Numerai Medals and Grandmasters: How the Ranking Works
How Numerai awards gold, silver, and bronze medals each round, the MMC thresholds needed to earn them, and why consistency beats peak scores.
The top 10% of medal-earning models hold 35% of every medal ever awarded — about 130,000 out of 371,000. The median medal winner has 7; only the 99th percentile clears 54. Numerai's medal system funnels recognition toward a small group of long-tenured, consistent models, and the distribution is steeper than most participants assume.
How are medals distributed across rounds and models, what MMC (meta-model contribution) thresholds do you need, and does consistency predict medal counts better than peak scores?
The stored medal columns in our database are empty, so the charts below compute medals directly from raw MMC using fixed per-round cutoffs: gold for the top 1%, silver for the top 3%, bronze for the top 15%.
How many medals per round?
Total medals per round rise and fall with field size.
Bronze dominates each round, with silver and gold as thinner bands on top. Totals climbed from near zero around round 200 to roughly 2,000 per round by round 800, then dropped sharply to about 500 around round 820 (a field-size contraction) before recovering toward 1,800 by round 1,200. Because the cutoffs are fixed percentiles, the tier split stays constant and only the absolute count moves with the number of submitting models.
What score do you need?
The minimum MMC for each tier moves with the field's performance distribution.
The percentiles are fixed, but the raw MMC they correspond to swings round by round. Gold (top 1%) typically lands between 0.02 and 0.04 MMC, silver (top 3%) between 0.015 and 0.025, and bronze (top 15%) between 0.01 and 0.02. Early rounds (before round 300) were noisier, with gold thresholds spiking to 0.06. Medals are relative: you beat your peers, not an absolute bar.
The gap between bronze and gold is smaller than you might expect. Often just 0.01 to 0.02 MMC separates a bronze-qualifying score from a gold one, so small score differences produce very different medal outcomes.
The medal distribution across models
How are total medal counts spread across all models?
The distribution is heavily right-skewed on a log scale. Out of about 33,000 models that ever earned a medal, the median model has 7. The 95th percentile sits at 36; the 99th at 54; only a handful clear 80, and the single highest count is 118. The top decile of medal winners alone holds 35% of all medals awarded — longevity (more rounds, more chances) and skill (winning round after round) compound into a power-law tail visible on any individual model's performance history.
Consistency vs brilliance
Does the system reward steady above-average performance or occasional exceptional rounds?
Total medal count rises with average MMC. Models clustered near zero average rarely clear 20 medals, while the 80-plus medal tier is almost entirely models with positive average MMC (roughly 0.005 and above). The color gradient (rounds played, purple at the low end up to yellow at 300+) shows the biggest hauls come from long-tenured models that also maintain a positive mean. High rounds played without a positive average produces middling counts at best.
A model that reliably lands just inside the top 15% will pile up bronze medals steadily. One that alternates between the top 1% and the bottom half earns fewer medals overall, because below-average rounds contribute nothing. Percentile thresholds convert repeated above-average rounds into repeated medal accumulation.
Takeaways
Medals are relative, not absolute. Thresholds shift with the field. You earn medals by outperforming peers.
Longevity is non-negotiable. Every model with 80+ medals has played at least 219 weekly rounds — nearly the entire tournament's weekly history. With 220-round survival probability at just 38%, the median Grandmaster contender washes out before they can accumulate enough medals to qualify. See model survival.
Consistency beats isolated spikes. Average MMC and rounds played explain the visible pattern better than occasional exceptional rounds. This connects to MMC's role in payouts and the meta-model's need for stable inputs.
The medal system rewards repeated cutoff clearing. Longevity and stable positive MMC matter more than occasional spikes. For the round-level data behind these charts, browse the rounds list or the trends dashboard.