AnomIQ

Documentation

Metrics Glossary

Unlike standard screeners that rely on lagging indicators and arbitrary candle closes, AnomIQ processes raw WebSocket execution data in real-time. Here is exactly how we calculate our data.

Rolling Window Filters (5m / 15m / 60m)

Full documentation for all rolling-window filters, including Total/Buy/Sell variants, Z-scores, Whale Score, return and volatility on next page.

Open Rolling Window Metrics →

Zero-Lag Rolling Windows

#

All timeframe-specific metrics (5m, 15m, 60m) are calculated using sliding rolling windows. For example, a 5-minute metric is composed of the last 4 completely finished 1-minute bars plus the current active bar, updating tick-by-tick.

Why this matters
No artificial “reset” at candle close → fewer false negatives, and faster detection of real regime changes in flow.
Common mistakes
Comparing rolling 5m metrics to candle-based indicators 1:1. They don’t update on the same clock, so signal timing differs.
----

Global & Daily Baselines

----

These metrics provide the macro context for an asset over the current UTC day.

Today Volume Notional (UTC)

#

Total traded notional accumulated since 00:00 UTC for the current day.

UI field: Today Volume in $

Calculation
Sum of executed trade notional since 00:00 UTC: Σ(price × size)
Range / Units
$ (quote currency notionals). Non-negative, unbounded.
Interpretation
Higher means more capital has already flowed today → generally better fills, less slippage, more reliable anomalies.
Trading Application
Filters out low-cap noise. A common floor is > $500,000 so the asset can support your trade size.
Common mistakes
  • Forgetting this is UTC-based (your local day boundary may differ).
  • Using it alone as a “trend” proxy (it’s activity, not direction).

Yesterday Volume Notional (UTC)

#

Total traded notional for the previous UTC day (00:00–24:00 UTC). Used as a baseline for “what is normal.”

UI field: Yesterday Volume in $

Calculation
Sum of executed trade notional from yesterday 00:00 UTC to 24:00 UTC: Σ(price × size)
Range / Units
$ (quote currency notionals). Non-negative, unbounded.
Interpretation
Establishes baseline participation. A “hot” day is only meaningful relative to what that coin normally does.
Trading Application
Use it to avoid chasing optical illusions: a coin doing $200k today sounds fine until you realize it did $20M yesterday.
Common mistakes
  • Comparing today’s partial-day volume to yesterday’s full-day volume without normalizing by time-of-day.
  • Assuming yesterday is “typical” on news days (it might not be).

Relative Volume vs Yesterday (same UTC time)

#

Compares today's volume to yesterday's volume at the same UTC time-of-day.

UI field: Volume Ratio Relative

Calculation
(today_volume_since_00:00_UTC / yesterday_volume_until_same_UTC_time) × 100
Range / Units
%. Typical values cluster around 60–140% on “normal” coins; spikes can exceed 300%+.
Interpretation
Normalizes away daily routines. A high value means the coin is unusually active right now, not just “active today.”
Trading Application
A common trigger is > 150% which means it’s trading significantly hotter than it usually does at this exact hour.
Common mistakes
  • Using it without a liquidity floor (thin coins can “spike” on tiny absolute flow).
  • Ignoring context days (news/airdrops/halts make yesterday a bad baseline).

Liquidity Score (0–100)

#

Quality score representing 24h trade consistency and continuity. Higher = steadier executions, lower slippage risk.

UI field: Liquidity score 0-100

Calculation
Aggregates trade continuity over 24h (e.g., how many minutes have executions + how long the typical “dead gaps” are), then maps to a 0–100 score. Exact weights are internal, but the interpretation is stable.
Range / Units
0–100 (unitless).
Interpretation
100 → trades basically every minute. < 30 → illiquid / “dead” coin.
Trading Application
Keep scanners safe from slippage: a common floor is > 80.
Common mistakes
  • Assuming “high liquidity score” means “safe price action” (it only reduces execution risk).
  • Using it without notional volume context (steady micro-volume can still be untradeable size-wise).

Micro acceleration (5m vol / 15m vol)

#

Measures immediate momentum shifts by comparing 5-minute volume intensity to the 15-minute baseline. High values indicate a rapid burst of activity.

UI field: Micro acceleration 5min vol / 15min vol

Suggested floor: ≥ 150%

Calculation
First compute intensity ratios for 5m and 15m:
ratio_w = ((currentSum_w + live) / (baselineMean_w × w_minutes)) × 100
Then: micro = (ratio_5m / ratio_15m) × 100
Range / Units
%. ~100% means 5m intensity is roughly in line with 15m intensity; higher means 5m is disproportionately “hot”.
Interpretation
150% means the 5m intensity is ~1.5× the 15m intensity (a fast burst). It’s a short-term “ignition” detector.
Trading Application
Use to catch fresh demand spikes early. Pair it with liquidity/notional floors so thin coins don’t fake it with tiny absolute flow.
Common mistakes
  • Using micro acceleration alone (it’s timing, not direction).
  • Letting illiquid coins dominate results (always combine with liquidity score / notional floors).
  • Forgetting it’s relative-to-baseline intensity, not raw volume.

Macro acceleration (5m vol / 60m vol)

#

Measures breakout strength by comparing 5-minute volume intensity to the 60-minute baseline. Detects when short-term action diverges from the hourly trend.

UI field: Macro acceleration 5min vol / 60min vol

Suggested floor: ≥ 110%

Calculation
Compute intensity ratios:
ratio_w = ((currentSum_w + live) / (baselineMean_w × w_minutes)) × 100
Then: macro = (ratio_5m / ratio_60m) × 100
Range / Units
%. ~100% means 5m intensity matches the 60m regime; higher means the last 5m is unusually intense vs the hour.
Interpretation
Think “breakout pressure.” The bigger the number, the more the last 5 minutes diverge upward in activity from the hourly norm.
Trading Application
Useful for spotting sudden regime shifts (news, listing, liquidation cascades). Combine with direction/order-flow metrics to avoid chasing noise.
Common mistakes
  • Assuming macro acceleration implies trend continuation (it can be a blow-off top).
  • Ignoring baseline validity during outages/halts (baselineMean may be distorted).
  • Not filtering by minimum liquidity/notional.

Trend set acceleration (15m vol / 60m vol)

#

Measures sustained momentum by comparing 15-minute volume intensity to the 60-minute baseline. High values indicate the move is solidifying into a trend.

UI field: Trend set acceleration 15min vol / 60min vol

Suggested floor: ≥ 110%

Calculation
Compute intensity ratios:
ratio_w = ((currentSum_w + live) / (baselineMean_w × w_minutes)) × 100
Then: trend_set = (ratio_15m / ratio_60m) × 100
Range / Units
%. ~100% means 15m is behaving like the hourly regime; higher suggests the move is persisting long enough to matter.
Interpretation
This is your “trend confirmation” acceleration. If micro is ignition and macro is breakout pressure, trend-set is “okay, this isn’t just a 2-minute sneeze.”
Trading Application
Great for separating one-candle spikes from sustained participation. Combine with price-move filters to avoid high-volume chop.
Common mistakes
  • Expecting it to trigger “early” (it’s intentionally slower than micro).
  • Using it as a buy signal without checking direction (it measures activity persistence, not bullishness).
  • Not controlling for low absolute volume environments.