Correlation Is Not Causation

Sync recommends. A named human decides. On the Decision Case, co-movement is a report about timing. It is not the record that one condition caused another, and it is not the record that an action is authorized.

Two tags rise on the same screen. Two alerts land in the same hour. A dashboard trend turns down while a second trend turns up, and the meeting treats the pair as the cause. Nobody wrote the check that one produced the other. Nobody stored the measurement that would show the link. Nobody closed a verified outcome against an authorized change. The Decision Case still has to show the inspection, the reading, or the verification that was actually recorded. Moving together is not that line.

This essay follows Confidence Is Not Evidence and stands beside Assumption Is Not Evidence. Confidence keeps a strong gut feel, a high model score, years of tribal certainty, or a manager's certainty from being stored as a recorded check or a verified outcome. Assumption keeps a belief, a habit, a prior shift's story, or an unverified model output from being stored as a recorded fact. This essay is the chart people treat as the cause. A pair of lines that move together does not move the statement into the observed-fact column. Correlation is not causation. A labeled assumption that two signals travel together is still an assumption. Being sure about the pair does not record the cause.

Facts come before assumptions. The same gap has three neighboring forms. Unknown Is Not Clear records missing evidence as unknown, not as clear or ready. Blank Is Not Zero keeps an empty field from being stored as a measured zero. Silence Is Not Clearance keeps no reply, no alert, and an empty inbox from being read as approval. Correlation is what the room uses to skip those records. Two tags that rose together do not record the missing evidence as clear. A trend drawn across a blank does not measure none. Coincident alerts do not write the decision the operator never recorded.

Orville Davis states the Decision Case order in Field Manual v0: Question, Evidence, Recommendation, Human decision, Action, Verification, and Learning. The manuals index lives at /manuals. Evidence is step 02. The Evidence chapter keeps a visible split between observed fact, assumption, and what is still missing. Correlation is not a fourth column that overrides the split. A proposal whose only basis is that two signals moved together has no causal basis a later reader can inspect. A proposal with no basis is not ready for a human decision. Honesty boundaries sit beside the spine: recommend is not authorize, evidence is required, and Stage-1 does not treat co-moving tags, coincident alerts, or dashboard trends as a live causal fact.

Correlation is not a recorded causal check

A recorded causal check is something a later reader can inspect on the case: the condition that was present, the change a named person authorized, the measurement taken before and after that change, and the verification that the authorized action did what it was supposed to do. Correlation is a statement about co-occurrence. It names that signals moved in the same window. It does not name which one produced the other, whether a third condition produced both, or whether either produced the outcome the question is about. Those are different records. The chart does not write the check.

The meeting can note the pair. Noting it records that someone saw the pair. It does not record the cause. Cause information is the check: what was observed, what was changed under a named decision, and what the verification showed. The co-movement stays in the assumption column, or it stays off the case as a display. Tightness of the fit does not change the column. A cleaner trend is still not a recorded causal check.

Sync may hold the pair beside a recommendation. Holding it does not authorize the work. The recommendation is a proposal. The person who can accept the operational consequence records accept, reject, escalate, or return. If the only support is that the lines moved together, the proposal is not ready for that act. The system does not auto-accept because the trends agree.

Co-moving tags are not the cause

Co-moving tags are two or more signals that rise and fall in the same window. A temperature and a vibration. A flow and a pressure. A motor current and a downstream level. The screen makes the pair look like an explanation. The pair is a timing report. Shared load, a common upstream, a season, a shift change, a clock that is off, or a third driver the case never named can move both tags without either being the cause the question needs. The operator can write that the tags moved together. The sentence does not identify which condition produced the failure, the delay, or the next action.

Stage-1 evidence is the record held on the case. A live connector that pulls historian or control-system tags sits outside this edition. A pair of tags that arrived without the records behind them is not a causal check. Simulated or seeded telemetry is a practice record. Co-moving practice tags are still practice tags. They are not a live plant fact, and they are not a substitute for the assumption the operator still has to label. Honesty Boundary Is Not Optional says Sync states what is simulated, seeded, or out of scope. Naming that limit does not promote the pair into observed causation. The Honesty boundaries chapter keeps Stage-1 from claiming a live result the edition does not hold.

