Spot gold, measured on real candles — and refused when it should be.
Gold moves on liquidity, not on trendlines. ORIN reads XAU/USD structure off the vendor’s own candles — the sweep, the zone, the break — puts the invalidation level down before it puts a size down, and tells you when the read does not stand up.
A real instrument, and a different one — an AMEX-listed ETF that keeps equity hours and gaps over the close.
Gold futures. ORIN refuses them and says why, rather than grading a chart it cannot measure.
No signals. No win-rate claims. Decision support only.
Four names for gold. One that ORIN prices.
Both panels are the product answering for itself — the real resolver, and the same market-hours function the engine checks before it will grade anything.
XAU/USD · Gold Spot / US Dollar · commodity
XAU/USD · Gold Spot / US Dollar · commodity
XAU/USD · Gold Spot / US Dollar · commodity
GLD · SPDR Gold Shares · stock
Gold futures need a paid futures data vendor (Databento/Polygon class). ORIN will not grade a chart it cannot measure. It suggests XAU/USD.
"NAS100" did not resolve to an instrument ORIN can price.
This is the same function the engine consults before it will grade anything: when the market is shut, a read is suspended rather than graded on stale bars.
No sample data. Nothing here is a worked example — it is the engine’s own output for six strings and two clocks.
What ORIN actually reads on gold
Every level below is geometry over candles. None of it is a model’s opinion about a picture.
ORIN is built on Smart Money Concepts primitives rather than a stack of lagging indicators, which matters on gold because gold is a liquidity instrument. But the honest version of that sentence is narrower than the one most of this category writes, so here is exactly what a XAU/USD read contains.
- The most recent liquidity sweep — a wick that takes out a prior pivot and closes back inside it. One sweep is cited, and only if it happened within the last six bars. A sweep from forty bars ago is history, not a reason.
- One live zone — the most recent order block or fair value gap that price has not yet mitigated. Not every zone on the chart. The one that is still open.
- The last break of structure or change of character, with the price it happened at. BOS is continuation; CHoCH is the first evidence the regime flipped. Gold produces convincing fakeouts of both, which is why they are labelled separately rather than merged into "a break".
- Clustered support and resistance — swing pivots bucketed by proximity and counted by touches. This is the level that has actually been defended, as opposed to the line you drew.
- Trend, momentum, volatility regime and where price sits against its own mean.
On gold today those measurements come back as a chart read — trend, structure, stretch and volatility, each as a sentence you can argue with — and not as a grade or a plan. Since engine 1.12.0 ORIN grades a setup and gives it an entry, stop and target only where that setup has beaten random entries in a registered test, and no gold setup has passed one yet. Where one has, in other markets, the same measurements feed nine weighted factors whose weights are published from the engine’s own table.
What happens when you upload a gold chart
You can hand ORIN a screenshot instead of typing a ticker. What it takes from that picture is narrower than you would expect, and the narrowness is the feature.
What the industry implies
UngroundedRead the picture, draw on the picture
A model looks at your screenshot and identifies the pattern
Support and resistance are read off your image
Boxes and arrows are drawn back onto the chart you uploaded
A stale screenshot produces a confident current read
Nothing in that flow can tell that the picture is from Tuesday.
What ORIN does
MeasuredRead the label, measure the market
The image is transcribed for one thing: which instrument and which timeframe
Volunteered levels are stripped before they reach the engine
If the transcription strays into market talk — a direction, a level, a pattern — the whole read is rejected rather than filtered.
Your image comes back unmarked, because nothing has read it
Structure is measured on the vendor’s candles and drawn on ORIN’s own chart
If your screenshot’s price axis disagrees with the live market, the read says so rather than annotating a stale picture.
The product’s own words, on screen after every screenshot read: "That is the only thing taken from the picture. Every level in the read below was measured on the vendor’s own candles, exactly as if you had typed the symbol."
Gold keeps the FX week, not the stock market’s
Spot gold is classed as a commodity here and trades the currency week: open Sunday 17:00 New York, closed Friday 17:00. That is not a cosmetic label — it decides when a read is live, when it is suspended, and how long a cached series stays valid.
It is also the cleanest way to see that XAU/USD and GLD are not the same instrument. At two in the morning on a Wednesday, spot gold is open and the gold ETF has been shut for ten hours. The panel at the top of this page runs both clocks side by side off the same function the engine consults.
