Part 3: Elliott waves · Chapter 12 of 14 · 7 min
Many counts for the same chart
The same high is wave 5, wave 3 or wave 1 depending on the threshold and the starting point: with three thresholds the valid impulses are 44, 60% of the candles have at least one and 35% at least two. After the end of an impulse price makes a new extreme in 62% of cases, as after any swing (67%).
The same high, three numbers
Chapter 8 showed that changing the swing threshold changes the degree. Here we look further: the number the same high is called by. Figure 1 shows, between 8 and 30 January 2026, three valid impulses (chapter 7 rules) on the same prices:
- A, 3 ATR swings from 8 January: from 4,410.01 $ to 5,110.47 $.
- B, 2 ATR swings from 16 January: from 4,540.03 $ to 5,625.34 $.
- C, 2 ATR swings from 21 January: from 4,755.59 $ to 5,457.92 $.
Role of the same two highs in the three counts:
| High | A (3 ATR) | B (2 ATR, from 16 Jan) | C (2 ATR, from 21 Jan) |
|---|---|---|---|
| X: 21 January at 04:00, 4,889.61 $ | end of wave 3 | end of wave 1 | precedes the start of the count |
| Y: 26 January at 04:00, 5,110.47 $ | end of wave 5 | end of wave 3 | end of wave 1 |
High Y is wave 5 in A (a count that ends there), wave 3 in B (a count that continues to 5,625.34 $) and wave 1 in C. All three readings meet the rules, and in real time there is no way to know which will be the one that "counts": A says the rise is over, B and C say it goes on. With three thresholds and different starting points, each swing can take several labels.
Read the chart in words
H4 gold chart from 8 to 30 January 2026 with three broken lines on the same prices (three counts) and the two highs X and Y marked with a letter.
| When | What | Price |
|---|---|---|
| 8 Jan 2026 12:00 UTC | 3 ATR from 8 Jan | 4,410.01 → 5,110.47 |
| 16 Jan 2026 12:00 UTC | 2 ATR from 16 Jan | 4,540.03 → 5,625.34 |
| 21 Jan 2026 16:00 UTC | 2 ATR from 21 Jan | 4,755.59 → 5,457.92 |
| 21 Jan 2026 04:00 UTC | Point X | 4,889.61 |
| 26 Jan 2026 04:00 UTC | Point Y | 5,110.47 |
Times in UTC.
How many counts are possible
On H4 gold the valid impulses are 19 at 1.5 ATR, 14 at 2 ATR and 11 at 3 ATR: 44 in all. Over the 1,652 candles from January to October:
| Candles covered by | Number |
|---|---|
| at least one valid impulse | 988 (60%) |
| at least two | 574 (35%) |
| at least three | 346 (21%) |
| an upward and a downward impulse together | 13 (1%) |
The maximum is 7 valid impulses on the same candle. Each of the 14 impulses at 2 ATR has at least one other valid impulse overlapping in time, and 8 have one in the opposite direction (usually the impulse that begins where the first one ends). Of 77 pairs of overlapping impulses, 22 have opposite direction. Figure 2 shows the number of valid impulses covering each day.
Read the chart in words
Daily gold chart from December 2025 to October 2026 with below, in bars, the number of valid impulses covering each day.
At any moment, for more than half the time, at least one valid count exists, and almost always more than one. A count that "works" can almost always be found; it is not the existence of a count that says something about price.
Is the end of an impulse special?
If P5 were the end of a structure, after it price should not make new extremes beyond P5 more often than after any swing. On the 13 impulses at 2 ATR with 60 candles of data after the confirmation of P5, price exceeds P5 in 8 cases (62%). After any swing it does so in 139 cases out of 206 (67%). The difference goes the wrong way for the theory and sits inside the error: P5 is not a special end.
What makes a count falsifiable
A count says something only if, before looking, it declares:
- the swing threshold and the degree;
- the rules (the three of chapter 7);
- the level that refutes it, and what it forecasts (chapter 11);
- a sample of counts, not the best case, and the comparison with a random path.
Without these four things, a count fits the chart: you only have to move the threshold or the starting point.
Procedure in four steps
- Declare threshold, rules and refuting level before looking at the chart.
- Count all the valid counts, at several thresholds, not just the best one.
- Look at how many labels the same swing gets.
- Compare the end of the count with an ordinary swing.
Exercises
Chapter exercises
1. On the chart (January 2026) the letters sit on four highs. Which letter is the end of wave 3 in the 3 ATR count from 8 January and the end of wave 1 in the 2 ATR count from 16 January?
Read the chart in words
| When | What | Price |
|---|---|---|
| 21 Jan 2026 04:00 UTC | Point A | 4,889.61 |
| 26 Jan 2026 04:00 UTC | Point B | 5,110.47 |
| 28 Jan 2026 20:00 UTC | Point C | 5,625.34 |
| 14 Jan 2026 12:00 UTC | Point D | 4,645.15 |
Times in UTC.
2. With three thresholds the valid impulses are 44 and 60% of the candles have at least one. What does it mean?
3. After P5 of an impulse price exceeds P5 within 60 candles in 62% of cases (8 out of 13), after any swing in 67%. What do you conclude?
4. What makes a count falsifiable?
Have a question or want to share your exercise?
Post in the community, or join the free signals room on Telegram.
Educational content, not financial advice. Trading involves risk.