Wallets with perfect win rates and no sell-offs look suspicious, yes, but this isn't a decisive trigger for insider trading every time.
What if the prediction market is tiny, and someone with a large enough size can actually influence the outcomes just by placing orders?
They might not have insider info, just a market-moving position; this invalidates the easy assumption.
We need to put liquidity on the watchlist before calling it insider activity, like checking if the local market for second-hand strollers is so small that one buyer can set the price.