If the case cannot show the asset configuration, the failure and maintenance history, and the human judgment that produced the line, the co-moving pair is a chart without a trail. Evidence Lineage Is Not Optional already requires that trail. Correlation does not shorten it. The pair can sit on the case as a display, labeled as a display. The label does not promote the pair into observed fact, and it does not promote it into a verified outcome.

Coincident alerts are not the decision

Coincident alerts are notifications that fire in the same window. Two threshold breaches. An anomaly flag beside a pager page. A red tile and a silenced channel that landed together. Seeing both, acknowledging both, or clearing both is not a Decision Case and not a decision. Alert Is Not Decision already keeps the alert off the decision. Coincidence does not put it back. Two alerts in one hour do not record that one caused the other. They do not record that either caused the outcome the question is about.

Clearing the pair, muting the channel, or auto-closing the ticket must not be read as authorization, verification of outcome, work-order execution, or plant control. The operator can note that the alerts were coincident. The note is a timing report. It is not accept, reject, escalate, or return. It is not the check that nobody performed. Sync may emit alerts and recommendations. A named human still opens the case on a question, cites approved evidence, and decides.

Dashboard trends do not authorize action

Dashboard trends are the lines on the display: a falling health score, a rising count, two series that cross, a week that looks worse than the week before. The display can surface that something changed in the window the chart covers. Dashboard Is Not Decision keeps that display off the decision. A trend that moves with another trend is still a display. Green, a crossing pair, or a steep slope does not authorize the work, verify the outcome, write the work order, or control the plant.

The operator reads the trend as a report about the chart. The operator does not read it as the causal check. A recommendation drafted from the trend is still a recommendation. Sync recommends. Humans decide. The named person accepts, rejects, escalates, or returns against evidence a later reader can inspect. If the only support is the slope, the proposal is not ready for that act.

This essay states no OEM limit and no operating threshold. A trend that crosses a line the meeting invented is not a recorded causal check, and this page does not supply the line. The case holds the measurement that was stored, or it holds unknown. It does not hold a limit this essay made up so the chart would look decisive.

Unknown, blank, and silence stay what they are

Unknown stays unknown. Filling the unknown line with a co-moving pair, a coincident alert, or a dashboard trend does not change the state to clear. Bad or not-ready evidence stays blocked. A trend that looks related to the blocked item does not relabel blocked as ready. The operator must not treat unknown as go because two signals moved together. Go would mean the evidence is sufficient for the next recorded step. Co-movement is not sufficiency.

Blank stays blank. Zero is a verified measurement of none. A trend drawn through an empty field collects a line. It does not collect the value. Recovery figures that were left blank stay blank. They are not inferred from a chart that moved in the same week. A blank verification field is not a verified outcome of none, and a coincident alert does not close it.

Silence stays silence. No reply is not approval. An empty inbox is not clearance. A quiet dashboard is not controlled. A trend on a quiet screen is still a trend. It does not write accept, reject, escalate, or return. It does not record a check that nobody performed. Correlation about the quiet does not turn the missing response into the named decision.

Sync recommends. Humans decide.

Authorization is the named act, and the act still needs a basis the case can show. A drafted next action is a proposal. Recommend Is Not Authorize keeps that proposal off the decision. A proposal that cites co-moving tags is still a proposal. The chart can be tight. The model can be sure. Neither the chart nor the score accepts the operational consequence. Human Decision Is Not Optional says a named person accepts, rejects, escalates, or returns. Until that act is recorded, the case is still a proposal. Correlation does not authorize action. The system does not auto-accept because two alerts fired together.

Learning is what a later question is allowed to inherit. The closed case shows what was believed, what was authorized, and what was checked. Verification Is Not Optional says authorization records who decided, and the case stays open until the check is written down. Learning Requires a Verified Outcome says the later question inherits that verified outcome, not a hope. What was believed may include a labeled note that two tags moved together. Learning does not inherit the co-movement as if the check had confirmed the cause. A later crew that repeats the work because the last chart looked related, without the verification on the prior case, is inheriting a display. The Learning chapter publishes the closed record. It does not publish correlation as a result.

Co-moving tags, coincident alerts, and dashboard trends are not a recorded causal check. They do not authorize action. Sync recommends. A named human decides.

The neighboring boundaries still hold

This boundary is not the neighboring ones. Each of those essays still holds. None of them promotes correlation to a recorded cause.

Assumption Is Not Evidence is the statement. Confidence Is Not Evidence is the surety attached to the statement. Correlation Is Not Causation is the chart attached to both. A belief labeled as an assumption does not become evidence because two lines moved together. A high model score on that pair does not become a measurement. The label stays. The observed-fact column stays empty until a check, a measurement, or a verified outcome is on the case.

Unknown, blank, and silence are still the missing record. Correlation does not interpret them into clearance. A case that is unknown, blank, or quiet is not made ready by a chart that looks related. A related chart is not a measurement.

The check that closes the case is still verification. The trail behind a recommendation is still evidence lineage. An alert is still not a decision. A dashboard is still not a decision. A later case still inherits the verified outcome, not the correlation that traveled with an unverified story. Those records are either on the case or they are not. If they are not, co-movement does not supply them.

Where the public statement lives

Field Manual v0 is the public contents of the Decision Case. Start at the manuals index or open Sync Field Manual directly. Evidence is step 02. Observed fact, assumption, and what is still missing stay distinct. Correlation does not collapse that split. A proposal with no basis is not ready for the human decision that follows. Verification is step 06. A trend that moved after the work is not the check. Learning is step 07. A later case inherits the verified outcome, not the chart the last shift treated as the cause. Honesty boundaries keep the edition from treating co-moving tags, coincident alerts, or dashboard trends as a live causal fact. Later editions can deepen a chapter. The spine stays in this order.

What this article is not claiming

This is an essay about the record, not a customer case study. It names no plant, states no savings figure, and claims no prevented failure. It states no OEM limit and no operating threshold. It does not claim that Sync executes plant work, writes a work order, isolates equipment, or controls a plant. It does not claim that correlation is causation, that co-moving tags are a recorded causal check, that coincident alerts are a decision, or that dashboard trends authorize action. It does not claim that self-guided onboarding is a live product path.

Stage-1 readiness means a signed-in user can complete the Decision Case — question, evidence, recommendation, human decision, action, verification, and learning — and Field Manual v0 describes that journey. This edition does not describe plant execute, a live connector tag pull, SMTP invite delivery, or automatic revocation of access on expiry as live. It does not describe Sync writing work orders, isolating equipment, or controlling the plant. Simulated or seeded telemetry and assets are practice records. They are not live plant results, and co-moving tags on a practice screen are not a recorded causal check. Self-guided onboarding is not claimed as a live product path.

The series continues with Trend Is Not Truth, on why a rising or falling line, a co-moving tag cluster, a week-over-week delta, or a "we've been trending this way" narrative is not a recorded causal check and not a verified outcome. Companion reading: Confidence Is Not Evidence on why a strong gut feel, a high model score, or a manager's certainty is not a recorded check, Assumption Is Not Evidence on why a belief, a habit, a prior shift's story, or an unverified model output is not a recorded fact, Unknown Is Not Clear on why missing evidence is recorded as unknown, not as clear or ready, Blank Is Not Zero on why an empty field is not a measured zero, Silence Is Not Clearance on why silence, no reply, no alert, and an empty inbox are not authorization, Alert Is Not Decision on why seeing or clearing an alert is not a decision, Dashboard Is Not Decision on why a display is not authorization, Verification Is Not Optional on why the case stays open until the check is recorded, Evidence Lineage Is Not Optional on why a recommendation you cannot reconstruct is not an industrial decision, and Learning Requires a Verified Outcome on why a later case inherits the verified outcome, not a chart. A Reliability Assessment asks whether the records can support a conclusion. A Strategic Pilot is a governed proof around one operating decision. The Evidence chapter is the public statement of the split between observed fact, assumption, and what is still missing.