Day Trading Setups
Halfback / 50% Retracement Continuation Entry
Just the literal midpoint of the last swing leg, no fib grid, no multiple levels. Widen or narrow the tolerance band around it and watch which real pullback bars touch the zone, hold, or fail straight through.
Set the zone width, log the touches, watch the hold rate
Every setup starts with a real swing leg. Halfback is just the average of that leg's start and end, drawn live. The slider sets the tolerance band around it as a percent of price, shaded as a magnet zone on the chart. Click New Setup to play a fresh pullback bar by bar, every real bar whose high or low enters the zone logs as a touch. The table recomputes instantly whenever you move the slider, against every touch logged so far, not just the current setup.
Building the swing leg…
| Touches | Held | Failed through | Hold rate |
|---|---|---|---|
| 0 | 0 | 0 | — |
How it works
- The swing leg is real, generated first. A directional impulse of several bars sets a start price and an end price, both real numbers taken straight from the generated candles.
- Halfback is just the average of those two, nothing more. (start + end) / 2, computed live, no Fibonacci ratios involved.
- Every touch is logged from the actual pullback bars, once. A bar whose real high or low overlaps the current zone counts as a touch, and what happens to that specific bar doesn't change afterward.
- Widening the slider re-scores the same history, it doesn't regenerate it. Every touch and its hold-or-fail outcome across every setup you've run gets checked again against the new zone width the instant you move the slider.
Where this breaks
A wider zone racks up a better hold rate by counting weaker touches as wins
Widening the tolerance band doesn't just catch more touches, it also counts shallower, less committed pullbacks as having "held," because a bar that barely grazes a wide zone's edge gets scored the same as one that plunged into the middle of a narrow one and still bounced. A hold rate that looks great at 1.2% tolerance can be doing most of that work on touches that were never a real test of the level in the first place, which is exactly why the raw touch count next to the hold rate matters as much as the rate itself.