Five dimensions. Weighted. Benchmarked against $30B+ in anonymized real LP investment data across 432 GPs. Every offering on this page is a private offering, scored against a rubric specific to its offering type. There is no generic fallback.
Every private offering is classified along 4 axes before scoring begins. The rubric the IC uses is specific to the offering type. A multifamily value-add offering and a seed-stage SaaS bet are not scored against the same questions.
Equity, debt, or hybrid. Based on where you sit in the capital stack, not the marketing language. A "preferred equity" tranche behaves like debt and gets scored that way.
Fund, single-asset, co-investment, feeder, or fund-of-funds. Based on how many underlying offerings the vehicle holds. A "Fund II" that buys one apartment building is scored as single-asset.
Real estate, venture, growth equity, buyout, crypto, oil and gas, infrastructure, consumer, manufacturing, life sciences, agriculture, other.
Within asset class. Real estate breaks into core, core-plus, value-add, opportunistic, development, stabilized for equity, and bridge, mezz, preferred equity, construction for debt. Operating companies break into seed bet, growth equity, buyout LBO, roll-up.
If a deck is pitched as "core-plus" but target IRR is 15% and LTV is 70%, the IC scores it as value-add. Sponsors often soften the label for positioning. We don't let that bias the rubric.
Every offering is scored across the same five dimensions. The weights are fixed. Dimensions where you disclose nothing get dropped and the remaining weights renormalize.
Rollup math. Within a dimension, parameter scores are averaged equally. The five dimension averages are then combined with the weights above. An offering scoring 4 / 4 / 4 / 4 / 4 across all dimensions earns a 4.0 overall. An offering that's strong on Returns and Structure but weak on Sponsor will not clear 4.0, because Sponsor carries the largest weight.
The simplest framing: a 4 or higher means at-or-above market on every dimension that matters. Below 4 is not a no, but it doesn't get distributed.
Top-quartile sponsor. Returns above benchmark with conservative assumptions. Hard-to-replicate market position. LP-friendly structure. Minimal risk. Rare.
Solid sponsor with full-cycle track record. Returns at or above benchmark midpoint. Healthy market. Terms in line with norms. Manageable risk. This is what gets distributed to the 12,568 LPs.
Below benchmark midpoint, or missing some disclosure, or market-standard structure with no edge. Worth a deeper look. Not a no.
Sponsor thin on track record, returns below benchmark, soft market, terms tilted to GP, or elevated risk. Defensible only if the rest of the offering compensates.
First-time sponsor with no relevant experience. Returns below stress floor. Market in decline. Predatory terms. Or a hard-rule breach (DSCR < 1, LTV > 85, etc.).
One of the most-submitted offering types. 33 scored parameters across the five dimensions, plus three bonuses. Other offering types follow the same structure with parameters tuned to the asset class. Param counts range from 28 to 37 depending on offering type.
This is the routing the IC ran. Click any dimension below to see its scored parameters.
Bonus: institutional backing.
Bonuses: refi liquidity (+1.0), secondary liquidity (+0.5).
Same five-dimension structure. Different parameters per asset class and strategy. Your rubric is selected after classification.
See full coverage matrix →Every offering type below has its own calibrated rubric. If your offering classifies outside this matrix, the system flags it before scoring. We will not score an offering type we have not built a rubric for.
Crypto: only liquid L1 and DeFi. No locked, seed-round, validator, or yield strategies. No L2, infra, or app-layer. Natural resources: oil only. No royalty, midstream, mining, timber, or renewables. Infrastructure: one fund-level offering type. No single-asset infrastructure or other strategies. Debt: real-estate only. No operating-company or infrastructure debt.
If your offering lands in one of these gaps, classification flags it. We add rubrics on a per-submission basis when demand justifies the calibration work.
Most parameters are scored on the 1 to 5 scale. A few breach conditions cap the parameter or dimension regardless of the rest of the offering.
We publish the parameter names, the dimension weights, the band meanings, and the hard caps. We do not publish the exact benchmark ranges, the skepticism thresholds, or the silent-fallback values.
If we did, decks would be tuned to them. The integrity of the IC depends on the GP not knowing what number it would consider a 4.
Verdict within 48 hours. Distribution earned by score, not payment.
Submit to the IC