Where session logic does appear in grading, it is a discipline filter and nothing more: a rule pack can require that the graded bar falls inside a clock window you named, in a timezone you named. It checks the bar’s own timestamp rather than the time you happen to be reading the chart, which sounds obvious and was not — the first version answered a question about entry time with the time the analysis ran.
A session on XAU/USD
You bring the context. ORIN measures what is on the chart when you ask.
- Pre-London
You mark the session. ORIN does not.
Where the Asian range sits, and which side has the resting liquidity, is your read — ORIN has no detector for it. What ORIN can tell you before London is what the structure on the chart currently is: the last break, the live zone, whether price is extended against its own mean. - London open
The sweep either happens or it does not
If it does, run the timeframe you actually trade. A sweep inside the last six bars is cited with the level it took, and the unmitigated zone left behind by the move is the one ORIN builds around. If nothing swept, the read will say the setup is a different one, or that there is not one. - Before you size
Read the invalidation before the entry
Where ORIN gives a plan, it names the structure the stop sits behind; on gold, where it gives none yet, the invalidation level is still named. If that structure is not one you would have used, the disagreement is the useful output — you have found the exact bar where your read and the measurement part company, which is worth more than agreement. - When it refuses
Take the refusal at least as seriously as the plan
No entry, no stop, no targets — and the invalidation level still on screen. Gold in a compressed pre-news hour produces exactly this, and standing down is the correct response to it. - After
Log it, and know what the log is for
A logged outcome enters your own record and, in aggregate, the platform’s. It does not tune a model against you. Calibration becomes personal only once you have resolved trades in the same confidence band, and the platform curve needs a hundred resolved outcomes before it steers anything at all.
Nothing in that session depends on ORIN knowing what time it is in Tokyo. It depends on you bringing the context and ORIN measuring what is on the chart when you ask.
How far a gold read actually sees
The number is two, and one of them is derived. Here is why, and what that costs you.
Gold is a metered instrument: its candles come from a paid vendor and every request is charged. Crypto is not, which is why crypto reads fetch their higher timeframes as separate series and gold does not.
On XAU/USD, a read analyses the timeframe you chose and builds one higher rung by aggregating those same bars. The rung is labelled as a roll-up in the engine and it is labelled as one here. It is genuinely useful — a 1H view assembled from 15m bars still tells you whether you are trading with or against the larger swing — but it is not a second chart, and it cannot see anything the analysed bars did not already contain.
Two consequences worth knowing before you rely on a gold read. The higher-timeframe conflict that can veto a crypto setup outright cannot fire here, because there is no independently fetched leg to disagree. And multi-timeframe agreement still carries weight in the score — that weight is being paid to a roll-up of the bars already being graded.
- 15m — fetched from the vendor
- 1H — aggregated, rolled up from the 15m bars
- 4H — fetched from the vendor
- 1D — aggregated, rolled up from the 4H bars
- 1D — fetched from the vendor
- 1W — aggregated, rolled up from the 1D bars
Nothing above the analysed series is fetched. The higher rung is a roll-up of the same bars, labelled as one by the engine, and it cannot see anything those bars did not already contain.
Invalidation first, then the stop, then the size
Gold’s range is not stable. A stop distance that was conservative in a quiet Asian session is reckless thirty minutes into the London–New York overlap. So ORIN never starts from a distance — it starts from the level that would prove the read wrong. That is how every ORIN plan is built; on gold, where no setup has passed a registered test yet, the level is shown and the plan is withheld.
Find the level that disproves it
Put the stop behind it
Let the size fall out
"No trade" is an output, not an error
A low read on gold is a legitimate result and it is the one most likely to save you money. ORIN treats it as a first-class answer rather than as a failure to produce one.
A read is refused when no setup on the chart has passed a registered test against random entries — today, every gold chart — and also when the reward does not cover the risk to the first target, when the grade lands at or below D, when a hard rule in the lens you applied has failed, when the data is more than three bars stale, or when the market is shut. Each refusal comes with the sentence that caused it, not a generic badge.
When ORIN refuses, the entry, the stop and the targets are withheld. Earlier versions drew a full trade plan underneath a "no trade" badge, which is a way of refusing that still hands over everything needed to ignore the refusal. The invalidation level is still shown, because that is structure and it is true whether or not anybody trades it.
