Chart Patterns
Symmetrical Triangle Intraday Breakout
Two converging trendlines squeeze price toward an apex, and the shape itself never says which way it will break.
Fit both lines, then call it blind
Drag the two gold handles onto the real swing highs to set the upper line, and the two blue handles onto the real swing lows to set the lower line. The tool checks both slopes and how much of the real candle range sits between them, live, as you drag. Once the fit is genuinely valid, call Breakout Up or Breakout Down before anything resolves, then watch price play forward from the apex.
Gold handles set the upper line · Blue handles set the lower line · drag any of the four freely
Upper line: pending
Lower line: pending
Apex: pending
Containment: pending
| Call | Calls | Correct | Accuracy |
|---|---|---|---|
| Breakout Up | 0 | 0 | — |
| Breakout Down | 0 | 0 | — |
How it works
- Drag the gold handles onto two real swing highs to set the upper boundary, and the blue handles onto two real swing lows to set the lower one. Both lines are genuinely diagonal, not locked flat.
- The tool checks both slopes live. The upper line has to point down and the lower line has to point up, and it scores how much of the real candle range actually sits between the two lines across the window.
- Once the fit clears the containment bar and the two lines converge within a reasonable number of bars, the setup arms. That is when it asks for a direction, before anything has played out yet.
- Price then runs forward from the apex on its own generated path and reveals which way it actually broke. Your call gets scored against that real outcome.
Where this breaks
The pattern predicts a move, not a direction
A symmetrical triangle squeezing two real trendlines together is a genuine signal that the range is about to expand. It says nothing about which way. The scoreboard on this page is built to prove that: it should sit close to 50/50 no matter how many valid setups get called, because the coin deciding the actual breakout direction here is fair by construction. If it ever drifts meaningfully away from 50/50 over a large sample, that is sampling noise, not a discovered edge, since the shape of a real triangle carries no directional information on its own.