> For the complete documentation index, see [llms.txt](https://help.hydravision.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://help.hydravision.ai/products/asset-profiler/directional-bias.md).

# Directional Bias

The panel headed **"Directional bias"**, and the second badge beside the symbol.

{% hint style="warning" %}
**The common misreading.** That a **NEUTRAL** reading means bias doesn't matter on this asset. The bias setting matters on every asset, for a reason that has nothing to do with which way price went. See below.
{% endhint %}

### What it means

The Profiler measures how directional the asset was over your window and summarises it as a direction and a strength. Both come from the daily (D1) bars; the four-hour ADX is shown alongside for comparison only.

**Direction** is **LONG** or **SHORT** when the D1 weighted texture leans clearly one way (beyond ±0.05), and **NEUTRAL** otherwise.

**Strength** is one of three values:

| Strength     | When                                                                |
| ------------ | ------------------------------------------------------------------- |
| **STRONG**   | D1 ADX average above 40, or above 35 with a texture beyond ±0.40.   |
| **MODERATE** | Not strong, but D1 ADX average above 20, or a texture beyond ±0.30. |
| **RANGING**  | Neither: ADX at or below 20 and a texture within ±0.30.             |

The badge beside the symbol shows both, for example **NEUTRAL · RANGING** or **SHORT · MODERATE**.

| Row                          | What it tells you                                                                   |
| ---------------------------- | ----------------------------------------------------------------------------------- |
| **D1 ADX average**           | How strong the trend was, day to day. Above 25 is conventionally a trending market. |
| **D1 bars above ADX 25**     | What share of days qualified. More informative than the average.                    |
| **D1 weighted texture**      | Whether the movement leaned up or down, weighted by how decisive each bar was.      |
| **D1 slope**                 | The plain drift over the window, as a percentage per day.                           |
| **H4 ADX average**           | The same strength measure on the shorter timeframe.                                 |
| **Counter-trend multiplier** | A single figure for how trending the window was, from D1 ADX and texture together.  |

### The counter-trend multiplier

A summary of trend strength on a scale from 1.0 to 2.0. It is calculated from the D1 ADX average and the size of the D1 weighted texture, not measured from retraces.

* **1.0** when the D1 ADX average is 20 or below, a market with no meaningful trend.
* Above that it rises with both ADX (reaching its full share at ADX 50) and texture (reaching its full share at ±0.60), up to **2.0** when both are at or beyond those levels.

It describes how strongly the window leaned one way, which is the direction a grid positioned against the trend would have been working against. It is not a measured count of how much further price had to come back.

### Why bias matters even when the reading is NEUTRAL

This is the part worth reading properly.

**A grid only holds one direction at a time.** If sells are open, buys can't trade. That's the engine's design.

So with bias left off, whichever side opens first takes the clock, and the other side simply never gets room. What you end up trading isn't decided by which side had the better prospects. It's decided by which one happened to open first.

That effect exists on every asset, trending or not. It's a consequence of the engine, not of the market.

The setting that controls this is **Bias Selection**. Set to a side, it decides which side you're running; left at *No Bias - Buy or Sell*, the order in which grids open decides it. The panel's note says the same thing.

{% hint style="warning" %}
**Never compare the two sides by net profit.** Net profit is decided partly by how long each side happened to be allowed to trade, which is exactly the thing that's unfairly distributed. Compare **return per lot** instead. It's the only comparison that isn't distorted by occupancy.
{% endhint %}

### What this panel is not

It's a description of the window measured, not a forecast. Gold trended down through the second half of 2022; that says nothing about next quarter.

It's also not a recommendation to trade in the measured direction. The reason to set a bias is to control which side gets the clock, not because the Profiler thinks the trend will continue.

### What to do with this

* **Bias Selection** decides which side runs on every asset, whatever this panel says; without it, the first grid to open decides.
* When comparing your two sides in testing, use return per lot.
* Read the direction as history, not as a signal.