Whether chaos magic is real is something you can test for yourself but can't settle by arguing, because the practice doesn't ask you to believe it first and the only evidence that matters is your own logged results.

The law on this site is three lines long. Do the work. Check the results. Believe nothing.

This page is the second line.

What does "real" mean here?

It can mean three things, and they have different answers.

Does the practice do something to you? This site's reading: the focus, calm and clarity people get from a short ritual are ordinary psychology, and you can watch for them yourself.
Does it change what you do? A sigil that keeps a goal in the back of your mind can change your behavior without anything supernatural going on.
Does it change events outside you? This is the claim people mean when they ask. It's the hardest to test, and the rest of this page is about testing it.

Chaos magic itself treats belief as a tool, and treats the symbol systems it borrows as maps with no truth value of their own. The scholar Hugh Urban described it as traditional occult technique combined with applied postmodernism, where systems work only through the belief of the practitioner. You're free to read that as psychology. Plenty of practitioners do, and they still keep logs.

What do practitioners claim?

Gordon White, who died in May 2026, framed what magic does as enhancing the probability of certain outcomes over others, which makes a failed spell expected some of the time. That framing is useful because it can be checked. If a working shifts odds, enough of them should come in above the rate you'd see without it.

The same framing is also a trap. Any result can be explained after the fact, and any miss can be filed under "expected some of the time."

So you decide the test before you run the working.

How to test a working

1. Pick an outcome you can check

"My life improves" can't fail. "A client replies to my email this week" can. Choose something with a clear yes or no and a deadline.

2. Write down the odds before you start

How often does this happen to you without any magic? If clients reply to about one email in five, your baseline is one in five. Without a baseline, a hit means nothing. With one, a hit means a little.

3. Run the working once and file it

Make the sigil, charge it, forget it. File the date and the check-back date. The Sigil Foundry and the Shoal tool both file a receipt in the Log that keeps the date and drops the intent.

4. Run a control

Choose a similar outcome you won't work on, and log it too. If both land at the same rate, the working added nothing you can see.

That's the step almost everybody skips, because it's the one that can embarrass you.

5. Check on the date and mark it

Mark each entry held, partly or nothing. Don't touch the wording of the goal after the fact. Rewording a miss until it reads like a hit turns the log into fiction, and you're its only reader.

6. Repeat until the count means something

One result tells you nothing. Ten tells you a little. Thirty tells you more. A coin that lands heads three times in a row isn't doing a trick, so you need enough workings that luck stops being a good explanation.

Why memory is the main enemy

You will remember the hits and lose the misses. You will count a near miss as a success. You will credit the sigil for the thing that was going to happen anyway.

Everybody's memory does this. Yours isn't special.

Most workings won't beat a coin. Log enough of them and you'll know which of yours do. Log all three outcomes: held, partly, nothing.

What counts as a result?

A held result that beats your baseline, repeated across many workings, with a control that doesn't do the same. Anything less, this site files under stories.

You may find nothing. That's a legitimate result, and this site has no stake in a different one.

You may also find that the practice makes you do the thing, and the doing was the effect all along. Worth knowing. Write that down too.

A note on safety

Testing is no reason to run high-stakes workings. Keep test goals small and dull, the kind you'd shrug at either way. If a practice makes your sleep or mood worse, stop and talk to a person.

Where to go next

What chaos magic is has the background.
How to make a sigil is the working to test first.
Chaos magic for beginners sequences the whole Library.

Now do it: open the Log.

Is chaos magic real? How to test it