14 periods, Wilder)

Measures how one-sided recent price moves have been, from 0 to 100. Gains and losses over the last 14 candles are averaged with Wilder's exponential smoothing — not a simple moving average, which is a common shortcut that gives different values. Above 70 the asset is conventionally « overbought », below 30 « oversold ».

MACD — Moving Average Convergence Divergence (12, 26, 9)

The difference between a fast (12-period) and a slow (26-period) exponential moving average, plus a signal line : a 9-period EMA of the MACD itself, rebuilt point by point over the whole series rather than approximated. The histogram — MACD minus its signal line — is what the score uses : positive means rising momentum, negative means fading momentum.

Bollinger bands (20 periods, 2 standard deviations)

A 20-period moving average, plus and minus twice the standard deviation of the same window. The score uses %B : where the price sits inside the bands, 0 at the lower band and 1 at the upper one. A price pinned to a band signals a stretched move.

Moving averages — SMA 20, SMA 50, EMA 200

Simple averages of the last 20 and 50 closes, and an exponential 200-period average — seeded with a simple average over its first window, then smoothed over the 300 candles that follow, which is what makes it a genuine EMA rather than a plain mean. The score reads whether the current price stands above or below the 50 and 200 lines — the classic shorthand for short-term and long-term trend.

ATR — Average True Range (14 periods, Wilder)

The average size of a candle's true range — how much the price actually travels, gaps included. It does not enter the directional score : it measures agitation, not direction. It is what detects pegged assets (see below), and what the volatility regime read is built on.

Pegged assets — the exception

A stablecoin priced within 2 % of 1 USDT whose average range is below 0.05 % of its price is treated as pegged. Its score is forced to zero and its signal to Neutral : the micro-wiggles around parity produce RSI values that look like signals and are not. Only a lasting break of the peg would be information — and that, the site does not predict.

That second threshold was 0.5 % until 23 August 2026, and it was too loose to separate anything. XRP, trading at 0.998 USDT through a quiet stretch, was classified as a stablecoin on 17 and 19 August : scored zero, called Neutral, three days before it rose 49 %. The new threshold sits in a measured gap — across 960 rolling windows the widest range USDC ever showed was 0.0152 % of its price, while the calmest real asset ever recorded was 0.0970 %.

From indicators to a score

Each indicator contributes a bounded number of points, so that no single one can dominate the verdict. The exact contributions:

Indicator	Condition	Points
RSI (14)	below 30 — oversold	+25
	above 70 — overbought	−25
	in between	(RSI − 50) × 0.2, at most ±4
MACD histogram	sign and size, normalised by price	capped at ±25
Bollinger %B	below 0.1 — pinned to lower band	+20
Bollinger %B	above 0.9 — pinned to upper band	−20
Price vs SMA 50	above / below	+15 / −15
Price vs EMA 200	above / below	+15 / −15

The sum is clamped to [−100, +100]. Note the deliberate mix : the Bollinger and RSI contributions are mean-reversion readings (a stretched move earns points against it), while the moving-average contributions are trend-following. The score is a balance of the two, not a bet on either.

The verdict then follows fixed thresholds — the same ones everywhere on the site:

Score	Signal
+50 or more	▲ Strong buy
+20 to +50	▲ Buy
−20 to +20	■ Neutral
−50 to −20	▼ Sell
−50 or less	▼ Strong sell

When the AI issues the verdict, it receives this score along with the raw indicator values, the multi-horizon read, order-flow measurements, qualified news and macro context — and may weigh them differently. When it is unavailable, the thresholds above are the verdict, and the signal is marked technique. In that fallback the confidence equals the absolute score (a score of +34 becomes 34 % confidence).

How calls are verified

Every call is later confronted with what the market actually did, at fixed horizons of 1, 3 and 7 days. Three methodological rules keep that measurement honest :

Comparison to the asset's own baseline drift. In a market that loses 38 %, any Sell call looks right. So the question is never « did the price fall? » but « did it do better or worse than this asset's usual drift over a window of the same length? » — the median of the rolling windows over its last ~120 daily closes. For a directional call, a gap below 0.5 points is recorded as not significant : noise, not signal. A Neutral call is judged by its own rule — right only when the market stayed within that margin — so its share of misses is not comparable with the directional hit rate, and the track record shows the two families apart.
Completed candles only. Since August 16, 2026 a call is judged against a finished daily close, never against the still-open candle of the day.
Criteria frozen in advance. Horizons and thresholds are written once in the code. Adjusting them after seeing the results would be self-validation.
Correlation is acknowledged — in the arithmetic. Major crypto assets move together : five simultaneous calls are not five independent observations. All directional calls of one day are therefore averaged into a single observation, and the track record reports the resulting daily mean with its z-score (|z| below 2 is noise). It still refuses to conclude anything below thirty distinct days.

