Workflows

Three scenarios in the same card form. Each card opens with a setup, names the reading routine, names the failure mode, and lays the anti-pattern contrast next to the scenario so you can feel the shape of the misuse against the shape of the use. The closing section is a cross-cutting habits block you can keep after the specifics of these scenarios stop mattering to you.

None of these is a trade recipe. They are reading routines built around the pane, not entry triggers. The tool does not issue entries; you do.

Scenario 1 — Differentiated three-slot stack

A deliberate three-slot configuration where each slot has a job. The defaults are a starting point; this scenario is the configuration you grow into after the defaults stop teaching you anything new.

Setup

• Slot 01 — fast intra-day. Timeframe 5m. RSI length 9. RSI smoothing SMA 3. Signal SMA 3. Weight 33.3. Line width 2. Source close.

• Slot 02 — chart anchor. Timeframe 15m. RSI length 14. RSI smoothing SMA 5. Signal SMA 5. Weight 33.3. Line width 2. Source close.

• Slot 03 — patient higher-timeframe. Timeframe 60m. RSI length 21. RSI smoothing EMA 5. Signal EMA 5. Weight 33.3. Line width 2. Source close.

• Master smoothing off. On Bar Close? = true. Reference lines at 30 and 70.

The three slots differ across RSI length, smoothing length, and — on slot 03 — MA family. They are genuinely three different reads of three timeframes of the same symbol, not three copies of the same read.

Reading routine

On each chart bar you pay attention to:

1. Scan slot colors. Look for agreement or disagreement between the fast, anchor, and patient slots. The fast slot will flip first on most moves; the anchor slot follows; the patient slot holds longest.

2. Check the fast slot's value. A fast slot at 60+ full-tone during a trending session and a fast slot at 60+ faded during a stalling session tell you different things about the same value band.

3. Check the blend color and fill. The blend leans on the three slots at equal weight; when the slots genuinely disagree, the blend quiets toward 50 and the fill narrows.

4. Cross-check against the chart. Price and structure are the context the pane is reading in; the pane does not replace that read.

Failure mode

The most common failure on this configuration is reading the three slots back into "three votes" after you differentiated them. Differentiation reduces autocorrelation; it does not eliminate it. A sharp move on the underlying still propagates into all three slots — the fast slot first, the anchor slot next, the patient slot last. Alignment on a differentiated stack is more informative than alignment on an identical stack, but it is still not three independent observations. Treat alignment as "the cadences agree," not as "three independent opinions agree."

Under stress — a fast session, a news bar, a low-liquidity window — the fast slot will hop harder and the color will flicker. The anchor slot is your steadying read in that window. If you find yourself making decisions off the fast slot alone during stress, you have collapsed the configuration back into a single fast slot with visual companions. A specific self-check: when you notice yourself wanting to disable or hide the anchor and patient slots "because they are lagging the move," that is the moment to keep them on. The lag is the point; the anchor is your argument against reacting to the fast slot's every flicker. Turning them off during stress turns the differentiated stack into a fast-only stack exactly when you needed the other two most.

Anti-pattern contrast

Three slots at identical settings — same source, same 14-length RSI, same SMA 3 smoothing pair, same SMA 3 signal, only the timeframe varying. Superficially this looks like three timeframes reading the same instrument. Functionally it is one measurement sampled at three cadences. Alignment holds for long runs on trending instruments and almost never fires on choppy ones; the alignment alerts become a trend-persistence detector rather than a cross-cadence agreement read. In Base, each slot is a third of the alignment vote, so a single mis-differentiated slot has a larger impact on the alignment quality than it would on a ten-slot trim.

The difference between the differentiated stack and the identical stack is the difference between "three slots each answering a slightly different question" and "three slots all answering the same question more or less slowly." Build the differentiated stack.

Scenario 2 — Weight-zero observer with two voters

A configuration that keeps one slot visible and alerting without letting it steer the blend. Useful when you have a higher-timeframe slot you want to see and react to, but you do not want its weight pulling the blended read.

Setup

• Slot 01 — fast voter. Timeframe 5m. RSI length 14. RSI smoothing SMA 3. Signal SMA 3. Weight 50. Source close.

• Slot 02 — anchor voter. Timeframe 15m. RSI length 14. RSI smoothing SMA 3. Signal SMA 3. Weight 50. Source close.

• Slot 03 — observer. Timeframe 60m. RSI length 21. RSI smoothing SMA 5. Signal SMA 5. Weight 0. Hide Plot off. Per-slot alerts wired on bullish and bearish state.

• Master smoothing off. On Bar Close? = true.

The blend is a weighted average of slots 01 and 02 only. Slot 03 plots, runs its pipeline, evaluates its state, and fires its own bullish and bearish alerts — it does not pull the blend.

Reading routine

1. Read the blend against slots 01 and 02, both of which contribute at equal weight.

2. Read slot 03 as context. A slot 03 in full-tone while the blend is also lime is a higher-timeframe regime that aligns with the faster voters; a slot 03 in faded while the blend is lime is a higher-timeframe warning that the faster voters have not yet caught.

