5 Famous Setups, 5 Free Filters That Fix Them

Every setup below is famous, taught everywhere, and traded by thousands of people every session. Every one of them also has a measurable flaw that a single filter fixes, and in every case the filter is not a better entry. It is a condition. When to take the trade, not how.

I know this because I test setups mechanically for a living, on years of session data, and the same lesson keeps falling out. The edge is almost never in the trade. It is in the conditions you take it under. Here are five worked examples, with the numbers, free.

1. The Opening Range Breakout

The classic ORB takes the break of the first 15 minutes. Traded blind, it is close to a coin toss, and the reason hides in plain sight. Not all opening ranges are equal.

When the opening range is tiny relative to price, under roughly 0.2%, the break holds clean only about 46% of the time. Under breakeven for a symmetrical target. When the range is large, above roughly 0.6% of price, clean breaks jump to about 69%.

The fix takes one division. Range height over price. If the number is small, the auction has not committed to anything and the break is noise. Skip it. The breakeven line in my testing sits around 0.4%, and simply refusing to trade below it removes most of the losing population in one move.

2. The IB50

The famous Initial Balance midpoint trade: buy or sell the 50% retracement of the first hour's range in the direction of the early move. I tested it unfiltered across two years of data. Profit factor 0.92. It loses.

The same trade, conditioned, is how I trade every week. The conditioning is not a secret entry tweak. It is the day's statistical lean, which extreme formed first, and whether the historical sample for today's exact conditions is even worth acting on. Some days the honest answer is that there is no read, and the filter's output is no trade at all.

The fix is a question, asked before every session. What do the measured conditions say about today, and is the sample behind that answer thick enough to trust? I automated exactly that question into an indicator because I got tired of answering it by hand, but the principle costs nothing. An unconditional setup is a coin with a story. The conditions are the edge.

3. The Gap Fade

Overnight gaps close, and they mostly close the same session. Across six years of Gold sessions, 1,305 of them, the gap filled back to the prior close 96% of the time. That number is why the gap fade exists.

But the 4% that never fill are not random, and they are where the strategy dies. The failures cluster on two conditions. A gap far beyond the normal daily range, and a one-way tape from the opening bell. Small gaps fill almost without exception. Oversized gaps are where the disasters live, and they are visible before entry.

The fix is a stand-down rule. Measure the gap against the average range. If it is an outlier, the base rate you are betting on no longer applies, so do not bet it. You keep the 96% behaviour and step out of the tail that funds everyone else's losses.

4. The Silver Bullet

The most hyped setup in futures. A liquidity raid fails inside a fixed one-hour window, you trade the reversal. I built it mechanically, every rule written down, and ran 12 months of NQ data through it.

In the famous 10am window, the one all the content is about, it won 38.9% at 1:1. A losing trade, comfortably. The identical rules in the ignored 2pm window won 60% and finished the year profitable.

Same setup. Same rules. Four hours apart. Opposite result. The fix is not a better raid definition or a smarter entry. It is the clock. If you take nothing else from this piece, take this one, because it is the purest example I have ever tested of the general law. When beats what.

5. Every setup above

The uncomfortable one, and the reason this article has an expiry date built in.

Every split above was measured over a historical window, and the splits move. In my IB testing, the profitable direction itself flipped between the recent regime and the full two-year record. Both directions were profitable somewhere in the data. Never at the same time. A filter that was true over the last two years is a fact about the past, not a promise about next month.

The fix here is a habit, not a rule. Re-check the numbers on a schedule, and treat any filter you have not re-verified recently as expired. The traders who get hurt by conditional edges are not the ones who use them. They are the ones who froze them.

The pattern, and where the live numbers are

Five setups, five filters, and not one of them touched the entry. Size, conditions, size again, clock, and regime. The retail world optimises entries because entries are visible on a chart. The filters live in the statistics, which is why almost nobody applies them.

All five sets of numbers above come from my own testing and from the Market Edge engine, which computes these conditional splits across five markets and updates them as sessions close. The rules are free, they are all here, and they always will be. Knowing whether the filter still holds this month is the product. That is Tradestar Pro, and the live version of every number in this piece is in there.

A Market Edge rate is a fact, not a signal. These numbers are historical, the splits rotate with regime, and a strong rate is not a promise. Trade your own plan.