Settings

This page walks every control you are expected to touch, grouped by the question each group answers. For each group you get defaults, the effect of moving the control, the tradeoff you are accepting, the misuse that most often happens there, and — where the control interacts with something else on the pane — the pairing that turns one wrong setting into a compounding misread.

It is worth reading top to bottom at least once, even the groups you do not plan to enable yet. Several controls have effects that only become legible once you have seen them sitting next to other controls; if you arrive at the Keltner group before the Oscillator group, you will not have the frame for why sensitivity interacts with clamping. After the first pass, this becomes a reference page — return to the specific group when a knob is not doing what you expected.

Two postures govern the whole page. First, defaults are a legitimate starting point, not a prescription. Leaving them untouched is a valid configuration; changing them should follow a reason, not an aesthetic. Second, the Power User groups at the end are conditional. Every control inside them is dormant until the matching moving-average type is actually selected somewhere. Treating those knobs as required is the most common way a serious reader over-invests effort on this indicator.

A short map before you start:

• Oscillator — the global feel of the oscillator, including sensitivity.

• Blend Core — how the blended fast and slow are drawn and whether they are smoothed.

• Blend Div / KC / BBWP / DC — the four structural extras that live on the blended fast line.

• MA 01–05 — five independent slots.

• Power User — conditional controls for the less common MA types.

Oscillator

This group controls the global feel of every slot's oscillator and the reference guides that frame the pane.

The sensitivity trap. ATR Sensitivity is the single most-tuned control and the one most often tuned badly. The temptation is to raise it so the oscillator looks more decisive. What you are actually doing is pushing more of the movement into the bounded edges at 0 and 100, where the oscillator stops carrying resolution. You lose the ability to read inside the body of moves. A saturated oscillator looks confident and reports nothing. If your pane is sitting on the edges, drop sensitivity first, not after.

What the overbought and oversold guides are not. Changing Overbought Level or Oversold Level does nothing to the oscillator's logic. They are anchors for your eye. A reader who moves the 70 line up to 80 because "the pane is spending too much time over 70" is treating the guide as a threshold. It is not. The behavior you are uncomfortable with is a sensitivity question, not a guide question.

Blend Core

This group controls whether the blended fast and slow lines are drawn at all, and whether a smoothing pass is applied to them after the weighted sum.

Smoothing as a relationship to sensitivity. The two places a reader most often reaches for smoothing are when the pane feels noisy and when the oscillator keeps flirting with 0 or 100. The second case is not a smoothing problem; it is a sensitivity problem, and smoothing will lie to you about it by making the saturated line look tidy. Before you enable Enable Smoothing, sweep ATR Sensitivity between 0.5 and 1.5 and see whether the noise goes away. If it does, leave smoothing off. If you still want quieter lines after that, smoothing is the right answer and you are paying for it honestly — one smoothed bar of lag on the blended fast and slow.

Do not hide a live blend. Plot Blend off is occasionally useful when you only want the per-slot lines visible. Remember that the blend is still being calculated under the hood, and the Keltner, BBWP, Divergence, and Donchian extras all draw relative to that invisible blend. Setting up alerts on an invisible blend is a small recipe for confusion.

Blended Divergence

This group controls divergence detection between the blended fast line and chart price. Triangles mark confirmed pivots where the two disagreed.

What Pivot Len: actually is. This is a symmetric pivot strength. A pivot is confirmed only when the bar has Pivot Len: bars of lower highs on either side (for a high) or higher lows on either side (for a low). Shorter lengths produce more pivots that confirm sooner but are weaker. Longer lengths produce fewer pivots that confirm later but are stronger. Do not think of this as a "sensitivity" knob for divergence; it is a strictness knob.

The two places the triangle can sit. With Plot On Pivot? off, the triangle appears on the confirmation bar. That is the most honest presentation because the confirmation bar is the bar on which the divergence actually became known. With Plot On Pivot? on, the triangle is shifted backwards to rest on the pivot bar itself. The information is the same in both cases; only the presentation differs. The trap is reading a back-offset triangle as if the tool "saw it coming." It did not; it saw it on the confirmation bar and then moved the drawing.