3. The slot 03 alerts fire on confirmed chart bars. Use them as context notifications rather than as trade signals.

Failure mode

The alignment alerts (All RSI Slots Bullish and All RSI Slots Bearish) count slot 03 because it is enabled. Alignment is satisfied when all three slots lean the same way — the observer's zero weight does not remove it from the tally. If you wired the alignment alerts on this configuration, a reader who thought slot 03 was fully out of the alert surface will be surprised when alignment fires from a two-voter blend plus a silent observer.

The failure mode is not a bug; it is the consequence of the design. Blend weight controls the blend. Enable controls everything else — plotting, per-slot alerts, alignment counting. If you want slot 03 fully out of every downstream surface, disable it. If you want it as a context observer with per-slot alerts, the zero-weight configuration is correct.

Anti-pattern contrast

"I set the weight to zero to turn the slot off." Zero weight is not off. Off is the Enable checkbox. A weight-zero slot still plots (unless Hide Plot is also on), still feeds per-slot alerts, and still counts toward alignment. Readers who use weight-zero as a blanket disable will end up confused by the slot's continued visibility, continued alerting, or continued presence in the alignment tally. Call the thing by its function: Enable is the power switch, weight is the vote share.

Scenario 3 — Single-slot first session

A learning configuration, not a production one. You enable one slot, disable the other two, and use the pane with one RSI at one timeframe for a session or two. The goal is to install the color-first / value-second / blend-third reading order on the simplest possible instrument before you add complexity.

Setup

• Slot 01. Timeframe 5m. RSI length 14. RSI smoothing SMA 3. Signal SMA 3. Weight 33.3. Source close.

• Slot 02. Enable off.

• Slot 03. Enable off.

• Master smoothing off. On Bar Close? = true.

With one slot enabled, the pane reduces to: one slot line colored by its own RSI-vs-signal relationship, a blended RSI that equals the slot's RSI, and a blended signal that equals the slot's signal. The blend is not a weighted average of different slots — it is slot 01 shown twice in different line widths. That is the teaching simplification on purpose.

Reading routine

1. Read the slot's color. Full-tone or faded. Practice reading the color before reading the value.

2. Read the slot's value. Notice how the color and the value interact — where the color flips at the same value band, where the color flips at a different band, where the color holds across a wide value range.

3. Ignore the blend; it is redundant in this configuration. That is fine. You are here to install a reading habit, not to collect a composite read.

4. Run the configuration for a full session or two. When the color-first read becomes automatic, come back and add the other slots.

Failure mode

Treating this configuration as a production setup. The single-slot configuration loses the cross-timeframe evidence that is the whole point of the tool. It is a teaching configuration — run it to learn the reading order and then move on. If you stay on one slot because one slot feels cleaner, you are using the tool as a single-timeframe RSI, and a single-timeframe RSI is already available from other tools without this one's pipeline.

Anti-pattern contrast

"Start with all ten possible slots enabled at every default." This trim only has three slots, so the literal ten-slot version of this anti-pattern lives on CTX, but the shape of the mistake is the same: loading the full slot capacity before the reading order is installed means the reader is learning the pane while simultaneously processing three timeframes of evidence. The reading order fails to install and the reader compensates by reading the blend alone, which is exactly the habit the single-slot first session is trying to prevent.

Cross-cutting habits

Keep these after the scenarios above stop being the configurations you run.

• Read color before value on every session. The discipline is the single biggest habit this tool can teach, and it is the first one to erode under stress. If you find yourself reading values before colors during a fast session, that is the diagnostic. Slow down and re-read.

• Assume the blend alone is lying to you when the slots disagree. A quiet blend on three disagreeing slots is evidence about disagreement, not evidence about calm. If the blend is flat, look at the slots before you decide the pane is saying nothing.

• Write down every change you make to the defaults. A session journal is unglamorous and it is the fastest way to keep yourself from wandering back into an identical-three-slots configuration six weeks from now. Name the configuration. Name why you chose it. Name what you are watching for.

• Weight and Enable are not interchangeable. Enable is the power switch; weight is the vote share; Hide Plot is the visibility toggle. When you want something fully out of every downstream surface, use Enable off. When you want a blend-out but keep alerts, use weight zero. When you want a visual-only reduction but keep everything else on, use Hide Plot.

• When you want to understand what the tool is doing, flip On Bar Close? once and watch. The before/after on a 60m slot teaches you more about the tool in two minutes than any amount of documentation describing the behavior in words.

• Come back to this page after a month on the tool. The scenarios will read differently once the pane is familiar. The anti-patterns are the sections that age best; they name misuses your own habits will drift toward over time.

• When you catch yourself treating the blend as the authority, treat that as a diagnostic, not a preference. The blend is a composite you designed. If it has started settling arguments for you, the slots have stopped doing the work the tool was built to let them do. The correction is a reading move, not a settings change — read the slots first for a full session and let the blend move back into its headline role.

Where to go next

• The full honest-limits page, organized by damage potential — Limitations and Trust Boundaries.

• Every input as a tradeoff — Settings.

• The reading order in depth — Visuals and Logic.

• When the configuration you built looks broken — Troubleshooting.