ICX Moment Score v3.0: adding the card itself to the equation

The formula

Every exact card is now scored with the same equation: ICX = 0.15A + 0.15M + 0.30C + 0.25S + 0.15L. A is athlete strength, M is moment durability, C is card role, S is verified scarcity and L is exact-market liquidity. Each component runs from 0 to 100; the weighted result is rounded to the nearest whole number.

The model does not use asking-price level, a later appreciation multiple or an analyst’s desired outcome. A card cannot gain C points because it already became expensive. Card role uses chronology visible from the exact identity: origin, RC, inaugural-set anchor, major first or standard issue.

0.15A + 0.15M + 0.30C + 0.25S + 0.15L = ICX Moment Score

Why v2.2 missed some winners

We audited 41 exact cards with captured launch-price comparisons. Among the cards at 2× launch or better, 13 still scored below 60 under v2.2. Nine of those low scores were Living Set portraits or early origin cards whose market importance came from chronology and scarcity rather than a dated sporting event.

Kimi Antonelli #14 was the cleanest modern example. It was his RC-marked Formula 1 debut and had ten exact sales, yet a 60 athlete tier and 60 moment tier held the score at 59. Messi Living Set #1 opened an entire program but received only 20 for the sporting moment because it is a portrait. Both observations were mathematically correct under the old definition and strategically incomplete. Version 3.0 moves them to 71 and 72 respectively without using their later prices as inputs.

The error was not arithmetic. The model was measuring story strength while the user expected opportunity strength.

What the larger sample changed

The expanded audit now contains 49 base cards with at least two completed-sale observations: 40 at 2× launch or more and only nine below that line. On those 49, v3.0 records a 0.478 pairwise ranking result at the 2× threshold. Scarcity alone reaches 0.694. Reweighting the same five components cannot turn this selected sample into a reliable forecasting model; the best constrained alternative reaches only 0.503.

The useful patterns are narrower. Low exact supply repeatedly matters. Origin cards and difficult-to-replace identities matter. Athlete re-rating matters, but using a player’s later career status to explain an old card is hindsight, not prediction. Wembanyama VW-4 also shows that visual identity can separate one card from nearby issues with the same athlete, RC mark and similar supply. That deserves a blinded image rubric before it enters the formula.

The sample also mixes cards released in 2016 with cards released in 2025. A raw multiple gives the older card far more time to rise. The next calibration set therefore needs a fixed holding period, at least 100 labeled cards, 30 below-2× controls and a primary cohort with five or more exact sales. Until those gates are met, the honest version decision is HOLD v3.0.

The larger dataset improved the method by stopping an unjustified formula change.

A · Athlete strength · 15%

The athlete receives one of five published tiers: emerging or specialist subject (20), breakthrough prospect (40), established international star (60), era-defining champion (80), or all-time global icon (100). The tier reflects achieved career record and durable international relevance, not social-media momentum for a single week.

Prospects can rise when achievements become real. Each mapped athlete has one sitewide tier, review date and evidence trail; individual card records cannot override it. The 15% weight prevents current fame from overwhelming card-specific opportunity.

M · Moment durability · 15%

The moment uses five tiers: episode or routine result (20), recognizable career chapter (40), major milestone or breakthrough (60), career-defining title or record (80), and sport-history landmark (100).

The test is biographical: would this achievement survive in a serious career timeline? A World Cup title, first major, career record or defining debut can pass. A routine goal or attractive celebration cannot become important merely because its card is scarce.

C · Card role · 30%

Card role asks a different question: how replaceable is this card in the athlete or program chronology? Tier 1 is secondary or replaceable; Tier 2 is a standard issue; Tier 3 is a distinct chronology or set anchor; Tier 4 is a major first, RC or defining image; Tier 5 is an origin or flagship identity.

An RC logo does not guarantee appreciation, and a first does not erase a huge print run. C simply stops the model from treating a debut card and a routine mid-career card as identical when their sporting-moment tier happens to match. Living Set portraits receive chronology credit for their one-per-player and program-position role, not for a match that never occurred.

C is assigned from identity and chronology known about the card, never from what its price did later.

S · Verified scarcity · 25%

Scarcity is calculated from the exact-format print run N on a logarithmic curve: S = clamp(100 − 18 × log10(N ÷ 10), 0, 100). That gives 100 points to a /10, approximately 82 to 100 copies, 64 to 1,000, 46 to 10,000 and 28 to 100,000.

Only publisher-confirmed or serial-numbered exact supply counts. If the final run is pending, S remains unresolved rather than being recorded as zero. A base print run does not establish the population of a PSA 10, so a graded card keeps S pending without exact population evidence.

L · Exact-market liquidity · 15%

Liquidity has three mechanical parts: deduplicated exact-sale volume, current normalized exact availability and recency of the latest exact sale. But L is not confirmed until at least two completed sales match the exact card, parallel and format. Zero or one sale produces a provisional range.

Active listings describe supply only. Once the two-sale threshold is met they can add availability context, but they cannot confirm liquidity by themselves. Price level never enters L: a $500 ask and a $5 ask are both simply one listing.

Unknown is not zero: when ICX publishes a range

A new release should not be punished because the publisher has not posted its final run or copies have not reached the resale market. When S is missing or L has fewer than two exact sales, ICX publishes a provisional Moment Range instead of a fixed score. Known components remain fixed; each unresolved component is allowed to vary from 0 to 100. We do not publish a midpoint because that would disguise an assumption as evidence.

The card page then shows print-run sensitivity at exact supply of 100, 1,000, 10,000 and 100,000 copies. If liquidity is also pending, each print-run row remains a range. Once all five inputs meet their evidence threshold, the range collapses to one confirmed score.

Moment Score v3.0 compares collectible quality; it is not a probability of appreciation. The separate ICX Watch Priority on Market Pulse uses only facts available at release: athlete demand 20%, card identity 30%, scarcity 30%, moment 10% and source completeness 10%. It never adds points because a card later became expensive and is not a price forecast. Entry price, fees, condition, market depth and expected return remain separate purchase work.

• One five-part formula is used across every indexed card
• Card role recognizes origins, RC issues and inaugural anchors without using later price performance
• Unknown scarcity or fewer than two exact sales produces a range, not a zero
• No estimated midpoint is published
• Asking-price level never raises the score
• Every component and weighted contribution is visible
• Moment Score measures collectible quality; Watch Priority is a separate release-time research screen
• The formula will not be reweighted until the calibration cohort has enough true controls and equal holding periods