Blended Keltner

This group wraps the blended fast line with a Keltner-style envelope that is native to the oscillator.

This is not a price Keltner. The bands are drawn around the blended oscillator, not around price. They widen when the oscillator expands and contract when the oscillator compresses. Reading them as if they were price Keltner bands — expecting a tag to mean a price reaction — is the most common misread here. Use them to see whether the oscillator is pressing against its own envelope, not whether price is.

Clamping consequences. Every KC output is clamped into 0..100. If you push KC Mult: very large, the bands will plaster against the pane ceiling and floor and stop telling you anything. A flat-topped or flat-bottomed envelope is the tool warning you that the multiplier is off the useful range. Drop it back toward 2.0.

Length vs basis length. KC Length smooths the range estimate. Basis Len: smooths the midpoint. They do different jobs and do not need to match, but wildly different values create a visibly asymmetric envelope that experienced readers can misread as a skew.

Blended BBWP

This group shows oscillator-width percentile rank as columns at the base of the pane.

What BBWP here is actually measuring. BBWP on this pane ranks the width of the blended oscillator against its own recent history. It is not the BBWP you may know from price. A low column means the oscillator has been compressed recently. A high column means the oscillator has been wide recently. That is a statement about the oscillator, not about price. Compression of the oscillator can precede expansion of the oscillator; it does not necessarily precede expansion of price.

Length: vs Lookback:. These are the two most confusable BBWP controls. Length: is used for both the Bollinger basis window and the standard-deviation window — they are kept matched on purpose, because separating them would break the Bollinger construction. Lookback: is the rank horizon: how far back the current width is ranked against. Shortening Lookback: makes the rank noisier. Lengthening it requires more chart history.

The threshold is cosmetic. Changing Threshold: changes the colour of the columns. It does not change the percentile the column represents. A reader who raises the threshold to "get more green columns" is not making the underlying compression any more real.

Non-SMA basis choices. The default Basis Type: is SMA, which is classic Bollinger. Other choices will change the basis smoothing but will no longer be classic Bollinger. That is allowed; just know what you traded.

Blended Donchian

This group wraps the blended fast line in a Donchian corridor — the oscillator's own recent highs and lows.

This is a Donchian on the oscillator, not on price. When the blended fast tags the upper Donchian line, the oscillator has reached its recent high. Price may or may not have reached its recent high at the same time. The two events are related but not equivalent, and reading an oscillator-Donchian break as a price-Donchian break is a trap that looks convincing.

Default Basis Len: of 1. With the default, the basis line traces the direct midpoint without any extra smoothing. Raising this only smooths the midpoint drawing; it does not change the upper or lower Donchian lines.

Slots (MA 01 through MA 05)

STR runs five independent moving-average setups. Each slot has the same control surface. The per-slot Power User knobs live under each slot's matching PU group and only bite when the chosen MA type uses them.

Per-slot defaults, at a glance

What each control does

• Enable MA nn — master switch for the slot. Disabled slots do not plot, do not fire alerts, and do not contribute to the blend. An enabled slot with a non-zero weight contributes; an enabled slot with weight zero stays alive for plotting but does not steer the blend.

• Hide MA nn Plot — hides the slot's fast line on the pane while leaving the slot computing. This is the trap that catches people: a hidden slot with weight still contributes to the blend. If the blend suddenly looks different from what you expected, check whether a hidden slot is doing work you forgot about.

• Source: — the price series the slot reads from. The default is close. When a Ticker: override is set, this source is read on that ticker.

• Timeframe: — the higher timeframe the slot requests. Blank means use the chart timeframe. This value must be greater than or equal to the chart timeframe or the indicator stops with a runtime error naming the slot.

• Length: — the baseline MA lookback.

• Type: — the baseline MA style. EMA is the default; other choices bring in their matching Power User knobs.

• Slow Length: — the length of the slow line applied to the slot's fast oscillator. This is not an MA on price; it is an MA on the already-normalized 0..100 oscillator.

