Three Line Break changes which price observations become visible rectangles. It does not remove elapsed time, open-position exposure or events between those rectangles. Calculate a few lines by hand before using the chart in a strategy: a new rectangle’s starting edge is not necessarily a price available after its confirming observation.

What Three Line Break is and where it came from

TradingView’s Number of lines setting specifies the previous confirmed lines used for the comparison. A closing price must move beyond their relevant range to add a line; equality or an insufficient move adds none. Three therefore describes the lookback in constructed lines, not three mandatory new opposite-direction candles. Check the implementation and distinguish a projection line from a confirmed one.

ChartSchool discusses both two-line reversals and a break after three same-direction lines, and points to Steve Nison’s Beyond Candlesticks for further study. A shared chart name does not ensure that platforms construct the same sequence. The examples below declare their own starting state and strict three-line-range convention; they are not screenshots or compatibility tests of any platform.

How the three-line rule and the reversal work

Our fictional model receives only completed source-period closes. Seed it at 100 with no line. An initial move draws a line from the seed; continuation starts at the last line’s ending edge. Compare each new Close with the full range of the last three constructed lines, or all available lines before three exist. On reversal, start the new rectangle at the near edge of the previous rectangle. State and drawing conventions are separate from trade execution.

For this up-line sequence, L1 runs 100–102, L2 102–105 and L3 105–108. Their range is 100–108. A Close of 106 stays inside; 100 touches the lower boundary; only a Close below 100 adds a down line. One qualifying new Close is enough. An intraperiod dip below 100 followed by Close 106 adds no line in this completed-close model.

Do not substitute the last three source closes for the last three drawn lines. Starting with L1–L3 above, feed 107,106,105 and then 104: none creates a line. The last three source closes before 104 have minimum 105, but the constructed-line threshold remains 100. A rolling three-close channel would falsely call 104 a break here. Even a thousand further 107 observations would retain the same three lines if none added a line. Preserve or reconstruct chart state from a declared seed.

Eight fictional completed closes, four lines
ObservationCloseState after observation
t0100Seed only
t1102L1 up, 100 to 102
t2105L2 up, 102 to 105
t3104No line
t4108L3 up, 105 to 108
t5106No line
t6100No line: equality
t799L4 down, 105 to 99

The periods t0–t7 are equally spaced, but L2 to L3 spans two periods and L3 to L4 three. Drawing L4 from 105 does not document a short entry at 105 after observing 99; those levels differ by 6 units. No Bid/Ask, orders or costs are supplied, so this table cannot produce a realized net trading result.

How to read and use this chart step by step

Step 1 — set the chart and the right timeframe

Record the source interval, feed, session, time zone, starting state and Number of lines setting. TradingView offers Line Break, but implementation and input detail still matter. Its non-standard-chart documentation says these charts use lower-timeframe data and approximate tick-based construction. The technical analysis foundations explain how transformed chart observations differ from conventional candle inputs.

Step 2 — read direction from the line sequence

Read the displayed direction as a property of the constructed sequence, not evidence that a healthy trend must continue. Keep source timestamps alongside line indices. Funding cutoffs, market closures and economic events operate on the real calendar even when the chart adds nothing.

Step 3 — wait for a confirmed reversal

Use only the line state actually confirmed at the decision time. A projection based on an unfinished source bar can change. In our example, the t7 Close confirms a down rectangle; it does not reveal the next market quote or guarantee a reversal. Decide separately how an alert becomes an order and when an order may execute.

A signal level is not an executable price

Choose the entry, invalidation, exit and position-size rules independently of the drawing convention. Do not fill an order at 105 merely because L4 starts there after Close 99. TradingView warns that non-standard-chart backtests can use synthetic chart prices for fills, producing results that do not represent actual market execution. A transformed series can supply a signal while a separate market-price series supplies an execution model, but their information timing must be respected.

A stop distance and funding exposure need the actual intended product, volume and tradable price series. The chart label supplies no mandatory stop, target, percentage budget or preferred timeframe. A close-only input can omit an intraperiod stop-triggering price even when no new line appears; filtering that observation from the chart does not remove its effect on a position.

Aggregation may remove a line absent from the final Close

Start with an empty chart and seed 100. Close 110 creates an upward line 100–110; Close 90 breaks its bottom and creates a downward line 100–90. Final 100 does not exceed the available range[90;110], so adds nothing. Feeding only final block Close 100 produces no line.

Full stream:two lines; final Close only:noneUpper data path 100,110,90,100. Two neighbouring lines show 100–110 and 100–90. The lower aggregated-data axis has only 100 and no new line. Both use seed 100.Source closes:100,110,90,100100110901002 linesOne final Close:100After aggregation:0 new lines
Right-hand rectangles are constructed lines, not equal-duration candles. Both streams end at 100.

Changing the source interval changes the chart input even when final price is unchanged. One Close 100 cannot recover omitted 110 and 90. When comparing charts, record source interval, seed and line count; the “3 Line Break” label alone does not ensure an identical stream or history.

Assumptions: entirely fictitious Close stream; line count 3 and shared seed 100. Four successive completed closes 100,110,90,100. The coarser version receives only final 100. No High/Low or assumed execution prices. Official material: TradingView:Line Break.

Strengths and weaknesses you need to know

Changing the source interval can change the chart even if its final Close is identical. As the separate diagram shows, the completed stream 100,110, 90,100 with seed 100 creates two lines: 100–110 up and 100–90 down. The final 100 lies within the available range 90–110 and adds none. Feeding only the final 100 creates no line. One aggregated Close cannot recover the omitted 110 and 90.

Keep a log of source input, prior line state, confirmation time, order submission, fill and costs. The number of source periods before a reversal is variable; three is not a fixed signal delay. Compare N = 2, 3 and 5 as distinct versions and evaluate an untouched period after selecting settings. No profitability comparison is presented here.

How it compares with Renko, Kagi and Heikin-Ashi

The transformations differ. Renko charts use a specified price-movement amount; Kagi charts use their own direction-change convention. Heikin-Ashi retains time-based bars with synthetic OHLC combining current and preceding values, rather than merely averaging Open and Close. Those constructions establish no universal ranking for signal frequency, conservatism or profitability. Compare fully specified inputs and rules.

What to do tomorrow to get comfortable with Three Line Break

  1. Copy the eight-close table and hide each next observation. Record the line state before deciding whether to add a rectangle.
  2. Keep source-period indices and constructed-line indices in separate columns, along with the seed and strict-boundary rule.
  3. Replace the last 99 with 108, then separately with 109. At 108 no line forms; at 109 an up line runs 108–109, and the next three-line lower threshold moves to 102.
  4. Compare the transformed chart with market prices and timestamps. Record available execution prices, rather than the starting edge of a new rectangle.
  5. Treat any changed line count or source interval as a new test version. Historical or demo exercises check understanding and implementation; twenty situations do not establish readiness or an edge.