Most articles about chart patterns quote a "success rate" and stop there. We wanted a plainer answer: when a strategy trades wedge, triangle and channel breakouts with real stops, real targets and real fees, which shapes actually make money?
So we took every trade our Price Action strategy made on Bitcoin over seven years and split them by pattern. The short answer: triangles did almost all of the work. They made 38 of the 58 trades and +5,783 USDT, more than the strategy's whole net profit. Wedges roughly broke even, and channels barely traded at all.
- Triangles produced almost all of the profit: +5,783 USDT from 38 of 58 trades.
- Wedges came out close to flat: falling wedges +450 USDT, rising wedges -147 USDT.
- Channels barely traded: 3 trades in seven years, all of them ascending channels, together -481 USDT.
- A high win rate is not the same as profit: rising wedges won 64% of their trades and still lost money.
- The samples are small (0 to 21 trades per shape, on one coin), so read this as a clue, not a rule.
The test: what was traded, and how
The trades come from the published backtest of Navixa Price Action on TradingView: Bitcoin perpetual futures on Binance (BTCUSDT.P), 4-hour candles, from October 2019 to October 2026.
The rules were the same for every pattern:
- The strategy looks for eight shapes: rising and falling wedges; symmetrical, ascending and descending triangles; ascending, descending and horizontal channels.
- A pattern is drawn only after its last turning point is confirmed, and it never repaints.
- A trade opens only when a candle closes beyond the pattern line by a margin scaled to Bitcoin's volatility. A wick through the line is not enough.
- The stop, two targets and the position size are all set at the breakout close. After the first target, the stop on the rest of the position moves just past the entry.
- Each trade risks 2% of the balance. Breakouts on Sundays are skipped.
The account started with 10,000 USDT. Every result below includes 0.05% commission per side and 5 ticks of slippage; funding fees are not included. Over the whole period the strategy made +56.05% (+5,604.79 USDT) with a 6.03% maximum drawdown and a profit factor of 2.491. You can check every number on the full results page, and the backtest methodology explains how the tests are run.
The results, pattern by pattern
| Pattern | Trades | Closed in profit | Profit factor | Net result |
|---|---|---|---|---|
| Symmetrical Triangle | 21 | 71% | 2.84 | +2,176 USDT |
| Ascending Triangle | 10 | 80% | 4.70 | +1,748 USDT |
| Descending Triangle | 7 | 71% | 4.88 | +1,859 USDT |
| Falling Wedge | 6 | 83% | 2.86 | +450 USDT |
| Rising Wedge | 11 | 64% | 0.83 | -147 USDT |
| Ascending Channel | 3 | 33% | 0.05 | -481 USDT |
| Descending Channel | 0 | none | none | none |
| Horizontal Channel | 0 | none | none | none |
Grouped by family, the picture is even clearer:
| Family | Trades | Closed in profit | Net result |
|---|---|---|---|
| Triangles | 38 | 28 of 38 (74%) | +5,783 USDT |
| Wedges | 17 | 12 of 17 (71%) | +303 USDT |
| Channels | 3 | 1 of 3 (33%) | -481 USDT |
Profit factor is gross profit divided by gross loss. Above 1 a pattern made money after fees; below 1 it lost money.
Triangles: where the profit came from
All three triangle types made money, and none of them was close to break-even. Ascending and descending triangles had profit factors near 5, which means their winners made almost five times what their losers cost. The symmetrical triangle was the workhorse: the most trades of any shape (21) and the largest net result.
We can only guess at the reason, and it is a guess. A triangle has two clearly converging lines, so a close beyond one of them is a clean, specific event, and the stop has a natural place on the other side of the pattern. Whatever the cause, on this chart triangles were the shape the strategy could trust most.
Rising wedges: a 64% win rate that lost money
This is the most useful row in the table. Seven of the 11 rising-wedge trades closed in profit, and the pattern still finished at -147 USDT. The four losing trades cost more than the seven winners made.
