Troubleshooting

A symptom-first page. When something on your chart looks wrong, start here and work back to cause. Each symptom below is paired with the most likely causes in order, a fix for each, and — where it matters — a note on whether the "problem" is a setup error, a misunderstanding of what the indicator is reporting, or a genuine product limit.

Those three categories are worth naming before you dig in, because they ask different things of you.

• Setup error. You configured something in a way that is not serving your intent. The fix is a configuration change.

• Misunderstanding. The indicator is reporting exactly what it says it reports; your expectation and the indicator's behavior are out of sync. The fix is a reading change.

• Product limit. The indicator does not do the thing you want. The fix is to build or find something that does, or to accept the limit.

If you go through this page and your symptom is not here, the most productive next step is the Visuals and Logic page or the MTF and Repainting page, depending on whether the confusion is about what you are seeing or about when you are seeing it.

Symptom: the pane is empty

You applied STR and the pane is blank. No lines, no columns, no triangles.

Most likely cause: the Axiom Moving Average Library (Pro) did not load.

STR imports the Pro MA library for every baseline, slow line, and basis filter. If your account cannot resolve that library, the indicator cannot compute anything and the pane stays blank. This is not the indicator silently failing — it is a load failure that looks like silence.

• Category. Setup error.

• Fix. Confirm on the TradingView indicator listing that the library is available to your account. If you recently re-loaded the indicator, remove it and reapply. If the library genuinely will not load, contact support; STR cannot run without it.

Second cause: the chart has no history.

If you are looking at a replay from the very first bar, or an illiquid symbol with almost no trading history, none of the indicator's lookbacks can compute.

• Category. Setup error.

• Fix. Move to a chart with confirmed historical bars, or advance the replay past the longest lookback on the indicator (the BBWP lookback is the longest by default at 252).

Third cause: Plot Blend is off and every slot has Hide MA nn Plot on.

You can legally end up in a configuration where nothing is being drawn even though everything is still computing.

• Category. Setup error.

• Fix. Turn Plot Blend back on, or un-hide at least one of the active slots.

Symptom: a runtime error appears naming a specific slot

TradingView displays a runtime error with a message referring to a specific slot — something like "MA 02 requires a timeframe greater than or equal to the chart timeframe."

Cause. A slot's Timeframe: is set to a value lower than the chart timeframe. STR hard-stops in that configuration because a higher-timeframe request cannot honestly be served from a lower-timeframe chart context.

• Category. Setup error.

• Fix. Open the named slot and set its Timeframe: to the chart timeframe or higher. Alternatively, leave the slot's timeframe blank to use the chart timeframe.

This error is not a bug and it does not need a workaround. It is a guardrail. The moment you want a lower timeframe than the chart, you are asking for behavior STR does not provide, and the right move is to change the chart, not to fight the error.

Symptom: the blended fast is glued to 0 or 100

The blended fast (and possibly several per-slot lines) spend long stretches flat against the pane floor or ceiling.

Cause. ATR Sensitivity has been pushed past the useful range. The oscillator's bounded mapping saturates when the input distance gets large enough, and above the useful range it spends most of its time in the saturated region.

• Category. Setup error, with a shading of misunderstanding. The indicator is doing what the setting told it to do.

• Fix. Drop ATR Sensitivity back toward 1.0. Confirm the pane comes back to life. If you have been running a high sensitivity for a while and the pane now looks less exciting at 1.0, that is correct; the excitement was saturation, not information.

If sensitivity is already near 1.0: check ATR Length. A very short ATR length can amplify distances on an instrument with rare outlier moves, producing occasional saturation. Try 14 (the default) first.

Symptom: a slot line is the wrong colour, or stops updating

One of the per-slot lines looks dimmed all the time, or stopped moving, or is a colour you did not expect.

Cause 1 — the slot's fast is below its slow. Each slot's colour modulates by transparency when fast is below slow. A permanently dim line is a slot that has been bearish for a while.

• Category. Misunderstanding.

• Fix. None. This is the colour convention; your reading of it was different from its definition.

Cause 2 — the slot is disabled but its line is cached on the chart. Rarely, a previously enabled slot that has been disabled can leave a ghost line until the chart is reloaded.

• Category. Setup error if disabled, rendering quirk if truly cached.

• Fix. Reapply the indicator or reload the chart.

Cause 3 — the slot's timeframe has not populated yet. A higher-timeframe slot takes time to report a first value on a fresh chart. During that warm-up, the line either does not draw or holds at its initial na.