Each outcome is written once and never edited — misses included. The track record page shows the running tally (an account is required, the demo account works).

Glossary

ATR — Average True Range: Average size of recent price moves, gaps included. Measures agitation, not direction. Computed over 14 periods with Wilder smoothing.
Base 100: A way to draw several assets on one chart : every curve starts at 100 on the first day of the window, so what you compare is percentage drift, not position size or price level.
Baseline drift: How much an asset usually moves over a window of a given length, measured on its own history. The reference every verified call is compared against.
Bollinger bands: A 20-period average ± twice the standard deviation of the same window. %B locates the price inside the bands (0 = lower, 1 = upper).
Buy-and-hold BTC: The return you would get by simply holding Bitcoin over the same window — no analysis, no model, no decisions. The honest benchmark for any strategy : beating inaction is not enough, since inaction is just one particular portfolio.
Concentration: The weight of your largest position in the portfolio's total value. A high concentration means the portfolio's fate tracks one asset.
Confidence: 0–100 %, attached to every call. From the AI when it issues the verdict; equal to the absolute technical score in the deterministic fallback. Below 50 % the badge is visibly dimmed.
Drift (30 / 90 / 180 days): The percentage change of an asset over the stated window, part of the long-term read given to the model.
EMA — Exponential Moving Average: A moving average that weighs recent prices more heavily. The 200-period EMA is the site's long-trend reference line.
Granularity: The time step of a chart : daily closes for long windows, intraday steps for short ones. Portfolio history switches automatically with the selected period.
Horizon: The fixed delay — 1, 3 or 7 days — after which a call is scored against the market.
MACD: Difference between a fast (12) and slow (26) exponential average, with a 9-period signal line. Its histogram reads momentum : rising when positive, fading when negative.
Measurement vs reconstruction: A measurement is a portfolio value recorded live and never changed. A reconstruction is recomputed after the fact with current quantities — useful, but it silently becomes wrong if positions change. The site stores which is which and never lets one overwrite the other.
Pegged asset: A stablecoin holding parity with the dollar. Detected by behaviour (price within 2 % of 1, average range under 0.05 %), scored zero, always Neutral.
Retrospective strategy (backtest): A proposal generated at a past date with price data strictly cut off at that date. Useful to test the machinery — but the model still knows what markets did afterwards, so it proves nothing and is always kept apart from real strategies.
RSI — Relative Strength Index: 0–100 measure of how one-sided recent moves were, over 14 periods with Wilder smoothing. Above 70 : overbought territory. Below 30 : oversold.
Signal / call: The current verdict on one asset : ▲ Buy, ▼ Sell or ■ Neutral (plus strong variants), with its score, confidence and source.
SMA — Simple Moving Average: The plain average of the last N closes. The site uses 20 and 50.
Source (ia / technique): Who issued the verdict : the AI reading all the evidence (ia), or the deterministic score alone when the AI is unavailable (technique). Always displayed.
Sparkline: The small 30-day price curve next to each asset — blue when the window ends higher than it started, red otherwise, like every chart here.
Technical score: The −100…+100 number built from bounded indicator contributions (see above). Deterministic and computed server-side.
Tone (news): Each article is scored from −10 to +10 per asset by a lightweight model before it reaches the analysis — « Bitcoin plunges, Ethereum holds » would be wrong with a single number. It is shown everywhere as a small pill next to the article : −6 reads « clearly negative for this asset », 0 is a genuine neutral, and not assessed means the article was never qualified — an unknown is never displayed as a misleading zero. Promotional pieces are dropped entirely.
« Why » pages: Every asset links to Why this call : the current verdict broken down into its real ingredients — the score's exact components, what the model wrote, the news it could see. Every strategy links to Why this proposal, same idea for a reallocation. Neither shows decision percentages : the model does not expose a weighting, and inventing one would be a lie.
Weighted score: The average of your assets' technical scores, weighted by each position's share of the portfolio value. A marginal line barely moves it; your dominant line mostly decides it.

What this measures — and what it cannot

Everything above transforms past prices. None of it predicts : a reversal is observed, then absorbed — never anticipated. Central-bank announcements, regulatory shifts and geopolitical shocks are decided elsewhere and reach no candle, no order book and no funding rate before they happen. That is a limit of the field, stated here on purpose. The disclaimer at the bottom of every page applies to this one too.