• Slow Type: — the style of the slow MA.

• Line Width: — cosmetic.

• Weight: — how much this slot steers the blend. Weights do not have to sum to 100; the blend normalizes across active slots.

• Ticker: — optional symbol override. Empty uses the chart symbol. When set, the slot's source is read on that symbol.

Per-slot On Bar Close?. This control lives under the per-slot Power User group for each slot. It is the single most important repaint control in the indicator. Default is true, which means each slot waits for its higher-timeframe bar to close before reporting a new oscillator value — safer, one higher-TF bar of lag. Set to false and the slot uses the live higher-timeframe bar value, which can shift until that bar closes. The MTF and Repainting page walks through exactly what each side buys you.

Weights and the blend. Start with equal weights. A reader who moves one slot to 100 and the others to 0 effectively turns STR into a single-slot indicator with structural extras drawn around that one slot. That can be what you want, but the "blend" label is no longer doing the work the name implies. Make that choice deliberately, not by accident.

Cross-ticker slots. Setting a Ticker: override is a powerful tool and a quick way to mislead yourself. Used well, it lets you ask whether a related symbol's higher-timeframe read agrees with the current chart's. Used badly, it imports conviction from a correlated-but-different instrument into the blend of the symbol you are actually trading. If you use cross-ticker slots, know what question you are asking. The Workflows page covers a sanity-check routine for this.

Power User groups

Each slot has a Power User group (MA nn PU). The blend has its own Power User groups for Smoothing, Keltner, BBWP, and Donchian. The controls inside each Power User group are only consulted when the matching MA type has been chosen somewhere.

The same table applies to the Blend Smoothing PU, Blend KC PU, Blend BBWP PU, and Blend DC PU groups, each governing the basis type chosen in their matching extras group.

Treat these as conditional. A reader who tunes Jurik Ph: while still on an EMA slot type is tuning a dead knob. The setting is not wrong; it just does nothing until the Jurik MA is actually in play. The inputs exist as a group to keep the interface predictable, not to imply they all matter at once.

Settings interaction map

Three interactions are worth keeping in mind whenever you adjust settings in isolation. These are the pairings that most often produce surprising behavior.

Sensitivity × smoothing. If you crank ATR Sensitivity until the oscillator saturates against 0 or 100 and then enable Enable Smoothing to quiet the result, you have masked the saturation. The smoothed line will look clean and will stop telling you anything inside the body of moves. Rule of thumb: set sensitivity on defaults first, enable smoothing second only if you are sure the noise is not coming from saturation.

Weight collapse. Setting one slot's weight at 100 and the others at 0 is legal. The blend label still prints. What you have built is not a blend — it is a single slot with the structural extras drawn around it. That is sometimes exactly what you want; just name it internally. If you are leaning on a "blended" alert in that configuration, understand that the alert is reading a single-slot state.

BBWP length vs BBWP lookback. Length: is the Bollinger construction window. Lookback: is the rank horizon. A reader who changes Length: hoping to control "how far back BBWP looks" is tuning the wrong knob. For the rank horizon, change Lookback:. The two are not interchangeable and neither hides the other.

A fourth interaction worth mentioning: KC Length vs Basis Len: inside the Keltner group. They smooth different things — the range estimate and the basis midpoint — and extreme asymmetry between them is the most common source of "why does my Keltner look lopsided." Keep them close unless you have a reason.

When defaults are the right answer

Most of STR's defaults were chosen to draw a legible pane on a wide range of instruments without any tuning. If you cannot name, in one sentence, what the current pane is failing to tell you, leaving the knobs alone for another session is usually a better investment than changing something. The test for whether a tuning move is disciplined is whether you can state the question you were trying to answer before you moved the knob — and whether the change you made is consistent with that question.

The two tuning patterns to watch for in yourself: tuning toward a prettier chart (the pane looked cleaner afterwards, but you cannot say what you learned), and tuning after a losing session (the setting did not cause the loss; changing it to recover dignity does not address what did). Neither is uncommon, and both produce configurations that drift away from a pane you can actually study.