• Category. Product behavior, not a defect.

• Fix. Wait for enough bars to pass for the slot's timeframe to populate. On a 240-minute slot applied to a chart that just loaded, this can be a long wait.

Symptom: divergence triangles are printing later than I expect

Triangles appear several bars after what you recall as the pivot in price.

Cause. With Plot On Pivot? off (the default), the triangle prints on the confirmation bar — Pivot Len: bars after the pivot itself — because that is when the pivot is actually confirmed.

• Category. Misunderstanding.

• Fix. Either live with the confirmation-bar placement (which is the more honest posture because the divergence was not knowable until the confirmation bar), or enable Plot On Pivot? to back-offset the triangle visually to the pivot bar. In either case, the alert fires on the confirmation bar — that has not changed.

A deeper walk-through of this behavior, including why "earlier" plotting does not mean earlier information, is in MTF and Repainting.

Symptom: divergence triangles are printing where I do not see an obvious pivot

A triangle prints on the pane but scanning price you cannot find the pivot that produced it.

Cause 1 — Pivot Len: is short. At Pivot Len: of 20 (the default), pivots in choppy data can be shallow and easy to overlook when you scan price visually.

• Category. Misunderstanding plus setup choice.

• Fix. If you want fewer, stronger divergences, raise Pivot Len: to something like 30 or 40. Live with the change before you decide.

Cause 2 — the divergence is between price and the blended fast, not between price and price. A divergence does not require a large price pivot; it requires a pivot that the blended fast disagreed with. Shallow pivots can still produce divergences.

• Category. Misunderstanding.

• Fix. A reading change. Divergence is a two-input check; expecting it to align with visually obvious price pivots is the wrong expectation.

Symptom: BBWP columns never paint

The blended fast is visible, the blend is working, but the BBWP columns are empty.

Cause 1 — insufficient history. BBWP defaults to a 252-bar percentile lookback. Until 252 confirmed bars of blended-fast widths exist, BBWP cannot rank anything and correctly refuses to draw.

• Category. Product behavior, not a defect.

• Fix. Load more history, or reduce Lookback: to a smaller value if you are willing to accept a noisier percentile rank over a shorter horizon.

Cause 2 — Show BBWP is off. Check the BBWP group.

• Category. Setup error.

• Fix. Enable Show BBWP.

Cause 3 — the blended fast itself is na. If every slot is na on the current bar, the blended fast is na and BBWP has no width to compute.

• Category. Warm-up behavior.

• Fix. Wait for slot warm-up to complete.

Symptom: the blend looks like it is tracking a single slot

The blended fast is moving almost identically to one of the per-slot lines, even though you expected several slots to be influencing it.

Cause 1 — weight collapse. One slot's Weight: is at or near 100 and the others are at or near zero. The "blend" is that single slot.

• Category. Setup error if unintended, deliberate configuration if intended.

• Fix. Review weights and redistribute toward 33.3 / 33.3 / 33.3 if the intent was a blend.

Cause 2 — hidden slot contribution. A slot you forgot about, with Hide MA nn Plot on, is contributing heavily to the blend while its line is invisible.

• Category. Setup error plus misunderstanding about hidden-slot behavior.

• Fix. Un-hide every enabled slot and verify which slots actually have non-zero weights.

Cause 3 — two slots warming up, one slot populated. On a freshly loaded chart with a 60-minute slot and a 240-minute slot, the 240-minute slot may still be na. During that window, the blend is effectively the two populated slots. This reads as "the blend is tracking one timeframe" until the long slot catches up.

• Category. Warm-up behavior.

• Fix. Wait for the slowest active slot to populate.

Symptom: an alert did not fire when I expected one

You set up an alert, the condition visibly appeared on the pane, and nothing fired.

Cause 1 — the condition was mid-bar, not confirmed. Every alert is gated on a confirmed bar. A cross that happened intrabar and reverted before close does not fire.

• Category. Product behavior.

• Fix. Nothing. This is the honest alert posture.

Cause 2 — for divergence alerts, Show Div is off. Divergence alerts are gated on the visual enable. With the visual off, the alert is silent.

• Category. Setup error.

• Fix. Enable Show Div.

Cause 3 — for alignment alerts, a slot is not populated. All MA Osc Slots Bullish requires every active slot with a current value to be bullish. Warming-up slots are silently excluded from the count. It is possible the alignment count you believed was met was in fact met on a smaller population than you intended.

• Category. Product behavior, worth understanding.

• Fix. Confirm every enabled slot has a current value before trusting alignment. If a slot is chronically na on your chart, consider disabling it so the alignment count reflects only slots that actually contribute.

Cause 4 — the alert-manager frequency is not "once per bar close." Some frequency settings can fight STR's confirmed-bar gate.

• Category. Setup error in the alert manager.

• Fix. Use once-per-bar-close.

Symptom: an alert fired and I disagree

An alert fired and, looking at the chart, you do not believe the condition was true.

Cause 1 — you are reading the live bar, not the confirmed bar. Scroll back one bar and look at the state at close. If the state was true at close, the alert was right.

• Category. Misunderstanding.

• Fix. Re-read at the confirmed bar.

Cause 2 — for divergence alerts, the marker and the alert have different timings. With Plot On Pivot? on, the triangle is drawn at the pivot bar but the alert still fires at the confirmation bar. The visual and the alert are not on the same bar in that configuration.

• Category. Misunderstanding.

• Fix. Recognize that the plot position is cosmetic. The alert bar is where the divergence was actually known.

Cause 3 — for alignment alerts, a slot you thought was active was in fact disabled or na. The alignment count reflected the smaller population.

• Category. Setup error plus misunderstanding.

• Fix. Confirm every slot you believed was active is enabled and has a current value.

Symptom: the Keltner envelope is pancaked against the pane edge

The upper Keltner band is flat against 100 (or the lower band is flat against 0) and stays there.

Cause. KC Mult: has been pushed past the useful range. The envelope is clamped into the pane on purpose; when the mathematics wants to go outside 0..100, the clamp flatens the band along the boundary.

• Category. Setup error.

• Fix. Lower KC Mult: toward 2.0. A flat-against-the-edge band is reporting nothing, and the configuration is uninformative in that direction until the multiplier is reduced.

A secondary cause: a very short KC Length on a very wide oscillator can also produce edge-clamped behavior at moments of rapid expansion. Try a length closer to 20.

Symptom: the pane looks fine, but my alerts are firing constantly

You wired STR into an alert workflow and are receiving what feels like too many alerts.

Cause 1 — per-slot or blend alerts replay every confirmed bar where the state is true. Blended MA Osc Is Bullish fires on every confirmed bar the blend is bullish, not only on the cross. If you wanted "fire on cross only," that is a different alert you have to build.

• Category. Misunderstanding.

• Fix. Either accept the repeating behavior, or wrap the alert with a cross-detection pattern in your charting platform.

Cause 2 — you have many alerts enabled at once. Five per-slot bullish alerts plus a blend alert plus two alignment alerts is nine potential firings per confirmed bar.

• Category. Setup choice.

• Fix. Disable the alerts you do not need. A common pattern is to keep only the blend and divergence alerts, and let the per-slot alerts be for diagnostic work rather than active notification.

Symptom: something else feels wrong and I cannot name it

Not every problem fits a clean symptom. When the pane feels wrong and you cannot say why, three moves tend to resurface what is happening:

1. Reset to defaults. Remove STR, reapply with defaults, do not tune. If the pane now looks right, your configuration is carrying the confusion.

2. Read the pane layer by layer. Use the Visuals and Logic page as a script. If you can walk from back to front — Keltner, BBWP, Donchian, per-slot, blend, divergence — and name what each is doing, you have diagnosed the pane. If one layer refuses to be named, that is your starting point.

3. Check the repaint posture. If a previously stable chart is suddenly whippy, someone (possibly you, on a different session) may have flipped On Bar Close? off somewhere. The MTF and Repainting page walks the cost of each side of that switch.

If after those three moves the problem persists, the most useful note you can bring to support is a short description of what the pane is showing, what you expected, and which steps above you have already run. That triage saves a round of back-and-forth and lands the conversation directly on whatever the real issue is. Screenshots help. A one-line description of the instrument and timeframe helps. A note about recent configuration changes helps more than almost anything else — "it was drawing correctly yesterday and today it is not" narrows the search space dramatically when paired with what you touched in between.

One closing note. The most common cause of a symptom that does not appear on this page is a misread of what the indicator is reporting rather than a defect. Before escalating, it is worth re-reading the Visuals and Logic page with the current behavior in hand and asking whether the pane is in fact doing exactly what the page describes — just not what you expected. Misreads outnumber defects by a wide margin, and catching one at the desk is faster than catching one through a support round-trip.