Everyone in private markets tells you what a company is worth today. This board says what happens next — a sealed weekly probability that each company's QL Mark reprices higher over the next 90 days — and then grades itself in public when the horizon elapses. Baselines are stamped at seal time; sealed claims are never edited; the Brier score and calibration table are published win or lose.
Each call carries its full driver breakdown at issuance — the gap between the freshest secondary print and the blended mark (amplified by round staleness), the sector tide, and any thin-evidence shrink. Deterministic, disclosed, no black box. Sorted by decisiveness.
Keep only the calls this board is most decisive about — those at least this far from a coin flip — and see what the sealed probabilities imply before a single one is graded. This projects the board's own stated odds; it is not a track record. The first real grades land at the horizon.
| Company | Sector | P(mark ↑ in 90d) | Baseline QL Mark | Drivers |
|---|---|---|---|---|
Loading the sealed board… | ||||
Four interactive simulations drawn only from this board's own sealed claims and disclosed drivers. None of them is a record: every number here is a projection, and the grader alone writes the record — first on the horizon.
Grade this board once under an assumed reality and watch the win rate the grader would publish. Honesty bends the truth behind the claims: at 1.0 the board is perfectly calibrated, at 0 its calls are noise, above 1 the world is more extreme than it dares to say.
Pick a company (or click any row of the board). The fan draws 120 hypothetical 90-day mark paths whose drift is set so the share finishing higher equals the sealed probability — the odds are the board's, the wiggle is an assumed volatility, and the grey band is the ±2% push zone where the grader decides nothing.
Every call discloses its two drivers — the pull of the freshest secondary print against the blended mark, and the sector tide — plus any thin-evidence shrink. Re-weight them and watch where each call would have landed. The sealed board never moves; this is the what-if.
Where each sector's sealed calls sit on the 0–100% axis. Bubble size is the number of calls; the tail runs from a coin flip to the cohort's mean, so a long tail is a strong tide.
Aggregate only — no company is named here, so this map is the same at every tier.
When we say 70%, does it happen ~70% of the time? This table fills as boards grade — it is the product.
| Claimed band | Decisive calls | Realized ↑ |
|---|---|---|
No graded boards yet. | ||
The QL Mark is a model-marked estimate, not a quoted price; private-market data is sparse and every call carries stated evidence quality. Nothing here is investment advice, an offer, or a solicitation. Past record ≠ future results.
The first forward-graded record for private markets: weekly sealed probabilities that a private company’s QL Mark reprices higher over 90 days, graded automatically at the horizon, Brier-scored and calibration-tabled in public.
It is the model’s estimate frozen before the outcome. The public calibration table shows how often its X% claims actually resolved — the record is the product.
They are model outputs with a published accuracy record, not advice. Read the calibration table before reading any single number.