How does that happen? Look at how all 58 trades ended:
- 26% reached both targets.
- 38% reached the first target, and then the rest of the position closed at the moved stop, just past the entry.
- 27% were stopped out before the first target.
- 9% were closed by a breakout in the opposite direction.
The second group counts as wins, but they are small wins: part of the position is banked at the first target, and the rest closes near the entry. A stop before the first target, on the other hand, costs the full planned risk. A pattern with many small wins and a few full losses can show a high win rate and still lose money. That is why the profit factor and the net result, not the win rate, tell you whether a pattern earns its place.
Channels: too few trades to judge
In seven years of 4-hour Bitcoin candles, channels gave the strategy just three trades, all on ascending channels, and only one of them made money. Descending and horizontal channels never produced a breakout that met the rules.
Three trades say almost nothing about channels in general. What they do say is that on Bitcoin's 4-hour chart, channels clean enough to trade under strict rules are rare.
These are 58 trades on one coin and one timeframe. A shape with 3 to 11 trades can swing from a loss to a profit with one or two more trades, so don't decide which patterns to trade from this table alone.
Does this hold on other coins?
The pattern split above exists only for the published Bitcoin chart. We also ran the strategy, with the same default settings, on the 40 largest coins by market cap (stablecoins and wrapped tokens excluded) on the 1-hour, 2-hour and 4-hour charts: 120 tests in total. 94 of the 120 were profitable, with a median net result of +10.7% and a median profit factor of 1.35. That test does not record which pattern each trade came from, so we can't tell you yet whether triangles lead on every coin. The coin-by-coin numbers are on the results page.
Two things are worth noting from the multi-coin test. The Bitcoin result is better than the typical coin, so don't expect +56% everywhere. And the median coin was still profitable, so the result was not limited to Bitcoin.
Win rate vs real results: why "success rates" disagree
If you search for chart pattern statistics, you'll find numbers that rarely match each other. That is mostly because they measure different things. altFINS, for example, defines a pattern's success rate as price reaching the forecast target within one pattern length. The widely quoted statistics of Thomas Bulkowski come from stock charts, not crypto. Neither measures what a trader actually takes home after stops, partial exits and fees.
Our numbers answer a narrower question: under one fixed set of rules, with costs included, which shapes made money on Bitcoin? That makes them less general, but closer to what happens in a real account.
What to take from this
- Judge a pattern by its net result, not its win rate. A high win rate built on small wins can hide a losing setup.
- Plan the stop and targets before you enter. Every trade in this test had its exit plan fixed at the breakout close, which is what kept the largest loss at the planned risk.
- Wait for a close, not a wick. The strategy ignores breakouts that don't close beyond the line by a clear margin.
- Distrust small samples. Ten trades are an anecdote. Look for dozens of trades, ideally across several coins.
FAQ
Which chart pattern is the most reliable?
There is no single answer, because it depends on the market, the timeframe and how you define "reliable". In this seven-year Bitcoin test, triangles made almost all of the profit, but with 7 to 21 trades per triangle type the samples are small.
Is a rising wedge bullish or bearish?
Textbooks usually read a rising wedge as bearish, a pattern that tends to break down. Navixa's strategy does not assume a direction: it trades the side where a candle closes beyond the line. On Bitcoin's 4-hour chart, rising-wedge trades lost a little money overall.
Do chart patterns work in crypto?
Under the rules of this test, triangle breakouts on Bitcoin were clearly profitable after fees, wedges were close to flat and channels were too rare to judge. Patterns are a tool, not a guarantee; they work best with fixed exits and position sizing.
The bottom line
On seven years of Bitcoin 4-hour data, the breakout edge came from triangles. Wedges were close to neutral, channels hardly traded, and the rising wedge showed why a win rate alone can mislead you. If you want to see these rules at work, every trade is on the Price Action results page, and the strategy page explains how the stop, targets and size are set. For more on trading breakouts in general, see our guide on how to trade Bitcoin price breakouts.
