When a recorded order book doesn't explain a trade, I'd first want to know what was being recorded.
Bybit has two different snapshots worth distinguishing. Its standard one leaves out Retail Price Improvement orders, a separate class of orders on the exchange. The other reports their size alongside ordinary orders. Saving more price levels from the first snapshot won't turn it into the second.
For someone replaying a market, that changes the investigation. Before adjusting simulated fills to match a real trade, check whether the recording includes the kind of orders you're trying to account for.
The separate snapshot has exclusions too, so it isn't a complete record of every order. Neither snapshot, on its own, tells you why a particular trade filled or whether your order could have matched those quotes. But checking the source tells you which categories of orders your replay could see in the first place.
Bybit has two different snapshots worth distinguishing. Its standard one leaves out Retail Price Improvement orders, a separate class of orders on the exchange. The other reports their size alongside ordinary orders. Saving more price levels from the first snapshot won't turn it into the second.
For someone replaying a market, that changes the investigation. Before adjusting simulated fills to match a real trade, check whether the recording includes the kind of orders you're trying to account for.
The separate snapshot has exclusions too, so it isn't a complete record of every order. Neither snapshot, on its own, tells you why a particular trade filled or whether your order could have matched those quotes. But checking the source tells you which categories of orders your replay could see in the first place.
