Cohort Overview

Cohort Overview is the perfect feature if you want to understand the market through groups of Smart Traders, instead of analyzing individual wallets.

It helps you:

  • Identify where different trader groups are positioning
  • Understand market sentiment across cohorts
  • Gain insights faster with aggregated behavior

How to use this feature?

To start, click here (Cohort Overview) or go to ‘Perps’ → ‘Cohort Overview’.

To use this feature effectively, focus on 3 key areas:

  • Filtered by win rate, PnL performance, and capital size
  • Filtered by Position age
  • Understand the Data Panels

[1] Filtered by win rate, PnL performance, and capital size

This filter allows you to group traders into cohorts based on Winrate (skill level), PnL performance (profitability), Capital size (wallet size: shrimp, fish, whale, etc.),

With it, you can instantly spot insights at a glance:

  • Overall market exposure across different tokens
  • Clear directional bias (long/short) for each specific cohort
  • Defined sentiment on individual tokens so you can catch the rotation

To fully understand this section, make sure you read the cohort icons, as each icon represents a specific trader group and behavior profile.

[2] Filtered by Position age

This filter lets you analyze positions based on how long they’ve been open. And it works together with [1], this helps you understand:

  • Timing of entries
  • Whether a trend is new or already crowded

[3] Understand the Data Panels

Once filters are applied, you’ll see 3 main sections:

#1: Overview

Here, you get a quick snapshot of a cohort.

Example insights:

  • Shrimp cohort (equity: 0 to $250) is Very Bullish with 58.8% is long and exposure of 8.26x
  • 2.93% in position 19.7K/672.44K wallets
  • Performance: Loss

#2 Detail

This section gives you deeper insights into what the cohort is doing:

  • Top 10 tokens the cohort is opening positions in
  • Liquidation risk
  • Opened largest positions and which tokens they focus on

#3 Top wallet

You’ll need: