The market is an auction, not a chart.
Every price you're looking at already happened. Somebody wanted in badly enough to pay up, somebody else was happy to sell it to them, and the print is what got left behind.
So when I read a chart I'm not really reading a line. I'm trying to work out who was doing the paying, and where they quit. The chart is a transcript of that argument. It isn't the argument.
You can't know what happens next. You can know what kind of day it is.
That sounds like a downgrade. It isn't.
Whether a session is going to travel or just sit there and chop is something you can get a real handle on before lunch. Where it closes, you can't.
And knowing which one you're in changes what you do next: how big you go, how long you're willing to wait, which setups you're even allowed to look at. Most of the damage I've done to myself came from running a directional plan on a day that was never going anywhere.
Context is the product. Signals are not.
Nobody here is going to tell you what to buy.
You'll see every trade I take, why I took it, and how it ended. That's a record of what I did. It isn't an instruction for what you should do, and the difference matters more than it sounds.
Hand somebody entries and you've trained them to need you. Show them how you got there and eventually they don't.
Every number comes with its sample.
A percentage on its own is decoration.
If I can't tell you how many sessions something was measured over and which ones, it doesn't go up. Same rule for you in the room. "It works most of the time" isn't a claim, it's a feeling.
This is the cheapest standard there is and hardly anyone keeps it.
Beat the dumb version or it isn't an edge.
Before I'll call anything an edge it has to beat the laziest possible version of itself.
On NQ, continuation runs about 0.60 if you do nothing clever at all. That's the bar. If an idea can't clear it after costs then it isn't an idea, and the market will charge you for the confusion.
Hit rate is not edge.
I published a setup here that won 81% of the time and lost money. −0.6 points a trade. The wins were small and the losses weren't.
I killed it, and the kill turned out to be the useful part. Hit rate gets quoted everywhere because it flatters you and it's the easiest thing to screenshot. Expectancy after costs is the only number that pays for anything.
Most of what looks like structure is geometry.
Here's the test. Take the day's moves, keep every single one of them, shuffle the order. If your effect is still sitting there afterwards then it was never about the market. It was arithmetic the whole time.
The best-looking curve I ever built came back 97% geometry. That one stung. Publishing it was worth more than keeping it, and it's the only reason you should believe the ones that survived.
What dies gets published.
The Kill List isn't me being modest. It's the most useful thing on the site.
Something you can stop watching gives you attention back, and attention is what you're actually short of. Not ideas. You've got plenty of those.
A desk that only shows you the winners is showing you a highlight reel. Nobody ever learned to trade off a highlight reel.
You're the biggest variable in your own results.
Run a trading log end to end and the damage is rarely where you expect it. Usually there's no disaster trade, no size blow-up, nothing dramatic at all — just a small, persistent gap between the trades the plan called for and the trades that actually got taken. Repeated a few thousand times, that gap is the whole loss.
The analysis is usually fine. The operator is the variable.
So the same tools that get pointed at the market get pointed back at the desk that runs them, and they'll get pointed at you if you want them to. A standard that only applies to the market isn't a standard. And if you won't measure your own discipline, no method is going to save you.