Numerai settled staking return distributions
Settled model-round NMR return distributions by historical stake size across Classic, Signals and Crypto, with 20- and 60-round windows.
Coverage & dates
- Record date range
- to
- Date meaning
- Opening dates of the rounds in the published windows. Settlement dates are separate. The range can include gaps; it does not imply complete observations throughout.
- Collected
- · 01:36:04 UTC
- Prepared by
- nmrdash, from the public sources below.
Sources & methods
Settled NMR payout divided by positive historical round stake, one observation per model-round. Stake-size cohorts are re-ranked in each round using nearest-rank 50th and 90th percentile cutoffs; ties stay in the lower cohort. Return percentiles use linear interpolation and equal position weights. Profitable means strictly positive settled payout. Windows use the latest payout-resolved rounds of the most recent resolved staking regime, without mixing legacy and atomic rounds. Aggregate principal is a sum of round stakes, not unique capital; returns are neither compounded nor annualized.
The largest group is above the 90th stake percentile, the middle group is above the 50th through the 90th, and the small group is at or below the 50th. Ties stay in the lower group.
Each model-round has equal weight in return percentiles. Windows overlap and must not be added together. Legacy and atomic regimes remain separate; pending payouts are excluded.
Return fields are ratios: 0.01 means 1%. Net NMR payout and stake-weighted return are different from the median position return.
JSON fields
Paths use [] for array items and * for metric keys such as corr20 and corr60. Absent fields or null values mean unavailable information; a numeric zero remains a real observation.
| Field | Meaning | Unit |
|---|---|---|
fetched_at | Collection timestamp in UTC. This does not establish the observation date or score finality. | Timestamp |
definition_version | Version of the dataset’s measurement and export definition. | Text |
tournaments[].id | Tournament identifier: 8 = Classic, 11 = Signals, 12 = Crypto. | See definition |
tournaments[].atomic | Whether the published windows use atomic staking. | See definition |
tournaments[].pending_atomic_rounds | Number of atomic rounds still awaiting payout resolution. | rounds |
tournaments[].windows[].requested_rounds | Requested window size: 20 or 60 resolved rounds. | rounds |
tournaments[].windows[].rounds | Actual number of resolved rounds in the window. | rounds |
tournaments[].windows[].first_open | Opening timestamp of the first included round. | See definition |
tournaments[].windows[].last_open | Opening timestamp of the last included round. | See definition |
tournaments[].windows[].last_resolved | Latest settlement timestamp in the window. | See definition |
tournaments[].windows[].cohorts[].key | Stake-size group: large, middle or small. | See definition |
tournaments[].windows[].cohorts[].positions | Number of model-round positions in the cohort. | positions |
tournaments[].windows[].cohorts[].profitable | Positions with strictly positive settled payout; zero is unprofitable. | positions |
tournaments[].windows[].cohorts[].profitable_share | Fraction of cohort positions with positive settled payout. | fraction (0–1) |
tournaments[].windows[].cohorts[].p10 | 10th percentile of payout ÷ position stake, using equal position weights. | return ratio (1 = 100%) |
tournaments[].windows[].cohorts[].p25 | 25th percentile of payout ÷ position stake, using equal position weights. | return ratio (1 = 100%) |
tournaments[].windows[].cohorts[].median | Median of payout ÷ position stake, using equal position weights. | return ratio (1 = 100%) |
tournaments[].windows[].cohorts[].p75 | 75th percentile of payout ÷ position stake, using equal position weights. | return ratio (1 = 100%) |
tournaments[].windows[].cohorts[].p90 | 90th percentile of payout ÷ position stake, using equal position weights. | return ratio (1 = 100%) |
tournaments[].windows[].cohorts[].stake_min | Smallest position stake in the cohort; decimal string or null. | NMR |
tournaments[].windows[].cohorts[].stake_max | Largest position stake in the cohort; decimal string or null. | NMR |
tournaments[].windows[].cohorts[].stake_nmr | Sum of round position stakes; this can reuse capital across rounds. | NMR |
tournaments[].windows[].cohorts[].earned_nmr | Sum of positive settled payouts, as an exact decimal string. | NMR |
tournaments[].windows[].cohorts[].burned_nmr | Positive magnitude of negative settled payouts, as an exact decimal string. | NMR |
tournaments[].windows[].cohorts[].net_nmr | Earned minus burned NMR, as an exact decimal string. | NMR |
tournaments[].windows[].cohorts[].weighted_return | Total net payout divided by total round position stake. | return ratio (1 = 100%) |
Download
Download staking-returns.json · application/json
Definition version: numerai-resolved-stake-size-returns-v1. Decimal strings preserve exact amounts; use decimal arithmetic when working with NMR or USD totals.
See the release manifest for this file’s checksum and byte count. The download is public and needs no authentication or JavaScript.
Interactive charts can load newer snapshots. Compare collection and observation dates before combining a download with chart values or another release.