Outperformance Score (OPS)
A five-term blend of the Moonboard Score with sentiment, turnover, social scale and price stability, expressed as a rank within the scanned universe.
- Scale
- percentile Relative scale — the value states rank within the scanned universe, not magnitude.
- Formula
- 0.30·MB + 0.20·Sentiment + 0.20·Liquidity + 0.20·Social + 0.10·Stability → percentile
- Inputs
- moonboardScore
- sentiment
- volume24h
- marketCap
- socialVolume24h
- interactions24h
- socialDominance
- volatility
"Outperformance" describes the intent behind the metric, not a measured property. OPS contains no return data, no benchmark and no forward-looking term. It cannot and does not establish that a high-ranking asset outperformed anything, and a high OPS is not a prediction that one will. What it measures is stated plainly below.
Purpose
The other Moonboard metrics each isolate one dimension. OPS exists for the opposite reason: to produce a single ordering when you do not want to choose a dimension. It asks which assets look reasonable on several axes at once — discussed more than their size implies, positively toned, actually traded, socially visible, and not violently volatile.
It is deliberately a compromise metric. Nothing scores highly on OPS by excelling at one thing.
Inputs and construction
A raw composite is built first, then converted to a percentile. The composite is never displayed; only the rank is.
| Term | Weight | Range fed in | What it contributes |
|---|---|---|---|
| Moonboard Score | 0.30 | 0–100 | Size-relative attention, already bounded |
| Sentiment | 0.20 | 0–100 | Tone of the conversation |
| Liquidity | 0.20 | 0–100 after cap | Turnover, capped at 10 % of market cap |
| Social scale | 0.20 | 0–100 after cap | Absolute social footprint |
| Stability | 0.10 | 0–100 | Inverse of volatility |
The two caps behave very differently
Both the liquidity and social terms are capped at 100, but they saturate at opposite ends of the universe.
- Liquidity saturates at 10 % daily turnover. Most assets sit well below that, so this term genuinely discriminates across the normal range.
- Social scale saturates at a combined footprint of 1 000 000. Because
interactions24hdominates that sum and routinely reaches millions for well-known assets, the term is pinned at 100 for the entire large-cap segment and only separates smaller assets. Within the top of the market it contributes a constant 20 points and therefore no ordering at all.
The socialDominance × 1000 scaling inside that sum is also worth noting: with dominance typically below 5, that term adds at most a few thousand against interaction counts in the millions. Its practical weight is far smaller than its position in the formula suggests.
Normalisation
The composite mixes bounded and capped quantities on a shared 0–100 scale, which makes the weighted sum meaningful. It is then converted to a cross-sectional percentile — the share of assets with a strictly lower composite.
This final step discards magnitude. Two assets separated by a hair in composite terms can sit ten percentile points apart if the field is densely packed there, and two assets far apart in composite can rank adjacently if the field is sparse. OPS tells you order, never distance. See percentile ranking.
Interpretation
| OPS | Reading |
|---|---|
| ~20 | Bottom fifth of the composite ordering. Usually a weak Moonboard Score combined with thin turnover. |
| ~50 | Median. Typically strong on some terms and weak on others. |
| ~70 | Upper third — no single term collapsed. |
| ~90 | Top decile of the blend. Reliably means high Moonboard Score plus real turnover, since those are the two terms that still discriminate at the top. |
Worked example
Hypothetical asset, using Example A from the Moonboard Score page:
Note how little the social term contributes here — 1.94 of a possible 20 — because this asset's absolute footprint is small even though its size-relative footprint is exceptional. That size-relative strength already entered through the Moonboard Score term. The two social-flavoured terms in OPS measure genuinely different things, and only one of them rewards small assets.
Limitations
- No return data. Despite the name, nothing here measures or forecasts performance.
- Correlated inputs. The Moonboard Score already contains sentiment, turnover and social activity, so those inputs enter twice at different weights. The effective weighting is not the stated weighting.
- The social term is dead at the top of the market. Constant at 100 for most large caps, contributing no ordering there.
- Stability rewards inactivity. A dormant asset with near-zero volatility scores maximum stability, and the term cannot tell calm from abandoned.
- Universe-dependent. The percentile shifts when the scanned set changes, with no change in the asset.
- The weights are asserted, not fitted. They were chosen for interpretability. A different split would be equally defensible and would reorder the results.
Data quality
A zero or missing market cap returns 0 rather than an infinite liquidity ratio. Missing social fields are read as zero, depressing the social term. Because the Moonboard Score carries 30 % of the weight, anything that voids that score — chiefly a missing marketDominance — propagates directly into OPS.
Related
- Moonboard ScoreThe heaviest single term in OPS
- CORThe other multi-factor metric, standardised rather than weighted
- LMRTurnover on its own
- Percentile rankingWhy the final step discards magnitude