Day Trading Setups
Round Number / Psychological Level Magnet
Whole-dollar prices like $100.00 act as informal magnets and resistance points because of clustered resting orders and plain human anchoring, even when there is no technical level there at all.
Round numbers versus a matched control
A long generated session plays out. Every whole-dollar level it crosses gets marked, and an equal number of deliberately non-round control levels get marked at comparable spacing. The generator is biased so round numbers really do cause more pausing and reversing, then the tool tallies the honest rate for each, live, as you move the tolerance slider.
Generating the session…
How it works
- Every whole-dollar level the series crosses gets marked as a round level. For each one, the tool also marks a matched non-round control level at a fixed, comparable offset, so the two groups have the same count and roughly the same spacing.
- A pause or reversal is a real check on the bars right after a touch. When price comes within tolerance of a level, the tool looks at the next few bars, if price fails to close decisively beyond the level by more than the tolerance, that touch counts as held. If it powers through, it counts as broken.
- The tolerance slider changes how close counts as a touch, and both rates recompute instantly from the same underlying bars. Nothing about the price series regenerates when you move it, only which touches qualify and how they are judged.
- The comparison only means something because the control levels are not round. If round levels and control levels showed the same hold rate, that would be evidence against the magnet effect, not for it.
Where this breaks
It is a tilt, not a wall
The magnet effect here is a real but modest statistical tilt, typically a difference of single digits in percentage points between the round and control rates, not a level that reliably holds. Price blows through whole-dollar levels constantly, and on higher-priced instruments or during fast, high-volatility stretches the effect gets swamped entirely by normal noise, the round-level rate and the control rate can converge or even flip on a given session. Treat this as a small edge that shows up over many trials, not a level you should expect to hold on any single approach.