A lens is a betting opinion that agrees to keep score. It is not a slogan, a color treatment, or a fresh name for whichever side looks good today. The lens states its claim before the result, makes or withholds a call under a fixed rule, and carries the grade into the next comparable spot.
That last word does the heavy lifting: comparable. A record belongs to the claim in its exact context. It does not belong to the author, the page, the season, or the logo. A useful lens can miss on Sunday and remain useful. An ungraded lens can sound brilliant and still give the reader no reason to trust it.
The contract comes before the card
Every lens needs a claim that can be written without knowing the outcome. The claim names the sport, market, side of the comparison, information available at decision time, condition that creates a call, and conditions that produce no call.
A lens that fires on every game is often hiding weak boundaries. Real evidence usually applies to a narrower population. The no-call rule keeps the lens from wandering into situations its mechanism never covered.
The decision is frozen before the grading boundary. Final settlement then produces the grade under the market’s published rules. A push stays a push. A void stays a void. An excluded event does not become eligible because its result happened to support the thesis.
The interface can compress that contract into a card, but it cannot replace it. Behind the card, the claim version, decision timestamp, market snapshot, eligibility result, call, final grade, and exclusion reason must remain inspectable.
A lens is not a system and it is not a pick
A system is the machinery. It collects prices, transforms inputs, fits or applies models, ranks choices, and publishes decisions. One system can support several lenses because it can ask several distinct questions of the same board.
A pick is one terminal decision on one event, market, and moment. It can be graded, but it has no reusable memory by itself. The reader sees the answer without the stable rule that produced it.
The lens sits between those layers. It is narrower than a system and more durable than a pick. It says: when this defined situation appears, this is the direction of the claim, this is when the call was made, and this is how every eligible prior call was graded.
That structure turns a stream of picks into a falsifiable idea. Without it, a feed can rotate explanations after every game and still look consistent.
Context is part of the score
A record without context mixes unlike decisions. A spread call made against an opener is not the same claim as a spread call made near close. A side lens is not a total lens. A pregame read is not an in-play rule. A familiar name can remain while the underlying question changes.
The scoreboard therefore travels with a context key. That key can include sport, market, decision window, competition phase, eligibility filters, The purpose is not taxonomy for its own sake. It is to stop strong evidence in one setting from laundering a weak claim in another.
Week One makes this discipline visible. New roles, new combinations, and little current-season evidence pressure every prior assumption. A lens may still apply, but the page should show whether its evidence comes from the same context or an adjacent one.
When the application falls outside the proven context, say so before the pick is read. The label should not force the reader to discover the mismatch in a methodology drawer.
The scoreboard has a required grammar
For a side or total lens, the public record should be ATS wins-losses, ATS win rate, evaluation window, and sample size. Those pieces answer what was graded, how often it covered, when the evidence was collected, and how much evidence exists.
Units and ROI are not substitutes. They depend on stake policy, price assumptions, and bookkeeping choices that can move independently of the directional claim. A streak is not a substitute either. It selects one part of the sequence and invites the reader to ignore the rest.
Show misses where wins are shown. If results cluster by decision window, market state, or roster condition, surface that split. A lens card that leads with its best slice and buries the losing contexts is an ad, not an audit.
The current figure belongs on the live surface that owns the grades. This explainer does not freeze a performance number into editorial copy that will outlive the sample.
The four evidence gates
GRADING means the claim is live and eligible calls are being settled, but the evidence is not yet strong enough for the highest public role. That is the honest state for a new claim, a changed claim, or an old claim entering a materially different context.
PROVEN means the lens has cleared the product’s evidence standard in the displayed context. The gate should depend on settled calls, sample size, uncertainty, and stability across the evaluation window. It is not a lifetime title.
FEED_ONLY means the source contributes information without making a clean, gradeable call. Injury news, price feeds, and attributed observations can be valuable without choosing a side. Keeping them feed-only prevents coverage from posing as forecasting skill.
DECAYING means the lens has prior evidence but current matched grades are weakening. Market adaptation, data drift, a changed decision process, or context shift can all produce that state. The history remains visible while the old weight is withdrawn.
Promotion and demotion use the same standard
A point estimate is not enough to promote a lens. Small samples jump after one result, and wide uncertainty can support both a useful effect and no effect. The gate needs evidence that the claim separates from its baseline under the same rule that produced the calls.
Demotion is not punishment. It is how the product stops old confidence from surviving new evidence. A lens can move from PROVEN to GRADING, from GRADING to FEED_ONLY, or out of a recommendation role when its current matched sample no longer supports the display.
The reverse matters too. A recent streak should not trigger promotion. The claim must stay fixed, every eligible call must be counted, and the settled sample must support the change. A gate earned by rewriting the definition after each miss is branding, not evidence.
Promotion and demotion should be versioned events with reasons. The reader should be able to see what evidence changed and which context the new gate governs.
How to read a lens on a live slate
Start with the gate, then the context, then the record. A PROVEN label in a mismatched setting should carry less weight than a GRADING label whose current application matches the stated claim exactly.
Check whether the lens made a call or correctly stayed silent. A no-call is part of the contract when evidence is missing, provenance is not live pregame, or the available price falls outside the decision rule.
Then read the misses. The useful question is not whether the lens has ever been wrong. It is whether losses cluster around a condition the current card also contains. A lens that weakens after late market moves should not be read the same before and after one.
Finally, separate the lens from the wager. A sound directional claim can arrive at an unattractive price. The lens grades its claim under the frozen contract. Acting still depends on the current market and the bettor’s constraints.
Changes create a new version, not a rewritten past
An auditable lens preserves its claim version, decision timestamp, market snapshot, When the rule changes, the version changes.
The child can keep lineage to the parent. It cannot silently inherit the parent’s scoreboard. Filters, thresholds, new inputs, and new decision windows all change the population being graded.
This is why a lens is more than a card treatment. The interface may reduce the state to a gate and a short record line, but the contract underneath has to survive inspection. The page earns trust by making the rule and the misses easier to find, not by finding a stronger adjective.
The standard
Define the claim. Freeze the eligibility rule. Use only information available at the decision point. Grade every eligible call. Publish ATS wins-losses, win rate, evaluation window, and sample size. Expose the misses. Let the evidence control the gate.
Current records belong on the track record. Live calls belong on the NFL picks surface. Lens construction and remixing belong on the Build surface. The lens is the contract connecting all three.
