Multi-timeframe crypto screening

A coin can be oversold on the 15-minute chart and overbought on the 4-hour at the same instant. Both readings are correct.

Why the readings disagree

Every indicator has a lookback window measured in candles, not in time. RSI with a 14-period setting looks back 14 candles — which is three and a half hours on the 15-minute chart and over two days on the 4-hour. Those are two genuinely different questions about the same coin:

A coin in a strong multi-day rally that has just had a sharp three-hour pullback answers "yes" to both. There is no contradiction — it is a pullback within an uptrend, described from two distances.

The practical consequence

Screening on a single timeframe hides this entirely. You get a list of coins matching your condition with no indication of whether the larger structure agrees or disagrees, and those two cases usually call for opposite decisions.

The two classic uses of timeframe comparison:

Agreement as a stronger filter

Requiring the same condition on two candle sizes returns far fewer matches, and the ones that survive have both a short-term and a longer-term reading pointing the same way. Fewer, better-supported candidates.

Disagreement as the actual setup

Many traders deliberately want the disagreement: a higher timeframe establishing direction, and a lower timeframe showing a temporary move against it as the timing mechanism. Oversold on 15m within an established 4h uptrend is the textbook "buy the dip" structure. Screening on one timeframe cannot express this at all.

Choosing which two timeframes

A conventional rule of thumb is a ratio of roughly four to six between them — close enough to be related, far enough apart to say different things. 15m against 1h or 4h; 4h against 1d. Two adjacent candle sizes such as 4h and 8h will mostly repeat each other and add little.

StyleDirection fromTiming from
Position / swing1d4h
Multi-day4h1h
Intraday1h15m
Scalping15m3m / 5m

Note where the free line falls: the top row is entirely within altscreenerai's free timeframes, and the second row needs one paid candle size. Position and swing traders can work multi-timeframe without paying; intraday traders cannot, which is a reasonable place for the line to sit.

On altscreenerai

The Confluencer tab is built for this — it checks a condition across several candle sizes at once and shows where they agree and where they diverge, rather than making you run the same scan repeatedly and hold the results in your head.

Run this scan now → Free on 1d, 8h and 4h candles across 500+ pairs. No account, no card.