The four forms
Spoofing posts large orders there is no intention to fill, to make one side of the book look strong, then cancels them once other traders react. Wash trading fakes volume between related wallets. Cornering accumulates enough of the outstanding contracts to control the exit price. Outcome interference tries to change the real-world event itself.
The first three are trading behavior visible in market data. The fourth is not a market problem at all, and it is the one that keeps some event categories off regulated venues entirely.
What manipulation looks like in the data
Spoofing shows as repeated large orders that appear, move the quoted price, and cancel without filling. Cornering shows as open interest concentrating into a small number of wallets while volume stays low.
None of these patterns is proof on its own. A trader who cancels frequently may be managing risk in a fast market, and a large holder may hold a strong view. The useful output is a ranked anomaly with the underlying trades attached.
Where 0xinsider flags manipulation
The suspicious-trade feed lists each flagged trade with its wallet, market, and time, and the API adds the suspicion score and the pattern that triggered it. That framing is deliberate. A flag is a prompt to look, not a finding of wrongdoing.
