Breaker Block vs Change of Character
A breaker block is an area price broke through and later revisited from the other side, which requires the prevailing sequence to have failed. A change of character is that failure, the first swing break against the run, so one produces the other in the strict version of the method.
One of these is the moment a sequence breaks against itself. The other is the area that moment leaves behind. The strict version of the method only calls an area a breaker when a change of character created it, and treats breakers left by an ordinary continuation break as weaker. That claim can be counted, and on bitcoin, a market that trades every day, the change-of-character breakers came out no better.
What each one is
A change of character is the moment price closes through a swing level on the side opposite the run in force. In a sequence of higher highs and higher lows, it is the close below the last swing low. Change of character covers it.
A breaker block is an area that failed, after which structure changed and price came back to it from the opposite side. The site’s breaker block page builds it from the nearest candle of the opposite color before the break, an up candle when the break is downward, and its full range is the block.
The warning is what makes the area a breaker in the strict reading. Without a break against the prevailing sequence, an area that failed is simply a level that did not hold.
Where they differ
Moment against place. The warning happens in a moment; the area persists afterwards and can be waited at for days.
How often each occurs. The warning fires at every pullback deep enough to break a swing against the run. On bitcoin’s daily bars the rule left 82 complete breakers after a change of character and 112 after a continuation break, so the filter removes a bit more than half the candidates.
What each needs defined. The warning needs a swing rule. The area needs that plus which candle forms it and what counts as a return.
What each tells you. The warning tells you the sequence you were trading has stopped. The area tells you where to look if you want to trade the new one.
What the warning and its block share
Both rest on the same swing definition. Change how many bars a swing needs, whether wicks count or which timeframe is read, and both the warning and the area move.
Both are frequent. A market that keeps reversing every few bars supplies opposing breaks and failed levels constantly.
And neither supplies a stop. A change of character marks a level and a breaker marks an area, but the invalidation just beyond either has to survive the next wide bar, which the drawing says nothing about.
Which one to use: Breaker Block or Change of Character?
Take the warning as the event, not the entry. Its defensible use is to stop trading the old direction; treating it as a reversal signal is where losses come from.
Use the area as the place afterwards. Once the sequence has broken, the area gives somewhere to act with an invalidation you can size against.
Require both if you want fewer setups, not better ones. On the bitcoin count below, breakers left by a change of character were visited less often and held no better.
And when you have an area but no rule for the break, do not call it a breaker. It is a level that failed, which is a weaker and much more common thing.
Does the change of character make the breaker better on bitcoin?
The claim. A breaker created by a change of character marks the point where the old trend’s last defenders gave way, so price should respect it on the return. A breaker left by a continuation break is just a pause in a trend that was already running.
The count. Every break of a five-bar swing level on bitcoin’s daily chart from December 2014 to 4 October 2026, each block drawn by the same candle rule and given 20 days for price to come back.
| Breaker left by | Breakers | Came back within 20 days | Held on the return | Closed through the far side |
|---|---|---|---|---|
| Change of character | 82 | 38 (46.34%) | 10 (26.32%) | 5 (13.16%) |
| Continuation break | 112 | 69 (61.61%) | 29 (42.03%) | 15 (21.74%) |
One gap is borderline, the rest are noise. Came back: p = 0.041. Held: p = 0.142. Closed through: p = 0.313 (Fisher exact, two-sided). Three tests were run, and with that many a p of 0.041 is weak evidence; it would not survive a correction for multiple tests. Either way, the direction is the opposite of the claim: the change-of-character breakers were the ones price returned to less.
How long the returns took. When price came back, it took a median of 4 days after either kind of break; 7 of the 38 change-of-character returns and 15 of the 69 continuation returns came the very next day (Mann-Whitney test on the waits, p = 0.852).
Direction split the continuation group. Upward continuation breakers held on 23 of 42 returns (54.76%), downward ones on 6 of 27 (22.22%). After a change of character the figures were 6 of 21 upward and 4 of 17 downward. Bitcoin rose a long way over these years, which works against any level that asks price to stay down.
A change of character and its breaker: bitcoin, September 2026
The warning. On 15 September 2026 bitcoin closed at $75,612.51, below the $76,248.30 swing low of 2 September: a change of character, because the break before it had been upward.
The block was the last up candle before that close, 14 September, which opened at $76,830.28, closed at $78,163.38 and traded between $76,367.38 and $79,568.95.
The return came the next day. On 16 September price rallied to $76,474.08, just inside the block, and closed at $76,150.32, back below it: a hold on the return, one of the 10 in the table.
The original data
Data. Yahoo Finance daily bars for BTC-USD, opens included, dated by UTC day from 17 September 2014; the last full day used is 4 October 2026. A break in the same direction as the previous break is a continuation; the opposite direction is a change of character. The swing rule (a swing high must top the five bars on each side and is usable five days later) is the code that reproduces this site’s SPY break of structure list row for row.
Breakers and outcomes. Block = the nearest opposite-colored candle within 60 days before the break, whole range, kept only when the break close cleared it; 8 breaks failed that test. Return = the earliest of the next 20 days whose range reaches back into it; held = a close on the break side of the block; one breaker from late September 2026 is still inside its 20-day window and is left out. The bitcoin breaker block file lists every break with its type, block and outcome.
What is left out. The count stops at the return, so neither group is shown to forecast a move. The groups are modest in size, and the count is for one market on one interval.
When it fails
The usual failure is marking breakers at any failed level, and a trending market supplies them endlessly. Price goes through an area during an ordinary pullback, no swing has broken against the run, and the area is nonetheless marked and traded from the other side. That is a position against the prevailing direction taken because a minor level did not hold. The trend resumes, the stop goes, and the next pullback produces another. Requiring a change of character stops that particular error, but the count above shows it does not turn the remaining breakers into better trades.
The second failure is trading the warning as an entry. It fires on pullbacks.
A third is no swing definition. Both concepts then follow the outcome.
A fourth is no expiry on the area. Failed levels accumulate forever.
A fifth is ignoring direction. On bitcoin, blocks that needed price to stay down held least.
And a sixth is accepting a marginal break. A close a few dollars through a swing level is well within what an ordinary bar does unaided.
Related
Breaker block explains how the block is drawn. Change of character covers the event that creates it. And break of structure describes the with-trend break used as the comparison group here.
The character change is the event; the breaker is the address it leaves behind. If you only take breakers that follow a change of character, do it because it gives you fewer setups to manage, not because it gives you better ones, until your own count says otherwise.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.