I analyzed 250 top Polymarket politics wallets. 0 were safe to blindly copy.
But many were still useful.
That changed how I think about wallet leaderboards.
The useful output is not:
“Here are wallets to copy.”
It is:
• which wallets to ignore completely
• which wallets are worth studying
• what kind of signal they provide
• what the key risks are if you blindly copy them
• what rules would make them safer to use
1. Which wallets to ignore completely
Some wallets looked too hard to use safely:
• timing-dependent edge
• one-win PnL concentration
• automated/high-frequency activity
• open or unresolved gains
• no obvious manually repeatable strategy
Good trader ≠ good copy target.
2. Which wallets are worth studying
There were still real signals:
• 22% looked worth studying further
• 11% were building directional positions
• 10% had cleaner PnL profiles
• 7% were useful for market/theme discovery
• 1% were finding longshot/asymmetric ideas
So the signal was not zero.
It just was not “copy every trade.”
3. What signal they provide
Some wallets are useful for finding markets.
Some for spotting accumulation.
Some for studying repeatable strategies.
Some for finding cleaner PnL profiles.
Some only for longshot ideas.
Each should be used differently.
4. What risks come from blindly copying
The same scan found:
• 94% had high copy-trading risk
• 87% had low manual copyability
• 68% had execution/automation risk
• 48% looked like poor signal sources
Copying can fail because:
• they entered earlier
• the market already moved
• they trade too fast
• the edge relies on automation
• PnL came from one big win
• gains are still unresolved
5. What rules make it safer
For wallets that still look useful, I’d want guardrails:
• only copy accumulators near their average entry
• avoid chasing after large moves
• use high-frequency wallets for discovery, not copying
• discount unresolved/open PnL
• size longshot ideas very small
• avoid one-win wallets
• check whether the visible wallet shows the full strategy
So the better question is not:
“Who made money?”
It is:
“What is this wallet useful for, what could go wrong, and what rules stop me copying it badly?”
I put together a small wallet breakdown using this framework.
Happy to share if useful.