Elliott Wave has usually been one person's opinion, which makes it hard to trust. Ours is a count run by a written set of rules, revised in numbered steps, reviewed by a separate reviewer on recent revisions, and published as odds across several scenarios instead of one call. It does not tell you what the market will do. It tells you what could happen, how likely each case is, and the level that would prove the leading reading wrong.
The problem you have probably felt
Elliott Wave has a bad name, and some of it is earned. For most of its history a wave count has been one person's judgment, made by eye, on a chart. That makes it subjective, and subjective is hard to test.
Put three analysts in front of the same chart and you can get three different counts and three different opinions. It is not a character flaw. A count turns into a market view, the view turns into a position, and then it is only human nature to want to be right. Confirmation bias creeps in. The count starts bending toward the opinion instead of the other way around.
You have probably met the result: the analyst who has been calling the top for years, and calls it again every time the market goes up. A call that never resolves is an expensive thing to follow. If that is why you gave up on Elliott Wave, you were reacting to something real.
What changes when a machine runs the count by rules
The trouble was never the idea that markets move in waves. It was that the count depended on whoever was holding the pen.
Modern AI makes a different approach affordable. Write the rules down in full. Have a machine apply them to the data, by rule, with no ego to defend. When the market does something the count did not expect, the count changes, and the change is written down.
That is what we do. It does not remove judgment from the process. It puts most of the judgment into the rulebook and a written record of rulings, where it can be read, argued with and improved, instead of leaving it inside one person's head.
How it works, in plain words
We keep a dated, numbered record of the revisions on file, starting with the oldest one, Revision 24, and a written review record behind the recent ones. We are not claiming the count gets more accurate with age, and we are not publishing a win rate. What we are claiming is that the process can be inspected.
The leading count comes with the price that proves it wrong
Many market views never say what would prove them wrong, which is how a view can survive any outcome. An Elliott Wave count is built differently. It rests on a few hard rules, and some of them draw a line on the chart. In a standard five-wave advance, wave 2 cannot close below where wave 1 started, and wave 4 cannot move into wave 1's price territory. Break one of those lines and, in a standard impulse, that reading is not just under pressure. It is wrong. The three rules lesson walks through each one.
That gives the reading a concrete price, the level where it fails. For the leading scenario, the weekly Nasdaq page lists the levels that confirm it, a softer level that means it needs a second look, and a harder level that kills it. The rule lines above are the principle behind levels like these. Think of the kill level as a place to define risk: the price at which the reading is wrong, named in advance. What you do at that price is your decision.
It is also the answer to the analyst who keeps calling the top. When price breaks one of those rule lines, the reading is wrong and the count has to be revised, with the change written down. The process is built so that a count is not stretched to keep an old call alive.
What this is not
It is not a prediction. Nobody knows what the market will do next, and no method can honestly claim to. What the odds give you is a map: which cases are live, which are fading, and which price levels would shift the balance.
It is not advice. This is structural analysis for education and information. It does not tell you what to buy or sell, and it does not know your situation. You are responsible for your own decisions, and Elliott Wave reads can be wrong, ours included.
It is not a promise of results. We do not claim any win rate, track record or performance figure, and nothing here should be read as one.
Where to go next
If you want to see the method working, start with the free weekly read. If you want to understand the vocabulary first, the Learn primers teach it plainly. If you want it in your inbox, the free list is the place to start.
Common questions
Why do Elliott Wave counts differ between analysts?
Because most counts are made by eye, and the rules leave room for judgment. Two people can read the same chart, start the count at different points and reach different conclusions, each one defensible. Opinion and confirmation bias widen the gap, since a count tends to drift toward the view its author already holds.
How is a machine-run Elliott Wave count different?
The rules are written down and applied by rule, with rulings written down, and changes to the count are recorded in numbered revisions, with the reasoning. Judgment still exists, but most of it lives in the rulebook and a written record of rulings, instead of in one analyst's head.
What is a revision?
A revision is a numbered, dated write-up of the full count: what the market did, what it changed in the reading, and the current odds across the scenarios. As of October 19, 2026, the Nasdaq count stands at Revision 246.
What does it mean when an Elliott Wave count is invalidated?
Every count rests on hard rules, such as wave 2 not retracing past the start of wave 1 and wave 4 not entering wave 1's territory in a standard impulse. When price breaks the level a rule draws, that reading is wrong. The level is known in advance, so it works as a place to define risk. It is not an instruction to buy or sell.
Does Degree & Phase predict the market?
No. It assigns probabilities to the scenarios that could play out, updates them as the market moves, and names the price levels that would confirm or kill the leading one. It is structural analysis for education, not investment advice.