Does a third touch of resistance confirm a triple top?

No. A third touch still needs to become a qualifying swing under the chosen rules. The candidate then needs its support-break condition. A continued rise on that visit may prevent the candidate from forming at all.

Which low should define support when the two pullbacks differ?

This guide uses the lower low as horizontal support. With pullbacks to 94 and 95, the threshold is 94. A sloping line through the lows is a different detector and must be specified and tested separately.

Is a triple top stronger than a double top?

The extra test does not establish better performance. Requiring a third swing changes which setups qualify and when they can be recognized. A comparison needs fixed detection rules, equivalent trade rules and costs on the same evaluation sample.

Must price return to support after breaking it?

No. A retest is an optional entry filter. Some confirmed breaks move away without returning; others reclaim support. Define the retest window and acceptable price range before looking at the subsequent path.

Can Backtest Report count all three-top candidates?

A trade CSV cannot represent candidates that produced no trade. Log detector states separately, then use Backtest Report to analyze exported trades from each entry rule. State counts and trading performance should remain separate measurements.

TradingView chart patterns

Triple top pattern: support break and failure cases

A triple top pattern has three similar highs separated by two pullbacks, usually studied as a possible reversal after a rise. Three tests of resistance alone leave a candidate. TradingView confirms its built-in pattern after a close below the lower of the two intervening lows. Price can instead break upward on the third test, or reclaim support after a breakdown.

See the decision guide

Mark the highs and both support levels

Read a triple top chart pattern in order: H1, low L1, H2, low L2, H3. The highs form the resistance area. Under the horizontal neckline convention used here, support N is min(L1, L2). If L1 is 94 and L2 is 95, a close at 94.6 has broken the second low but has not broken N. Do not move N upward to label that earlier move a confirmed breakdown.

Three neighboring candles touching a similar price do not establish three separate swings. Require intervening declines and minimum spacing.
If the central high stands well above both outer highs, compare the structure with head and shoulders rather than expanding the tolerance until everything fits.
A third visit can still become an upward breakout. The number of tests alone provides no measured probability of reversal.

An explicit detector specification

The following rules are illustrative design choices for standard price candles. They are not TradingView defaults or recommended trading settings. Track the latest three confirmed swing highs and reject the set if any geometry rule fails. A fixed price-unit tolerance needs adjustment and fresh testing when the instrument changes.

Create one candidate when the third pivot becomes known. Freeze its highs, lows, N and U; do not replace them while that candidate is active.
At creation, reject a set if any completed close after H3 already exceeded U. While active, invalidate it on a close above U; expire it after bar 56 if no confirmation occurs.
A support crossing before the third pivot becomes known is not backdated into a signal. This specification requires a fresh eligible crossing, including on the pivot confirmation bar.
Consume the candidate after its first breakdown. Require a newly confirmed high before forming another set so the same structure is not counted repeatedly.

Decision	Illustrative rule	Worked values
Context before H1	H1 must be at least 5% above the lowest low of the 20 preceding bars; require all 20 bars to exist	A measurable prior-rise filter, not proof of a durable uptrend
Swing selection	A high must exceed the highs of the three completed bars on each side; reject ties	A high at bar 34 is first known at the close of bar 37
High tolerance	max(H1, H2, H3) - min(H1, H2, H3) ≤ 0.5 price units	100.0, 100.4, 100.2: spread 0.4 passes
Spacing and pullbacks	Each high-to-high interval is 8 to 30 bars. Each intervening lowest low is at least 3 units below both adjacent highs	Highs at bars 10, 22 and 34; intervening lows 94 and 95: both tests pass
Frozen support	Set N to the lower intervening low and U to the highest selected high plus 0.5	N = 94; U = 100.9
Confirmation window	From third-pivot confirmation through the next 19 bars, require previous close ≥ N and current completed close < N	Eligible bars 37 through 56; closes 94.4 then 93.6 create a crossing

Compare the upward breakout with the breakdown retest

These hypothetical paths use the example levels above. They are schematic event sequences, not historical market data. Path A branches during the third resistance test; the other paths assume the three-high candidate has become eligible at bar 37.