What we can prove about gold reads, today
These cells are empty. That is the current true answer, and it is published rather than filled in.
The calibration a gold read carries today comes from replaying this engine over historical candles — simulated fills, not resolved live trades. Neither you nor the platform has a live record in these bands yet, and the read says exactly that rather than borrowing someone else’s number.
How the calibration worksWhat you'll actually use on XAU/USD
Three tools this page's argument depends on, each free to open.
Confluence Engine
The nine weighted factors behind the score, and what each one is measuring.
confluence enginePosition Sizing
Size a gold position from a stop distance in dollars per ounce, without an account.
position size calculatorSmart Money Concepts
How sweeps, order blocks and change-of-character are defined here, in code rather than in vibes.
smart money conceptsWhat this will not do for you
The section most of this category does not publish.
It does not predict price
ORIN reads structure that is already on the chart. It has no view on what happens next and no feed that would give it one — no economic calendar, no flows, no positioning beyond what the venue publishes. Anyone selling prediction is selling something else.
It does not send signals
No entries pushed to your phone, no channel, no copy trading, no broker or exchange connection. It holds no API keys and cannot place an order. You bring the chart, ORIN measures it, you decide.
It does not know your circumstances
ORIN analyses a chart. It does not know your account, your drawdown, your tax position or your risk tolerance, and it is not licensed to account for them. That distinction — analysis, not personalised advice — is the entire regulatory line, and it is why nothing here is a recommendation.
It will not publish a rate it cannot show you
There is a public accuracy page and it currently publishes a counter rather than a curve, because the resolved sample is far below the floor a published rate would need. We would rather show the counter filling up than fit a number to a sample that cannot carry one.
Gold moves on things no chart contains
Real yields, dollar strength and geopolitical risk decide whether gold is bid for the week, and no chart tool sees a CPI print coming. Structure tells you how the market is currently positioned and where price has already been defended. It does not tell you what the print says.
No gold setup has passed a test yet
ORIN gives a grade, confidence and plan only where a setup has beaten random entries in a registered test. No gold setup has, so a gold read is a description of how the market is positioned, not a verdict on a trade. On the replayed record the gold grade did not rank setups reliably below the 4H, which is exactly why it is not shown until a gold setup earns it.
How the checklist reads gold specifically — the nine factors, session-range measurements and two worked reads — is documented on the XAUUSD analysis AI page. Gold is metered, so a read sees one fetched timeframe. Bitcoin is not — BTC/USD chart analysis fetches two and lets them overrule the setup. If you are trading gold inside an evaluation, the drawdown maths matters more than any of this — prop-firm risk analysis covers it, and the position size calculator runs without an account.
Questions XAU/USD traders actually ask
For reading structure quickly and consistently, yes — it does not get tired, does not get attached to a bias, and does not forget to check what the last break of structure was. For predicting where gold goes next, no. Nothing does that, and gold’s macro sensitivity makes the claim especially hollow.
Judge on four things: does it measure structure from candles rather than pattern-match a picture; does it show its reasoning or just a verdict; does it publish outcome data you can audit; and does it stay on the analysis side of the advice line. ORIN was built against those four, and the fourth is checkable here rather than asserted.
No. Real yields, dollar strength and geopolitical risk decide whether gold is bid for the week. Chart structure tells you how the market is currently positioned and where price has already been defended. It does not tell you what CPI prints.
No, and deliberately. Your image is read for one thing — which instrument and which timeframe — and comes back unmarked. Every level is then measured on the vendor’s own candles and drawn on ORIN’s chart. If the screenshot is stale, the read tells you rather than marking up an out-of-date picture.
Two, and one of them is built from the other. Gold’s candles are metered, so the higher rung is aggregated from the bars being analysed rather than fetched as its own series. Crypto fetches its higher legs; gold does not. Pull the higher timeframe up yourself if it is carrying the decision.
No. XAU/USD is spot gold on the currency week. GLD is an AMEX-listed ETF on equity hours that gaps over the close. Both resolve, separately and on purpose. GC — gold futures — is refused outright with the reason, because ORIN will not grade a chart it cannot measure.
Free accounts get 4 full reads each day with no card — the chart read, levels, invalidation and reasoning, whole, and a plan wherever a tested setup appears (none on gold yet). Paid plans raise that ceiling, show the confidence and confluence dials on the verdict, and add outcome tracking against your own logged trades.