If an already eligible candidate closes above U before a breakdown, record it as invalidated. That differs from Path A, where the third high never passes the geometry rules.
For an optional retest entry, allow the next 10 bars after confirmation. Require a high between 93.8 and 94.2 inclusive, a close below 94, and no earlier close above 94.
Under this custom failure rule, the first post-confirmation close above N ends the setup. TradingView uses a different Failed status: price exceeds the last top before reaching its target. Do not combine those definitions in one statistic.
A decline that never retests creates no retest trade. Retest entries therefore test a different sample from entries taken directly after the breakdown.

Illustrative paths. Arrows show event order, not equal time intervals or a price forecast.text

First two tests: H1 100.0 -> L1 94 -> H2 100.4 -> L2 95
Third test reaches the resistance area around 100.

A. Upward break on the third visit
   Third visit -> close 101.3 -> close 102.0
   No qualifying third high under the 0.5-unit tolerance.
   The anticipated triple top never becomes a candidate.

B. Candidate still waiting
   H3 100.2 -> eligible at bar 37 -> close 94.6 -> close 95.4
   The lower support at 94 remains unbroken.

C. Breakdown and retest
   H3 100.2 -> eligible -> close 94.4 -> close 93.6
   Later bar: high 94.1, close 93.7
   Support breaks, then the retest closes beneath it.

D. Breakdown and reclaim
   H3 100.2 -> eligible -> close 94.4 -> close 93.6
   Later close 94.5
   The breakdown happened; the later reclaim is a failure event.

Keep pivot time separate from signal time

A label placed at the third high can make a detector look earlier than it was. With three right-side bars, H3 at bar 34 cannot be confirmed until bar 37 closes. Store both times. TradingView documents the risk of plotting pivot information into the past; a historical drawing does not establish that an alert was available on the pivot bar.

Replay the bars around H3 and the support crossing. Inspect candidate creation, the first eligible close and the later state transitions separately.
A wick below support that closes back above it fails the close-based confirmation rule. An intrabar alert is a different rule and needs a separate evaluation.
Waiting for the current candle to close reduces changes from unfinished price data. It does not establish that the full script has no other repainting behavior.

Separate state counts from trade returns

Keep an event log with a candidate ID, H1/H2/H3 times, both lows, eligibility time, N, U and the final outcome. Count invalidated and expired candidates even though they have no entry. Among confirmed candidates, track retests and later reclaims without deleting the original confirmation. These counts describe the detector; a trade report answers a different question.

Specify stop orders, maximum holding time, position size, commission and slippage before backtesting. In a default TradingView strategy, an order generated after a close normally fills no earlier than the next bar open.
Test a separate date range without changing the rules to fit it. This page supplies no measured win rate, preferred timeframe or performance claim.
Ask Pineify AI Coding Agent to draft the candidate, confirmation and invalidation logic with explicit timestamps. Inspect and validate the generated source before testing.
Use Backtest Report for separate TradingView trade exports from each entry experiment. Maintain untraded state counts in the event log; the report does not automatically reconstruct pattern candidates from price charts.

Record	What it can answer	What it cannot establish
Candidate and event log	How many eligible sets confirmed, expired or invalidated; how many confirmed breaks reclaimed support	Trade profitability without entry, exit, size and cost rules
Breakdown-entry strategy export	Returns from a specified entry after confirmation	What a retest-only entry would have earned
Retest-entry strategy export	Returns from entries that met the additional retest condition	The number of all untraded candidates from trade rows alone

Where Pineify fits

Pine Script AI Coding Agent

Turn a written triple top specification into a Pine Script draft for review. Include pivot delay, frozen support, expiry and failure rules in the request. This educational page does not run a live pattern scanner.

Also useful: Backtest Report. Analyze exported strategy trades for each separately defined entry experiment. Use an independent event log for candidates that never became trades.

Draft triple top rules

Related resources

TradingView chart patterns Double top confirmation and failure Triple bottom confirmation Bar-close alert timing

Frequently asked questions

Educational information, not investment advice. All numerical prices, thresholds and paths are hypothetical. No backtest or live performance is reported. A confirmed support break can fail and does not guarantee a profit.

Sources and verification

Last verified: 2026-10-16

Published by Pineify. Pineify is an independent product and is not affiliated with TradingView. Sources support the technical explanations; examples do not establish trading performance.